A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | R | S | T | U |V | W | X | Y | Z
Optimal Portfolio
Control for Parabolic Type Infinite-dimensional Factor Model with Power
Utility Shin Ichi Aihara and A. BAGCHI
We consider the optimal portfolio construction problem for
maximizing a power-utility at the final time.For managing the
portfolio, we control the amounts of the bank account and several
bonds with different maturities. The dynamics of bond price is given
through the parabolic type infinite-dimensional factor model with
boundary noises. By using the dynamic programming approach, we
obtain the optimal portfolio in the incomplete market.
| |
Monte Carlo Static Replication of Barrier Options Emanuele Amerio, Antonio Vulcano, Gianluca Fusai
The static hedge for barrier options, initially proposed by Derman.et.al (1995), is theoretically very appealing, since no adjustment is required once the replicating portfolio is in place. However, their procedure is practically flawed in assuming a perfect knowledge of the future implied volatility surface. As a possible solution, in this paper we propose a statistical static hedge, also called Monte Carlo static replication, consisting in generating risk-neutral dynamics for the volatility surface and in determining the portfolio minimizing the tracking error, i.e. the difference between the portfolio value and the barrier option price. This minimization is accomplished by a least square approach along the barrier level, where the dependent variable is the barrier option price, while the independent variable is the replicating portfolio value. Through our statistical approach, we can use the R² coefficient in order to have a preliminary measure of the goodness of the Monte Carlo static hedge; check the convergence of the statistical replicating portfolio to the barrier option price obtained by direct simulation of the underlying; construct confidence intervals around the estimated portfolio weights and hence around the barrier option price; evaluate the performance of our methodology when employing realistic dynamics of implied volatility surfaces.
| |
Mean-variance hedging
for discontinuous asset price processes
Takuji Arai
Our goal in this paper is to give a representation of
mean-variance hedging strategy for models whose asset price process
is discontinuous as an extension of Gourieroux, Laurent and
Pham(1998) and Rheinlander and Schweizer(1997). However, we have to
impose some additional assumptions related to the variance-optimal
martingale measure.
| |
No Arbitrage
Conditions and Liquidity Fabian Astic and Nizar Touzi
We extend the fundamental theorem of asset pricing to the case
of markets with liquidity risk. Our results generalize those
obtained by Kabanov, Rásonyi and Stricker (2002, 2003) and by
Schachermayer (2004) for markets with proportional transaction
costs. More precisely, we generalize the notion of robust
no-arbitrage and prove that it is equivalent to the existence of a
so-called strictly consistent price system. Moreover, we give
another generalization of the usual no-arbitrage condition that has
the existence of a consistent price system as a dual
characterization, which is a question that remained open even in the
case of transaction costs. This concept, called weak no-arbitrage
condition, roughly states that one does not systematically create an
arbitrage by enlarging the "positive orthant".
| |
Accurate Yield Curve
Scenarios Generation using Functional Gradient Descent
Francesco Audrino, Fabio Trojani
We propose a multivariate nonparametric technique for generating
reliable historical yield curve scenarios and confidence intervals.
The approach is based on a Functional Gradient Descent(FGD)
estimation of the conditional mean vector and volatility matrix of a
multivariate interest rate series. It is computationally feasible in
large dimensions and it can account for non-linearities in the
dependence of interest rates at all available maturities. Based on
FGD we apply filtered historical simulation to compute reliable
out-of-sample yield curve scenarios and confidence intervals. We
back-test our methodology on daily USD bond data for forecasting
horizons from 1 to 10 days. Based on several statistical performance
measures we find significant evidence of a higher predictive power
of our method when compared to scenarios generating techniques based
on (i) factor analysis,(ii) a multivariate CCC-GARCH model, or (iii)
an exponential smoothing volatility estimators as in the RiskMetrics
approach.
| |
Optimal impulse
control for a multidimensional cash management system with nonlinear cost
functions Stefano Baccarin
We consider the optimal impulse control of a multidimensional cash man-
agement system which fluctuates randomly in accordance with a homogeneous
diffusion process in
: Holding/penalty costs are continuously incurred at a
rate which is a positive function of the state of the system. The controller can
at any time increase (or decrease) the cash funds by transferring money among
them and by selling (or buying) short-term securities. However the presence of
a fixed component in the transaction costs structure makes a continuous control
unprofitable. Under general assumptions we show that the value function of the
problem is the minimum solution of a quasi-variational inequality in a suitable
Sobolev space. From this solution it is always possible to deduce the existence of
an optimal impulse policy. Furthermore we show that the value function of the
corresponding problem on a bounded domain, enlarging the domain converges
to the value function in
: This result allows us to prove the convergence of
a numerical scheme in by using a numerical procedure already existing for
problems on bounded domains.
| |
Universal Exercise
Signals for American Options: A New Approach to Optimal
Stopping Peter Bank
We present a new approach to optimal stopping problems where the
usually considered Snell envelope is replaced by the solution to a
stochastic representation problem. The main advantage of this
approach is that it provides what we call a universal stopping
signal, i.e., a single stochastic process which simultaneously
yields optimal stopping rules for a whole family of stopping
problems. This result is illustrated by considering American put
options on a common underlying. We also discuss numerical aspects of
our approach and present an ecient algorithm to solve the central
stochastic representation problem.
| |
Maximum Likelihood
Estimation of Latent Affine Processes
David Bates
This article develops a direct filtration-based maximum
likelihood methodology for estimating the parameters and
realizations of latent affine processes. Filtration is conducted in
the transform space of characteristic functions, with a version of
Bayes' rule used for recursively updating the joint characteristic
function of latent variables and the data conditional upon past
data. An application to daily stock returns over 1953-96 reveals
substantial divergences from EMM-based estimates; in particular,
more substantial and time-varying jump risk. The implications for
option prices are discussed.
| |
Optimal stopping and
American options with discrete dividends and exogenous risk
Anna Battauz and Maurizio Pratelli
We analyze the effects of discrete dividends on option pricing.
First of all, we allow for the presence of a discrepancy between the
stock jump and the dividend amount, in order to be consistent with
the empirical evidence of excess returns at the dividend payment
date. Following Heath and Jarrow (Journal of Business, 1988) we
introduce a stochastic discrepancy factor to avoid arbitrage
opportunities in the market model, and characterize the set of
equivalent martingale measures as well as the minimal
martingale/variance optimal measure. Then, we fit the model to
Italian market data and analyze the impact of the discrepancy on the
prices of American call options written on such discrete-dividend
paying stocks. In a continuous-time model (like Black-Scholes) the
optimal stopping time of an American call option does not exist,
since formally it is not possible to exercise the option at the end
of the cum-dividend date. To this aim, we study an optimal stopping
problem with restrictions on exercise dates, in order to circumvent
the failure of the continuous-time model.
| |
Monte carlo method
using malliavin calculus on poisson space for the computation of
greeks. Marie-pierre Bavouzet-morel, Vlad Bally, Marouen Messaoud
We use the Malliavin calculus for Poisson processes in order to
compute sensibilities for European options with underlying following
a jump type diffusion. The key point of the calculus is to state an
integration by parts formula for general random variables: we have
to define the differential operators which are involved in the
weight following from this formula. Then, we compute them to perform
Monte Carlo simulations of this weight.
| |
Quickest Detection of
the Poisson Disorder with Exponential Delay Cost
Erhan Bayraktar and Savas Dayanik
We solve the Poisson disorder problem when the delay is
penalized exponentially. The relevance of the quickest detection
problems to the financial data analysis are discussed by Shiryaev
[Mathematical finance---Bachelier Congress, 2000 (Paris), Springer
Finance, Springer, Berlin, pp.~487--521]. In the Poisson disorder
problem, the intensity of a Poisson process changes at some
unobservable random time, and the objecive is to detect the change
point as accurately as possible. The change point delimits two
different regimes in which one employs distinct strategies (e.g.,
investment, advertising, manufacturing). We seek a stopping rule
that minimizes the frequency of false alarms and an exponential
(unlike previous formulations which use a linear) cost function of
the detection delay. Especially in the financial applications, the
exponential penalty is a more apt measure for the cost of delay
because of the compounding of the investment growth. The Poisson
disorder problem with a linear delay cost is studied by Peskir and
Shiryaev [Advances in Finance and Stochastics, Springer, Berlin,
295-312], which is a special limiting case of ours.
| |
Arbitrage in a
Discrete Version of the Wick Fractional Black-Schole Market
Christian Bender and Robert J. Elliott
We consider binary market models based on the discrete Wick product
instead of the pathwise product and provide a sufficient criterion for
the existence of an arbitrage. This arbitrage is explicitly constructed
in the class of self-financing one-step buy-and-hold strategies, (i.e.,
the investor holds shares of the stock only at one time step). Using
coefficients obtained from an approximation of a fractional Brownian
motion with Hurst parameter bigger than a half, the result is applied
to a discrete version of the Wick-fractional Black-Scholes market.
| |
Modeling of spot and
forward/futures contracts in markets for electricity and
weather Fred Benth, Steen Koekebakker, Jurate Saltyte-Benth
The problem of pricing forward/futures products and options
written on these in electricity and weather markets is analyzed.
Empirical studies motivate the introduction of Levy processes and
seasonal volatility for temperature and spot electricity. Based on
such models, we derive the forward/futures prices which are based on
averages of the underlying product to be delivered. A dynamics for
the option premium can be derived, at least numerically. Taking the
alternative Heath-Jarrow-Morton approach, we discuss no-arbitrage
models when the forward/futures can be overlapping in their delivery
period, and present models which we study empirically on data
collected from the Norwegian power exchange NordPool.
| |
Measuring Default Risk
Premia from Default Swap Rates and EDFs
Antje Berndt, Rohan Douglas, Darrell Duffie, Mark Ferguson
This work estimates recent default risk premia for U.S.
corporate debt, based on a close relationship between default
probabilities, as estimated by Moody's KMV EDFs, and default swap
(CDS) market rates. The default-swap data, obtained through CIBC
from 22 banks and specialty dealers, allow us to establish a strong
link between actual and risk-neutral default probabilities for the
69 firms in the three sectors that we analyze: broadcasting and
entertainment, healthcare, and oil and gas. We find dramatic
variation over time in risk premia, from peaks in the third quarter
of 2002, dropping by roughly 50\% to late 2003. This is joint work
with R. Douglas, D. Duffie, M. Ferguson, and D. Schranz.
| |
Incomplete
Information, Heterogeneous Beliefs and Bounded Rationality
Tony Berrada
We consider a continuous time pure exchange economy with
incomplete information populated by agents with heterogeneous
beliefs and learning rules. There is a bayesian learner who
constructs an optimal filter based on the available observations and
a number of other agents displaying different learning bias, namely
under / over reaction and over confidence. By simulation we study
the distribution of terminal and average consumption share in 2
agent (bias + unbiased) and 3 agent (bias 1, bias 2, unbiased)
economies. We find that for relatively long horizons (50 years), an
unbiased agent does not necessarily dominate when facing 2 other
biased agents. In this model, agents, biased or not, never assign 0
probability to observable events. We show that, as a result, there
are no agent with consumption equal to 0 at anytime. We also
consider the effect of irrational agent on equilibrium prices and
trading volume, and we find that depending on other agents learning
rules the average stock ownership of the rational agent varies
considerably. Finally we consider the impact of irrational agent on
the equilibrium stock volatility.
| |
Replication and
Mean-Variance Approaches to Pricing and Hedging of Credit Risk
Tomasz Bielecki, Monique Jeanblanc, Marek Rutkowski
The paper presents some methods and results related to the
valuation and hedging of defaultable claims (credit-risk sensitive
derivative instruments). Both the exact replication of attainable
defaultable claims and the mean-variance hedging of non-attainable
defaultable claims are examined. For the sake of simplicity, the
general methods are then applied to simple cases of defaultable
equity derivatives, rather than to the more complicated examples of
real-life credit derivatives.
| |
Fractional
Heath-Jarrow-Morton Models Jaya Bishwal
We introduce a class of one factor Heath-Jarrow-Morton term
structure models driven by fractional Brownian motion with Hurst
parameter H > 1/2. This class of models is important since it
captures the long memory behavior of the forward rate and it
captures, as a special case, all term structure models where the
short term represents a time-homogeneous univariate fractional
diffusion in the equivalent fractional risk neutral economy. We
introduce several fractional short rate models. We obtain the
fractional bond option pricing formula. Where as in the classical
case of short memory the option price depends only on the length of
time to exercise, here it depends both on the the exercise time and
the maturity time. We then study the asymptotic behavior of
estimator of the term structure's volatility. Finally we study
estimation in fractional stochastic volatility model as a filtering
problem.
| |
Towards a General
Theory of Good Deal Bounds Tomas Björk and Irina Slinko
We consider a Markovian factor model consisting of a vector
price process for traded assets as well as a multidimensional random
process for non traded factors. All processes are allowed to be
driven by a general marked point process (representing discrete jump
events) as well as by a standard multidimensional standard Wiener
process. Within this framework we provide the following results. 1.
We extend the Hansen-Jagannathan bounds for the Sharpe Ratio to the
point process setting. 2. We study arbitrage free good deal pricing
bounds for derivative assets along the lines of Cochrane and
Saa-Requejo (2000). Using martingale techniques we derive the
relevant Hamilton-Jacobi-Bellman equation for the upper and lower
good deal bound functions, thus extending the results from Cochrane
and Saa-Requejo to the point process case. 3. In particular we study
the case of a single price process driven by a scalar Wiener process
as well as by a marked point process. For this case we provide a
detailed analysis of the dynamic programming equation and the
optimal market prices of risk. As a concrete application we present
numerical results for the classic Merton jump-diffusion model.
| |
Modelling forward
curves for seasonal commodities
Svetlana Borovkova and Heylette Geman
We introduce a new way to model futures prices for seasonal
commodities such as natural gas, electricity and agricultural
commodities. We extend the well-known cost-of-carry relationship
between spot and futures prices, by introducing a deterministic
seasonal premium and a stochastic convenience yield. This leads to a
better understanding of the futures prices; separating the
deterministic seasonal component of the forward curve allows us to
study features of futures prices, that are normally obscured by
dominant seasonal effects. Our model is a two-factor model with the
factors given by the average level of the forward curve and the
stochastic convenience yield. We describe a method for estimating
the seasonal cost-of-carry model and apply it to natural gas and
electricity futures. We outline some properties of the stochastic
convenience yield and illustrate them on the examples of energy
futures. Applications of the model to derivatives pricing will be
discussed as well.
| |
Stochastic Volatility
Models: a Large Deviation Approach
Phelim Boyle, Shui Feng, Weidong Tian
When volatility is stochastic the market is incomplete and there
is an infinite number of equivalent martingale measures. This paper
analyzes the situation where we have a sequence of stochastic
volatility models which converge in the limit to a complete market
model with deterministic volatility. We examine the convergence of
derivative prices as the stochastic volatility model converges to
its deterministic limit. Using some recent results from large
deviation theory we are able to demonstrate convergence for a wide
class of diffusion processes. We also examine the speed of
convergence and establish a theoretical result which will be used to
analyze the hedging risk in stochastic volatility models in a
subsequent version of this Keywords: Stochastic volatility,
incomplete market, hedging, large deviations.
| |
Tractable Hedging - An
Implementation of Robust Hedging Strategies
Nicole Branger and Antje Mahayni
This paper provides a theoretical and numerical analysis of
robust hedging strategies in a diffusion-type setup including
stochastic volatility models. A robust hedging strategy avoids any
losses as long as the realised volatility stays within a given
interval. We focus on the effects of restricting the set of
admissible strategies to tractable strategies which are defined as
the sum over Gaussian strategies. While the cheapest robust hedge of
a mixed payoff-profile is given by a numerical solution of a
stochastic control problem, firstly analyzed in Avellaneda, Levy,
and Parás (1995), a tractable hedge still allows for a closed form
solution. The cheapest tractable hedge can be represented by one
long and one short position in convex claims where each claim is
hedged separately. We show that although a trivial Gaussian hedge
may be prohibitively expensive compared to the cheapest overall
hedge, this is not the case for the cheapest tractable hedge.
Furthermore, we use a Monte Carlo simulation to illustrate the
hedging performance and the distribution of terminal losses in a
stochastic volatility model. The results show that after taking the
different initial capital into account, the optimal tractable
strategies behave quite similar to the cheapest robust hedge.
| |
A Complete Market
Model for Implied Volatility
Oliver Brockhaus
This paper presents a new framework for modelling equities in
discrete time. Within this framework di erent smile dynamics such as
"sticky delta" and "sticky strike" can be represented. The approach
builds upon ideas by Schönbucher [3], Hobson and Rogers [2] and an
earlier paper by the author [1] in that • implied volatility is
modelled directly • the model is assumed to be complete but not
Markovian • the asset distribution is only generated for a small
number of deal relevant dates using implied sampling. The general
framework provides a parameterisation of all arbitrage free
distributions of an asset process observed at discrete times.
Specific parameterisation examples allowing to match both spot and
forward implied volatilities demonstrate the practical relevance of
the approach.
| |
Entropic Calibration
Revisited Ian Buckley, Dorje C Brody, Bernhard K Meister
Whilst the entropic calibration of the risk-neutral density
function is effective in recovering the strike dependence of
options, it encounters difficulties in determining the relevant
greeks. By use of put-call reversal we apply the entropic method to
the dual, time-reversed economy, which allows us to obtain the spot
price dependence of options and the relevant greeks. The failure of
the calibration to satisfy global constraints can be used as a
litmus test for the price consistencies in the market data, and thus
can be used as a powerful tool for risk management. Numerical
examples will include calibration to real market call prices at
single and multiple strike prices. Wherever possible, ideas will be
presented graphically and using animations.
| |
Optimal Dividend
Policy with Mean-Reverting Cash Reservoir Abel
Cadenillas, Sudipto Sarkar, Fernando Zapatero
Motivated by empirical evidence and economic arguments, we
assume that the cash reservoir of a financial corporation follows a
mean reverting process. The firm must decide the optimal dividend
strategy, which consists of the optimal times and the optimal
amounts to pay as dividends. We model this as an stochastic impulse
control problem, and succeed in finding an analytical solution. We
also find a formula for the expected time between dividend payments.
A crucial and surprising result of our paper is that, as the
dividend tax rate decreases, it is optimal for the shareholders to
receive smaller dividend payments. [Joint work with S. Sarkar and F.
Zapatero].
| |
A Family Of
Term-structure Models with Stochastic Volatility
Andrew Cairns and Samuel A. Garcia Rosas
In this paper we extend the class of multifactor term-structure
models proposed by Cairns (2004) to incorporate a more explicit form
of stochastic volatility. The models are built up within the
framework proposed by Flesaker \& Hughston (1996). Our general
aim is to work with models in which zero-coupon bond prices can be
expressed in the form \[ P(t,T)=\frac{\int_{T-t}^\infty e^{A(u) +
B(u)^T X(t)} du}{ \int_{0}^\infty e^{A(u) + B(u)^T X(t)} du} \] for
some $n$-dimensional, stationary diffusion $X(t)$ and for suitable
deterministic functions $A(u)$ and $B(u)$. We prove that the models
require a multivariate affine state-variable $X(t)$ as developed
previously by Duffie \& Kan (1996). The remainder of the paper
describes some numerical experiments for specific two and
three-factor models which incorporate one stochastic volatility
component. The models have a close relationship with recently
developed market models incorporating stochastic volatility. The new
models can therefore be used to provide practitioners with a
parsimonious benchmark against which more elaborate market models
can be compared.
| |
Some results on
quadratic hedging with insider trading
Luciano Campi
We consider the hedging problem in an arbitrage-free incomplete
financial market, where there are two kinds of investors with
different levels of information about the future price evolution of a stock,
described by two filtrations $\mathbf F$ and $\mathbf G =\mathbf F
\vee \sigma (G)$ where $G$ is a given r.v. representing the
additional information. We focus on two types of quadratic
approaches to hedge a given square-integrable contingent claim:
local risk minimization (LRM) and mean-variance hedging (MVH). By
using initial enlargement of filtrations techniques, we solve the
hedging problem for both investors and compare their optimal
strategies under both approaches.
| |
Generalised
Fractional-Black-Scholes Equation: pricing and hedging
Alvaro Cartea
In this paper we show that when the underlying risk-neutral
dynamics of securities follow either a regular Lévy process of the
exponential type (RLPE) or a maximally skewed Lévy-Stable process
(also known as the FMLS) it is possible to express the corresponding
Black-Scholes operator in terms of fractional derivatives. Based on
this Fractional-Black-Scholes operator we show how portfolios can be
hedged. Finally, we show an interesting connection between factional
derivatives, also known as the Riemann-Liouville operators, and the
Lévy measure of the corresponding processes we use.
| |
Reservation Prices on
Order Driven Markets Boyer Cécile
On order driven financial markets, a limit order investor faces
two types of risk: one concerns the execution of his order and the
other concerns the profit when the order is executed, through the
future price evolution. Investors trade for liquidity and/or
speculative purpose. In any cases, any executed order is more likely
to generate a negative profit ex post. This problem induces a
winner's curse risk. In this paper, we define a transfer premium in
order to describe the behavior of an investor facing such risks.
This premium corresponds to the amount of money an investor is
willing to pay for sure execution. This premium allows us to
characterize the order driven market reservation prices and, as a
result, to compare the prevailing spreads on quote and order driven
markets.
| |
The Risk of Optimal,
Continuously Rebalanced Hedging Strategies and Its Efficient Evaluation
via Fourier Transform Ales Cerny
This paper derives a closed-form formula for the hedging error
of optimal and continuously rebalanced hedging strategies in a model
with leptokurtic IID returns and, in contrast to the standard
Black-Scholes result, shows that continuous hedging is far from
riskless even in the absence of transaction costs. The paper
provides an efficient implementation of the hedging error formula
via FFT and demonstrates its speed and accuracy. We compute the size
of hedging errors for individual options based on the historical
distribution of returns on FT100 equity index as a function of
moneyness and time to maturity. The resulting option price bounds
are non-trivial, and largely insensitive to model parameters. Our
result is an extension of the Capital Asset Pricing Model and the
Arbitrage Pricing Theory, allowing for intertemporal risk
diversification, and at the same time it represents an important
generalization of the Black-Scholes pricing formula.
| |
Optimal financing
policies via a stochastic control problem with exit time
Roy Cerqueti
The life of a firm can be influenced by several events, whose
impact can change drastically the evolution of the dynamic
associated to the firm's value. We propose in these pages a
particular case of an equilibrium model, in which such event is
represented by the external financing. The complexity of the
financing modelling is due to the differences between the
financiers. In the work that we present, we want to construct a
model that can take in account two fundamental cases of external
financiers: a bank and an illegal financier. We start from a work
due to Peccati et al., but the aim of our research is different. We
want to search for the best interest rate that the financier has to
apply to a firm in order to catch up a certain objective. The
mathematical tools adopted in order to solve the problem are
associated to the stochastic control theory via dynamic programming
approach.
| |
Good-deal equilibrium
pricing bounds on option prices
Marie Chazal and Elyés Jouini
We consider the problem of finding bounds for European option
prices when the underlying asset dynamics is unknown. More
precisely, given a consensus on the actual distribution of the
underlying asset at maturity, we derive an upper bound for the call
option price which only depends on the second moment of the stock
terminal value under the risk-neutral probability measure. We
further restrict to equilibrium prices, i.e. that are obtained under
some probability measure which has a Radon-Nikodym density with
respect to the true probability measure which is a nonincreasing
function of the stock price at maturity. We formulate an associated
dual problem and obtain some sufficient condition for strong duality
and existence to hold, in a fairly general context. Explicit bounds
are provided for the call option. Finally, we provide a numerical
example.
| |
DrawDown Measure in
Portfolio Optimization Alexei Chekhlov,
Stanislav Uryasev, and Michael Zabarankin
A new one-parameter family of risk measures called Conditional
Drawdown (CDD) has been proposed. These measures of risk are
functionals of the portfolio drawdown (underwater) curve considered
in active portfolio management. For some value of the tolerance
parameter Alpha, in the case of a single sample path, drawdown
functional is defined as the mean of the worst (1 - Alpha)* 100%
drawdowns. The CDD measure generalizes the notion of the drawdown
functional to a multi-scenario case and can be considered as a
generalization of deviation measure to a dynamic case. The CDD
measure includes the Maximal Drawdown and Average Drawdown as its
limiting cases. Mathematical properties of the CDD measure have been
studied and efficient optimization techniques for CDD computation and
solving asset-allocation problems with a CDD measure have been
developed. The CDD family of risk functionals is similar to
Conditional Value-at-Risk (CVaR), which is also called Mean
Shortfall, Mean Excess Loss, or Tail Value-at-Risk. Some
recommendations on how to select the optimal risk functionals for
getting practically stable portfolios have been provided. A
real-life asset-allocation problem has been solved using the
proposed measures. For this particular example, the optimal
portfolios for cases of Maximal Drawdown, Average Drawdown, and
several intermediate cases between these two have been found.
| |
Time Series Properties
of Cross-Sectional Equity Returns Gib Bassett, Chen Chen, Rong Chen
We investigate time series properties of the cross sectional
distributions of equity returns. Changes in cross sectional
dispersion affect the ability of portfolio managers to exploit
skill. It also changes portfolio risk. The analysis is based on
monthly returns of the 1000 largest stocks on the NYSE from 1991 to
2003. Dispersion is measured by the variance, and also by
differences in the extreme quantiles. We show that the quantile
analysis picks up properties of the changing distribution that are
missed by focusing on just the variance. We find that: (1)
dispersion has varied widely over the decade; (2) there is
short-term predictability in dispersion; and (3) dispersion is
associated with market volatility. We suggest theoretical reasons
for the connection between dispersion and volatility and indicate
how it is related to the cross sectional dispersion of market ?s and
?s.
| |
On Modeling
Firm-Specific Correlations between Bonds and Stocks
Li Chen, Damir Filipovic, H. Vincent Poor
In this paper, a model is presented to establish the correlation
between the returns on individual stocks and the yield changes of
individual bonds issued by the same firm. The dynamics of a firm's
credit migration, earnings process and default event are jointly
characterized in a general affine setting together with
consideration of market risk. This new modeling strategy provides a
unifying framework of pricing corporate bonds and stocks subject to
default risk. Closed-form price formulas for both securities are
derived. Furthermore, the model is implemented using integrated data
from both bond and stock markets. Its empirical fitting ability and
the pricing performance are investigated. It turns out that by
incorporating the information from the bond market, the model
exhibits a strong predictive power on stock prices.
| |
General arbitrage
pricing model: probability and possibility approaches
Alexander Cherny
The paper has 5 main goals: 1) We propose the general arbitrage
pricing model. (It is similar to the general model introduced by
Harrison and Kreps, but there is a number of important differences.)
This model includes various models used in the theory of pricing by
arbitrage as particular cases. Within the framework of the general
model, we obtain the Fundamental Theorem of Asset Pricing; the form
of fair prices of a contingent claim; the form of fair prices of a
controlled contingent claim (this notion is introduced in the
paper). 2) The obtained general results are applied to several
particular models (one-period model, multiperiod model,
continuous-time model, etc.) The "projection" of general results on
these models leads us, in particular, to the revision of the
Fundamental Theorem of Asset Pricing in the continuous-time setting.
3) The general approach mentioned above allows us to narrow the
class of risk-neutral measures (and thus to make the intervals of
fair prices shrink) by taking into consideration the current prices
of traded securities (options, bonds, etc.). 4) Furthermore, the
obtained results are extended to models with friction, i.e. models
with proportional transaction costs; restrictions on short selling;
costs of short selling. 5) Finally, we introduce the possibility
approach to pricing by arbitrage. When using this approach, one does
not need to know the original probability measure. It is shown that
all the results described above can be transferred to the
possibility framework.
| |
Pricing Swap Credit
Risk with Copulas Umberto Cherubini
We apply copula functions to evaluate counterpart risk in swap
transactions. Using copulas allows to generalise the approach
proposed by Sorensen and Bollier (1994), allowing for dependence
between swap rates and counterparty default. Counterpart risk is
represented by a sequence of vulnerable swaptions, which are priced
using Cherubini and Luciano (2002) approach. Using copulas grants
maximum flexibility with the choice of term structure and default
risk models, as well as with the specification of the dependence
structure between interest rate and credit risk. Closed form hedging
and pricing formulas are derived for extreme dependence cases and
for copula functions of the Fréchet family. An empirical application
based on actual market data has shown that dependence affects both
the level and the slope credit spreads, particularly for the case in
which a credit institution is paying fixed. The effect is reversed
in the case in which the financial institution pays floating.
| |
Free boundary near the
maturity for an American option on several assets
Etienne Chevalier
We consider an American put option on a linear function of d
dividend-paying assets. The value function of this option can be
viewed as the solution of a free boundary problem, which can be
formulated as a variational inequality. When d=1, the behavior of
the free boundary near the maturity of the option is well-known:
when the payoff function is regular near the limit of the boundary
at maturity, the behavior is parabolic. In less regular situations
however, an extra logarithmic factor appears. We will extend the
study of the free boundary near maturity to the case d>1. A
parametrisation of the stopping region at time t is given. That
enables us to define and give a convergence rate for this region
when t goes to the maturity.
| |
A General Benchmark
Model for Stochastic Jump Sizes
Morten Christensen and Eckhard Platen
This paper extends the benchmark framework of Platen (2004). It
introduces a sequence of incomplete markets, having uncertainty
driven by an m-dimensional Wiener Process and a marked point
process. By introducing an idealized market, in which all
fundamental economical variables exist, but may not all be traded, a
generalized growth optimal portfolio (GOP) is obtained and
calculated explicitly. The problem of determining the GOP is solved
in a general setting, which extends existing treatments. It also
provides a clear link to a fundamental variable of the economy, the
market price of risk. The connection between traded securities,
arbitrage and market incompleteness is analyzed. This paper aims to
provide a framework for analyzing the degree of incompleteness
associated with jump processes, a problem well known from insurance
and credit risk modelling. Furthermore, by staying under the
empirical measure, the resulting benchmark model has potential
advantages for various applications in finance and insurance.
| |
Reset and Withdrawal
Rights in Dynamic Fund Protection
Chi Chiu Chu and Yue-Kuen KWOK
We analyze the nature of the dynamic fund protection which
provides an investment fund with a floor level of protection against
a reference stock index (or stock price). The dynamic protection
feature entitles the investor the right to reset the value of his
investment fund to that of the reference stock index. The reset may
occur automatically whenever the investment fund value falls below
that of the reference stock index, or only allowed at pre-determined
time instants. The protected funds may allow a finite number of
resets throughout the life of the fund, where the reset times are
chosen optimally by the investor. We examine the relation between
the finite-reset funds and automatic-reset funds. We also analyze
the premium and the associated exercise policy of the embedded
withdrawal right in protected funds, where the investor has the
right to withdraw the fund prematurely. The impact of proportional
fees on the optimal withdrawal policies is also analyzed. The holder
should optimally withdraw at a lower critical fund value when the
rate of proportional fees increases. Under the assumption that the
fund value and index value follow the Geometric Brownian processes,
we compute the grant date and mid-contract valuation of these
protected funds. Pricing properties of the protected fund value and
the cost to the sponsor are also discussed.
| |
Mispricing of S&P
500 Index Options Jens Carsten Jackwerth, George M. Constantinides, Stylianos Perrakis
We address the implied volatility smile of the S&P 500 index
options before and after the October 1987 crash. In a single-period
model, the cross sections of one-month calls and puts violate
stochastic dominance even with realistic bid-ask spreads and
transaction costs on the options and the index (SPDRs). The
violations are frequent even prior to the crash, in contrast to the
extant literature that considers the post-crash pronounced smile to
be the primary challenge to economic theory. In a multiperiod model,
we address potential stochastic dominance violations via the bounds
on bid and ask option prices developed by Constantinides and
Perrakis (2002) because the linear programming methodology applied
in the one-period model provides unusably weak restrictions. We find
that the violations of stochastic dominance persist both before and
after the crash.
| |
By Force of Habit: An
Exploration of Asset Pricing Models using Analytic Methods
Thomas Cosimano, Yu Chen, Alex Himonas
Classical analysis is used to explore solution methods for asset
pricing models. Campbell and Cochrane's (1999) habit persistence
model provides a prototypical example to illustrate these methods.
We show that the integral equation for the price-dividend function
yields a unique, bounded, continuous and infinitely differentiable
solution. Using real analysis we are able to demonstrate that the
price-dividend solution is analytic within a small interval.
Switching to complex analysis we are able to find the maximum radius
of convergence so that the price-dividend function may be portrayed
accurately as a Taylor series for any consumption growth within
[-16.25%, 16.25%] per month. I have a preference for presenting this
paper on or after Thursday afternoon, since I have to teach in South
Bend on thursday morning. Thanks. Tom Cosimano
| |
A Comparison between
the SSRD Model and a Market Model for CDS Options Pricing
Laurent Cousot and Damiano Brigo
We investigate implied volatility patterns in the Shifted Square
Root Diffusion (SSRD) model as functions of the model parameters. We
introduce a candidate market model for Credit Default Swap (CDS)
options that is consistent with a market Black-like formula. We
introduce an analytical approximation for the SSRD implied
volatility that follows the same patterns in the model parameters
and that can be used to have a first rough estimate of the implied
volatility following a calibration. We also find an increasing
CDS-rate volatility smile for the adopted SSRD model, with one
exception in one negative interest-rate / intensity correlation. We
find a decreasing pattern of SSRD implied volatilities in the
correlation. We finally find one testing set out of four where
correlation has a possibly relevant impact on CDS option prices and
on the implied CDS volatility smile shape. This work is joint with
Dr. Damiano Brigo.
| |
Optimal Contracts and
Principal-Agent Problems in Continuous Time
Jaksa Cvitanic, Abel Cadenillas, Fernando Zapatero
Principal-agent problems involve an interaction between two
parties to a contract: an agent and a principal. Through the
contract, the principal tries to induce the agent to act in line
with the principal's interests. In our framework, the agent can
control both the drift (the "mean") and the volatility (the
"variance") of the underlying stochastic process. The question is:
What is the optimal contract from the principal's point of view,
among all the contracts that the agent is willing to consider? Main
applications include optimal reward of portfolio managers and
optimal compensation of company executives. There are two main
modeling frameworks for these problems: 1) the case of full
information; 2) the case of hidden information in which the
principal cannot observe the agent's actions. In the first case we
develop a methodology based on martingale/duality methods for
stochastic optimization, that can be applied to general
semimartingale models. In the second case we assume that the
randomness is driven by Brownian Motion, and we derive necessary and
sufficient conditions (the "stochastic maximum principle") for the
problem, in the form of Forward-Backward Stochastic Differential
Equations. The latter methodology covers a number of less general
frameworks and examples considered in the existing literature.
| |
The Swing Option on
the Stock Market Martin Dahlgren and Ralf Korn
The valuation of a Swing option for stocks under the additional
constraint of a minimum time distance between two different exercise
times is considered. We give an explicit characterization of its
pricing function as the value function of a multiple optimal
stopping problem. The solution of this problem is related to a
system of variational inequalities. We prove existence of a solution
to this system and discuss the numerical implementation of a
valuation algorithm. Finally we present numerical examples, which
highlight the typical form of the optimal exercise strategy.
| |
Heterogenous Beliefs,
Trading Risk, and the Equity Premium
Alexander David
Agents who have heterogeneous beliefs about the states of
fundamental growth agree to not hedge each other perfectly, leading
to heterogeneity in consumption growth. In addition to fundamental
risks, they face trading risks, the risks of incurring losses due to
equilibrium prices responding to beliefs of other agents. These
trading risks are priced, causing larger premiums than in a similar
benchmark economy with homogeneous beliefs. Agents with a
coefficient of relative risk aversion of less than one have a
`speculative' demand for risky assets in periods of high
disagreement, but choose to remain in safe assets in periods of low
disagreement, resulting in a low riskless rate in such times.
Calibrated to fundamentals as well as survey data, the model
resolves most features of the equity premium puzzle.
| |
A self exciting
threshold term structure model Marc Decamps, Marc Goovaerts, Wim Schoutens
One-factor models assume that all the information about the term
structure of interest rates can be summarized by a single state
variable which is usually the short-term rate. Among many others,
the Vasicek, the CIR and the CEV models define the short rate
process as a linear diffusion with continuous scale and speed
densities. In this contribution, we permit the scale and speed
densities to be discontinuous at some level. We derive stochastic
representations for the transition probability as well as for the
state-price density using the skew brownian motion and the skew
three-dimensional Bessel process recently introduced by the same
authors (2003). Similarly to Linetsky (2002) and Gorovoi and
Linetsky (2003), we obtain eigenfunction expansions for the price of
general contingent claims when analytical expressions exist for the
continuous case. We interpret the resulting term structure as a
continuous-time version of the Self Exciting Threshold
AutoRegressive models (SETAR) popular in time series analysis.
Following Goldstein and Keirstead (1997), we adapt the
Heath-Jarrow-Morton procedure to forward rates with discontinuous
scale density and we discuss possible generalization to Libor market
models. Finally, we calibrate a SET model with two Vasicek regimes
to the U.S. yield curve.
| |
Reward-Risk Portfolio
Selection and Stochastic Dominance Enrico De Giorgi
The portfolio selection problem is traditionally modelled by two
different approaches. The first one is based on an axiomatic model
of risk-averse preferences, where decision makers are assumed to
possess a utility function and the portfolio choice consists in
maximizing the expected utility over the set of feasible portfolios.
The second approach, first proposed by Markowitz (1952), is very
intuitive and reduces the portfolio choice to a set of two criteria,
reward and risk, with possible tradeoff analysis. Usually the
reward-risk model is not consistent with the first approach, even
when the decision is independent from the specific form of the
risk-averse expected utility function, i.e. when one investment
dominates another one by second order stochastic dominance. In this
paper we generalize the reward-risk model for portfolio selection.
We define reward measures and risk measures by giving a set of
properties these measures should satisfy. One of these properties
will be the consistency with second order stochastic dominance, to
obtain a link with the expected utility portfolio selection. We
characterize reward and risk measures and we discuss the implication
for portfolio selection.
| |
Unconditional Return
Disturbances: a non Parametric Approach
Rita DEcclesia and Robert G. Tompkins
In this paper, we propose an alternative historical simulation
approach. Given a historical set of data, we define a set of
standardized disturbances and we generate alternative price paths by
perturbing the first two moments of the original path or by
reshuffling the disturbances. This approach is either totally
non-parametric when constant volatility is assumed; or
semi-parametric in presence of GARCH (1,1) volatility. Without a
loss in accuracy, it is shown to be much more powerful in terms of
computer efficiency than the Monte Carlo approach. It is also
extremely simple to implement and can be an effective tool for the
valuation of financial assets. We apply this approach to simulate
pay off values of options on the S&P 500 stock index for the
period 1982-2003. To verify that this technique works, the common
back-testing approach was used. The estimated values are
insignificantly different from the actual S&P 500 options payoff
values for the observed period.
| |
Option valuation in a
non-affine stochastic volatility jump diffusion model
Griselda Deelstra and Ahmed Ezzine
This paper proposes an alternative option pricing model in which
the stock prices follow a diffusion process with non-affine
stochastic volatility and random jumps. Our class contains and
generalises the usual square root stochastic volatility model of
Heston (1993) and the affine jump diffusion models.
Approximative European option pricing formulae are derived by
transforming a non-linear PDE in an approximate linear PDE which is
explicitly solved by using Fourier transformations. We check that
these approximative prices are close to the (very time-consuming)
Monte Carlo estimates. We also state the original affine model
corresponding to the approximations.
Model parameters are estimated from joint time series of the S&P
500 index and option prices and by using the simulated method of
moments. We evaluate the impact of the different submodels on
option prices and on implied volatility. In particular, the
relative performance based on mean-squared error (MSE) is measured
for each model. We further study the sensibility of the implied
volatility curves on the model parameters.
| |
Pension funds with a
minimum guarantee under short selling and borrowing constraints
Marina Di Giacinto and Fausto Gozzi
We propose a continuous time stochastic model of optimal
allocation for a defined contribution pension fund with a minimum
guarantee. Traditionally, portfolio selection models are interested
in maximizing the total expected discounted utility from consumption
and from final wealth, whereas our target is to maximize the total
expected discounted utility from current wealth. In our model the
dynamics of wealth takes directly into account the flows of
contributions and benefits, so that in general the portfolio is not
self-financing and the level of wealth is constrained to stay above
a "solvability level". The fund manager can invest in a riskless
asset and in a risky asset but borrowing and short selling are
prohibited. Applying dynamic programming techniques, we discuss the
existence, uniqueness and regularity of the value function which
solves the related Hamilton-Jacobi-Bellman equation. Lastly we
obtain the existence and uniqueness of the allocation strategy in a
feedback form.
| |
An algorithm for early
detection of volatility change
Vladimir Dobric
Consider the standard geometric Brownian motion stock price
model with a constant volatility and drift. The goal is to detect as
soon as possible, with probability at least p, that one of the two
parameters has changed within a fixed time interval L. The problem
is solved using a stopping time algorithm. For the given p and L
explicit upper and lower envelopes are drawn starting at the current
price and into the future. If the stock price crosses either of the
envelopes within the time L, with probability at least p, the
crossing is due to parameters changes. The algorithm is based on new
estimates of the distribution of exit times of Brownian motion
through "almost" modulus of continuity boundaries. Those estimations
are obtained using the natural wavelet expansions of Brownian
motion. The algorithm can be modified to hold for any stock price
model based on an increasing function of a Gaussian-Markov process.
| |
On the Market Price of
Volatility James Doran and Ehud Ronn
In this paper we examine the extent of the bias between
Black-Scholes (1973)/Black (1976) implied volatility and realized
term volatility. To examine this bias we institute a stochastic
volatility data generating process, and demonstrate the bias through
Monte Carlo simulation of the underlying parameters. This provides
us with the numerical justification for testing the importance of a
risk premia for volatility. We implement empirical tests for the
market price of volatility risk by analyzing at-the-money options on
the S&P 500 and S&P 100. Further, we extend the study by
considering options on natural gas contracts. Our findings suggest a
negative market price of volatility risk, and this risk contributes
to the bias between Black-Scholes/Black implied volatility and
realized term volatility.
| |
Asymptotic Analysis of
Portfolio Trading with Transaction Costs
Petr Dostal
We consider an agent who invests in a stock and a money market,
but he/she does not consume. We seek for a strategy of investment
such that the asymptotic behavior of expected utility is as good as
possible. We restrict ourselves to the utility functions with
hyperbolic absolute risk aversion. Instead of considering the
singular case corresponding to logarithmic utility, we maximize the
asymptotic behavior of total wealth form view of ergodic theory. We
restrict ourselves to such strategies that do not trade when the
position of the investor is inside a given interval (a,b) and on the
other hand they buy or sell the stock in order to keep this position
within the interval [a,b]. This approach together with the above
mentioned restrictions enable us to obtain almost explicit results.
The stock market price is supposed to be a geometric Brownian motion
while the transaction costs are assumed to be proportional to the
amount of sales and purchases of shares.
| |
The lognormal
approximation in financial and other computations
Daniel Dufresne
Sums of lognormals frequently appear in a variety of situations,
including engineering and financial mathematics. In particular, the
pricing of Asian or basket options is directly related to finding
the distributions of such sums. There is no general explicit formula
for the distribution of sums of lognormal random variables. This
paper looks at the limit distributions of sums of lognormal
variables when the second parameter of the lognormals tends to zero
or to infinity; in financial terms, this is equivalent to letting
the volatility, or maturity, tends either to zero or to infinity.
The limits obtained are either normal or lognormal, depending on the
normalization chosen; the same applies to the reciprocal of the sums
of lognormals. This justifies the lognormal approximation, much used
in practice, and also gives an aymptotically exact distribution for
averages of lognormals with a relatively small volatility; it has
been noted that all the analytical pricing formulas for Asian
options perform poorly for small volatilities. Asymptotic formulas
are also found for the moments of the sums of lognormals. Results
are given for both discrete and continuous averages. More explicit
results are obtained in the case of the integral of geometric
Brownian motion.
| |
The Levy Libor
Model Ernst Eberlein and Fehmi Oezkan
Models driven by Lévy processes are attractive in finance
because of their greater flexibility compared to classical diffusion
models. First we derive conditions for arbitrage-freeness of the
Libor rate process in a Lévy Heath--Jarrow--Morton setting. Then
we introduce a Lévy Libor market model. In order to guarantee
positive rates, the Libor rate process is constructed as an ordinary
exponential. Via backward induction we get that the rates are
martingales under the corresponding forward measures. An explicit
formula to price caps and floors which uses bilateral Laplace
transforms is derived. This is joint work with Fehmi Oezkan.
| |
Optimal Stopping
Problems for Asset Management
Masahiko Egami and Savas Dayanik
We study two optimal stopping problems of an institutional asset
manager hired by ordinary investors. The investors entrust initial
funds to the asset manager and receive coupons from the asset
manager; in return, the asset manager collects dividend or a
management fee. The asset manager has the right to terminate the
contract and to walk away with the net terminal value of the
portfolio after the payment of the investors' initial funds without
any responsibilities for any amount of shortfall. The asset
manager's first problem is to find a nonanticipative stopping rule
which maximizes her expected discounted total income. To address the
possibility of the investors losing initial funds, the asset manager
could offer limited protection in the form that the contract will
terminate as soon as the market value of the portfolio goes below a
predetermined threshold. Her second problem is to find the fair
price for this protection and the best stopping rule under this
additional clause. We assume that the market value of the asset
manager's portfolio follows a geometric Brownian motion subject to
instantaneous downward jumps with jump times arriving according to
an independent Poisson process. The problems and the setting are
motivated by those faced by the managers for the portfolios of
defaultable bonds as in collateralized debt obligations (CDOs).
| |
Pricing Claims on Non
Tradable Assets Robert Elliott and John van der Hoek
A discrete two time period model is considered where there are
both tradable and non tradable assets. The indifference bid and ask
prices for the non tradable asset are determined for general
utilities. The results generalize those of Musiela and
Zariphopoulou.
| |
Properties of European
and American barrier options
Jonatan Eriksson
We investigate monotonicity in the volatility and convexity in
the underlying asset for barrier option prices when using a time and
level-dependent volatility. It is well known that for non-barrier
contracts, convexity of the contract function implies that the price
is increasing in volatility and convex in the underlying asset. We
show that under some additional conditions depending on the interest
rate this result holds for certain barrier options. We give an
example to illustrate the necessity of the additional conditions.
Results are obtained for barrier options of both European and
American type.
| |
A Series Solution for
Bermudan Options Ingmar Evers
This paper presents closed-form expressions for pricing Bermudan
options in terms of an infinite series of standard solutions of the
Black-Scholes equation. These standard solutions are combined for
successive exercise dates using backward induction. At each exercise
date, the optimal exercise price of the underlying asset is the root
of a one-dimensional nonlinear algebraic equation. Numerical
examples demonstrate the convergence of the series to the solution
obtained using alternative methods.
| |
Duality and Derivative
Pricing with Lévy processes José Fajardo and Ernesto Mordecki
The aim of this work is to use a duality approach to study the
pricing of derivatives depending on two stocks driven by a
bidimensional Lévy process. The main idea is to apply Girsanov's
Theorem for Lévy processes, in order to reduce the posed problem to
the pricing of a one Lévy driven stock in an auxiliary market,
baptized as "dual market". In this way, we extend the results
obtained by Gerber and Shiu (1996) for two dimensional Brownian
motion. Also we examine an existing relation between prices of put
and call options, of both the European and the American type. This
relation, based on a change of numeraire corresponding to a change
of the probability measure through Girsanov's Theorem, is called
put{call duality. It includes as a particular case, the relation
known as put{call symmetry. Necessary and sufficient conditions for
put-call symmetry to hold are obtained, in terms of the triplet of
predictable characteristic of the Lévy process.
| |
On the Valuation of
Options in Jump-Diffusion Models by Variational Methods
Vadim Linetsky, Liming Feng, Michael Marcozzi
We consider the valuation of European and American-style options
under jump-diffusion processes by variational methods. In
particular, the value function is seen to satisfy a parabolic
partial (spatial) integro-differential variational inequality. A
theoretical framework is developed and an analysis of a finite
element implementation presented. A key feature is the introduction
of separate approximation domains for both the state space and jump
process variables. When coupled with any semi-implicit time
integrator, this procedure presents a full discretization which is
of optimal efficiency; the additional computational cost of
evaluating the value function associated with a jump-diffusion as
compared to pure diffusion process is, as a practical matter,
negligible. Multi-dimensional computations are presented that
validate the applicability and efficiency of the method.
| |
Which input in the
calibration of a stochastic volatility model?
Gianna Figa' Talamanca
In this paper we investigate the informational content, for
pricing purposes, of historical data of the underlying stock and of
past implied data from European option prices in the context of a
specific stochastic volatility model suggested by Heston (1993).
First of all, some calibration techniques, based respectively on the
partial information contained in the stock historical data or in the
implied volatilities data, are analysed in terms of robustness and
efficiency in the estimates provided. Then the techniques are
applied for fitting Heston model to market data obtained for the
MIB30 and the SP500 stock indexes. Preliminary results show a more
powerful pricing forecast for the technique based on implied data of
option prices when suitable restrictions are imposed on input data
to be considered in the calibration. Finally, a mixed calibration
technique which takes as input both historical and implied data is
suggested, which seems capable to exploit all the information
available, "taking the best" from the partial techniques analysed.
| |
Nonparametric
estimation of Exponential Levy Models for Asset Prices
Jose Figueroa-Lopez and Christian Houdre
Accurate asset price models is of crucial importance in modern
mathematical finance. Recent works have introduced asset price
models driven by Levy processes with superior performance compared
to the classical Black-Scholes model. However, the high
computational intensity involved in the calibration of such models
has prevented them from being more widely used in practice. In this
talk, novel methods of model selection and nonparametric estimation
for Levy processes are presented. The estimation relies on
properties of Levy processes for small time spans, on the nature of
the jumps of the process, and on recent methods of estimation for
spatial Poisson processes. The procedures are illustrated in the
case of Gamma Levy processes as well as variance Gamma processes,
models of key relevance in asset price modeling.
| |
Credit Derivatives in
an Affine Framework Damir Filipovic and Li Chen
We develop a general and efficient method for valuating credit
derivatives based on multiple entities in an affine framework. This
includes interdependence of market and credit risk, joint credit
migration and counterparty default risk of multiple firms. As an
application we provide closed form expressions for the joint
distribution of default times, default correlations, and credit
default spreads in the presence of counterparty default risk.
| |
High Dimensional
Radial Barrier Options Neil Firth and J. N. Dewynne
Pricing high dimensional American options is a difficult problem
in mathematical finance. Many simulation methods have been proposed,
but Monte Carlo is numerically intensive, and therefore slow. We
derive an analytic expression for a new type of multi-asset barrier
option using Laplace transform methods. The solution is assumed to
be radially symmetric in the normalized non dimensional variables,
hence the name "Radial Barrier Options". In the single-asset case
our results reduce to published results for American binary barrier
options.
| |
Asset Substitution and
Debt Renegotiation Christian Riis Flor
In a structural modeling framework this paper analyzes how the
asset substitution problem is affected when debt renegotiation is
possible. The wellknown effect of asset substitution is that it
destroys ex ante firm value. In this paper asset substitution can be
used as a threat in a debt renegotiation, instead of simply being
enforced. Thus, if the firm's earnings decrease, the equity holders
will threaten the debt holders with either an asset substitution or
a default in order to obtain lower coupon payments and, hence,
reestablish the firm's optimal capital structure. Including asset
substitution in the renegotiation game mitigates the negative ex
ante effect but does not necessarily eliminate the problem.
Furthermore, the paper analyzes the optimal threat and ex ante costs
depending on the riskiness and expected growth rate of the
substituted assets.
| |
Default and Volatility
Time Scales Jean-Pierre Fouque, Ronnie Sircar, Knut Sølna
In the first passage structural approach, default occurs when
the underlying reaches a default barrier. In the classical Black-Cox
model the underlying follows a geometric Brownian motion with
constant volatility. Probability distributions of first passage
times are used to compute default probabilities. One of the
undesirable features of this model is that the yield spreads at
short maturities is almost zero which is in contradiction with
observed yield spreads. We propose here to look at the effect of
stochastic volatility on the yield spreads. We show that volatility
time scale is an essential concept in understanding this effect.
Perturbation methods are used to approximate defaultable bond prices
for instance. From the probabilistic point of view we propose
approximations to the probability distributions of hitting times of
"Brownian motion with stochastic diffusion".
| |
Markov Models for
Interacting Defaults and Counterparty Risk Rüdiger
Frey and Jochen Backhaus
In this talk we will be concerned with dynamic models for
portfolios of dependent defaults. We concentrate on models for
credit contagion, where the default of one company has a direct
impact on the default intensity of other firms. We introduce a
Markovian model and discuss the various types of interaction. We
present limit results for large portfolios in a homogeneous model
with mean-field interaction and analyze the impact of credit
contagion on the portfolio loss distribution. Finally we discuss the
pricing of basket credit derivatives in our model.
| |
A Financial Approach
to Machine Learning with Application to Credit Risk
Craig Friedman and Sven Sandow
We review a coherent, financially based approach for measuring
model performance and building probabilistic models that learn from
data. We give information theoretic interpretations of our model
performance measures and provide new generalizations of entropy and
Kullback-Leibler relative entropy. For investors with utility
functions in a three-parameter logarithmic family, our model
building method leads to a regularized relative entropy
minimization. We review applications of this methodology to two
credit problems: estimating the conditional probability of default,
given side information and estimating the conditional density of
recovery rates of defaulted debt, given side information.
| |
Liquidity Discovery
and Asset Pricing Michael Gallmeyer, Burton Hollifield, Duane Seppi
Long-dated securities are are risky, in part, because of
uncertainty about the preferences of potential counter-parties and
the terms-of-trade at which they will be willing to provide
liquidity in the future. We call such randomness liquidity risk. We
argue that liquidity risk is a important source of asymmetric
information in addition to private information about future cash
flows. We model the endogenous dynamics of liquidity risk, the risk
premium for bearing liquidity risk, and the role of market trading
in the liquidity discovery process through which investors learn
about their counter-parties' preferences and future demands for
securities. We show that market liquidity is a forward-looking
predictor of future risk and, as such, is priced. Our model also
provides rational explanations for price support levels and flights
to quality."
| |
Theory and Calibration
of Swap Market Models Stefano Galluccio
We introduces a general framework for Swap Market Models which
includes the co-terminal, the co-initial and the co-sliding model
specification. The standard LIBOR Market Model appears as a special
case of the co-sliding class. We aim at classifying Market Models
according to their practical use in pricing and risk managing
complex interest-rate derivatives. We show that swap market models
can be easily handled both in theory and practice; for the
co-terminal class, we introduce and numerically compare several
approximating analytical formulas for caplets, while swaptions can
be priced by a simple Black-type formula. A novel calibration
technique is introduced to allow simultaneous consistency with
caplet and swaption markets. In particular, we show that the
calibration of the co-terminal swap market model is faster, more
robust and more efficient than the same procedure applied to the
LIBOR market model. We numerically study efficient pricing
algorithms for exotic derivatives. Finally, suitable
smile-consistent extensions of the diffusion-based dynamics are
introduced and meaningful classes are selected according to trading
and risk-management practice.
| |
General Quadratic Term
Structures of Bond, Futures and Forward Prices
Raquel Gaspar
For finite dimensional factor models, the paper studies general
quadratic term structures. These term structures include as special
cases the affine term structures and the Gaussian quadratic term
structures, previously studied in the literature. We show, however,
that there are other, non-Gaussian, quadratic term structures and
derive sufficient conditions for the existence of these general
quadratic term structures for bond, futures and forward prices. As
forward prices are martingales under the T-forward measure, their
term structure equation depends on properties of bond prices' term
structure. We exploit the connection with the bond prices term
structure and show that even in quadratic short rate settings we can
have affine term structures for forward prices. Finally, we show how
the study of futures prices is naturally embedded in a study of
forward prices and show that the difference between the two prices
have to do with the correlation between bond prices and the price
process of the underlying to the forward contract and this
difference may be deterministic in some (non-trivial) stochastic
interest rate settings.
| |
A Hidden Markov Model
of Default Interaction
Giacomo Giampieri, Mark Davis, Martin Crowder
The occurrence of defaults within a bond portfolio is modeled as
a simple hidden Markov process. The hidden variable represents the
risk state, which is assumed to be common to all bonds within one
particular sector and region. After describing the model and
recalling the basic properties of hidden Markov chains, we show how
to apply the model to a simulated sequence of default events. Then,
we consider a real scenario, with default events taken from a large
database provided by Standard & Poor's. We are able to obtain
estimates for the model parameters, and also to reconstruct the most
likely sequence of the risk state. Finally, we address the issue of
global vs. industry-specific risk factors. By extending our model to
include independent hidden risk sequences, we can disentangle the
risk associated with the business cycle from that specific to the
individual sector.
| |
Beyond Single Factor
Affine Term Structure Models
Javier Gil-Bazo
This paper proposes a new approach to testing for the hypothesis
of a single priced risk factor driving the term structure of
interest rates. The method does not rely on any parametric
specification of the state variable dynamics or the market price of
risk, and simply exploits the constraint imposed by the no arbitrage
condition on instantaneous expected bond returns. In order to
achieve our goal, we develop a Kolmogorov-Smirnov test and apply it
to data on Treasury bills and bonds for both the U.S. and Spain. We
find that the single factor cannot be rejected for either dataset.
| |
The Market Price of
Credit Risk Kay Giesecke and Lisa Goldberg
We describe the relationship between actual probability of
default and defaultable security prices. Our starting point is
I-squared, a first passage model of default based on incomplete
information. This model incorporates the unpredictable nature of
default and thereby accounts for positive short spreads and the
abrupt drops in defaultable security prices that occur at default.
To connect prices with actual default probabilities, we analyze
post-default recovery and the credit risk premium. Our recovery
model is a generalization of the fractional market value convention
introduced by Duffie and Singleton for intensity-based credit
models. We derive I-squared pricing formulae for defaultable
securities subject to fractional recovery. The I-squared credit risk
premium has two components. One accounts for investors' aversion
towards diffusive price volatility. The other reflects aversion
toward the price jumps that occur at default.
| |
An Intensity-Based
Approach to Valuation of Mortgage Contracts Subject to Prepayment
Risk Yevgeny Goncharov
This paper gives a rigorous treatment of modeling of mortgage
securities subject to sub-optimal prepayment risk. A general model
is developed and a new alternative representation of a mortgage
price obtained. The proposed classification of approaches to the
option-based and mortgage-rate-based (MRB) is validated by the fact
that these two approaches require different analytical and numerical
tools. With the option-based specification of the prepayment
intensity, our model is the first continuous-time model; this gives
advantage in numerical treatment of the model. We propose to
generalize MRB approach by considering the endogenous mortgage rate.
The mortgage rate process is defined and existence of a solution is
proven for the option-based specification of the prepayment process.
| |
Wiener chaos and the
Cox-Ingersoll-Ross model Matheus Grasselli and T. R. Hurd
In this paper we recast the Cox--Ingersoll--Ross model of
interest rates into the chaotic representation recently introduced
by Hughston and Rafailidis. Beginning with the ``squared Gaussian
representation'' of the CIR model, we find a simple expression for
the fundamental random variable X. By use of techniques from the
theory of infinite dimensional Gaussian integration, we derive an
explicit formula for the n-th term of the Wiener chaos expansion of
the CIR model, for n=0,1,2,.... We then derive a new expression for
the price of a zero coupon bond which reveals a connection between
Gaussian measures and Ricatti differential equations.
| |
Impulse Response
Analysis and Immunization in Affine Term Structure Models
Martino Grasselli and Claudio Tebaldi
Affine Term Structure Models (ATSM) are traditionally used in
finance as a statistical model for bond portfolio immunization and,
more recently, to modelize the stochastic response of the term
structure to unanticipated macroeconomic shocks and monetary policy.
A natural stochastic approach to both problems requires some new
tools from stochastic analysis. Their computation in ultimate
analysis involves the Jacobian for the ATSM‡ow in the forward
measure, which appears to be the natural multifactor generalization
of the duration measure. We reduce its computation to the solution
of the same set of Riccati ODE required for pricing.
| |
Necessary Conditions
for the Existence of Utility Maximizing Strategies under Transaction
Costs Paolo Guasoni and Walter Schachermayer
For any utility function with asymptotic elasticity equal to
one, we construct a market model in countable discrete time, such
that the utility maximization problem with proportional transaction
costs admits no solution. This proves that the necessity of the
reasonable asymptotic elasticity condition, established by Kramkov
and Schachermayer [KS99] in the frictionless case, remains valid
also in the presence of transaction costs.
| |
Robust Utility
Maximization for Complete and Incomplete Market Models
Anne Gundel
We investigate the problem of maximizing the robust utility
functional infQ2Q EQu(X) for some set of subjective measures Q. We
give the dual characterization for its solution for both a complete
and an incomplete market model. To this end, we introduce the new
notion of reverse f-projections and use techniques developed for
f-divergences. This is a suitable tool to reduce the robust problem
to the classical problem of utility maximization under a certain
measure: the reverse f-projection. Furthermore, we give the dual
characterization for a closely related problem, the minimization of
expenditures given a minimum level of expected utility in a robust
setting and for an incomplete market.
| |
Explicit solution of a
stochastic irreversible investment problem and its moving
threshold Ulrich Haussmann, Maria B. Chiarolla
We consider a firm producing a single consumption good, that
makes irreversible investments to expand its production capacity.
The firm aims to maximize its expected total discounted real profit
net of investment on a finite horizon. The capacity is modeled as a
controlled Ito process where the control is the investment, an
increasing process. The resulting singular stochastic control
problem and the associated optimal stopping problem are solve
explicitly in the case of CRRA production functions. The moving free
boundary is the threshold at which the shadow value of invested
capital exceeds the capital's replacement cost. Then we use the
equation of the free boundary to evaluate the optimal investment
policy and the corresponding optimal profits.
| |
On Covariance
Estimation for High-Frequency Financial Data
Takaki Hayashi and Nakahiro Yoshida
We consider the problem of estimating the covariance/correlation
of two diffusion prices that are observed at discrete times in a
nonsynchronous manner. A de facto standard approach in the
literature, "realized" estimator, which is based on the sum of
cross-products of intraday log-price changes measured on
regularly-spaced intervals over a day, is problematic because choice
of regular interval size and data interpolation scheme may lead to
unreliable estimation. We present a new estimation procedure
recently proposed by Hayashi and Yoshida(03)(04), which is free of
such "synchronization" of data, hence, free of biases or other
problems caused by it. In particular, the estimators are shown to
have consistency as the observation frequency (or the market
liquidity) tends to infinity, which is not possessed by realized
estimators.
| |
Valuing Real Options
without a Perfect Spanning Asset
Vicky Henderson
The real options approach to corporate investment decision
making recognizes a firm can delay an investment decision and wait
for more information concerning project cashflows. The classic model
models of McDonald and Siegel (1986) and Dixit and Pindyck (1994)
value the investment decision as a perpetual American option and in
doing so, essentially assumes the real asset underlying the option
is traded, or that there is a perfect spanning asset available. Most
real projects however can only be partially hedged by traded
securities. Our model relaxes this assumption and assumes only a
partial spanning asset can be found. In this model, we obtain in
closed form the value of the option to invest and the optimal
investment trigger level, above which investment takes place. These
both depend on the correlation between project cashflows and the
spanning asset, risk aversion of the firm's shareholders, and
volatilities of project cashflows and the partial spanning asset. We
observe that the value of the option to invest and the trigger level
are both lowered when the spanning asset is less than perfect. This
implies the firm should invest earlier than the classic models
suggest. Although the partial spanning model contains the classic
model as a special case, it is much richer. In particular, there are
situations where the classic model recommends the firm always
postpones investment, whereas if a highly (but not perfectly)
correlated spanning asset were assumed, the firm should invest at a
certain trigger level.
| |
On the tradeoff
between consumption and investment in incomplete financial
markets Daniel Hernandez-Hernandez and Wendell H. Fleming
In this paper we are concerned with the tradeoff between long
term growth of the expected utility of wealth and consumption. The
goal is to find a consumption policy for which the optimal rate of
capital growth is zero, i.e. a policy for which balance between
consumption and investment is reached. The asymptotic limit of this
investment problem when the HARA parameter $\gamma\to-\infty$ is
also studied.
| |
Pricing electricity
risk by interest rate methods
Juri Hinz, Lutz von Grafensein, Michel Verschuere, Martina Wilhelm
We address a method for pricing electricity contracts based on
valuation of ability to produce power, which is considered as the
true underlying for electricity derivatives. This approach shows
that an evaluation of free production capacity provides a framework
where a change--of--numeraire transformation converts electricity
forward market into the common settings of money market modeling.
Using the toolkit of interest rate theory, we derive explicit option
pricing formulas.
| |
Arbitrage-free bounds
for basket options David Hobson, Peter Laurence, Tai-Ho Wang
In this paper we investigate the possible values of basket
options. Instead of postulating a model and pricing the basket
option using that model, we consider the set of all models which are
consistent with the observed prices of vanilla options, and, within
this class find the model for which the price of the basket option
is largest. This price is an upper bound on the prices of the basket
option which are consistent with no-arbitrage. In the absence of
additional assumptions on the market it is the least upper bound on
the price of the basket option. Associated with the bound is a
simple super-replicating strategy involving trading in calls.
| |
The integral of a
geometric Brownian motion is indeterminate by its moments
Per Hoerfelt
This talk proves that the integral of a geometric Brownian
motion is indeterminate by its moments. The proof is based on
geometric inequalities in Gauss space and the Pedersen criterion for
the moment problem. The question if the integral of a geometric
Brownian motion is indeterminate by its moments was raised by Yor.
The presented paper corrects previous false proofs showing that the
integral of a geometric Brownian motion is indeterminate by its
moments.
| |
Stochastic Cascades,
Credit Contagion, and Large Portfolio
Ulrich Horst
We analyze an interactive model of credit ratings where external
shocks, initially affecting only a small number of firms, spread by
a contagious chain reaction to the entire economy. Counterparty
relationships along with discrete adjustments of credit ratings
generate a transition mechanism that allows the financial distress
of one firm to spill over to its business partners. Such a
contagious infectious of financial distress constitutes a source of
intrinsic risk for large portfolios of credit sensitive securities
that cannot be ``diversified away.'' We provide a complete
characterization of the fluctuations of credit ratings in large
economies when adjustments follow a threshold rule. We also analyze
the effects of downgrading cascades on aggregate losses of credit
portfolios. We show that the loss distribution has a power-law tail
if the interaction between different companies is strong enough.
| |
Matched asymptotic
expansions for discretely sampled barrier options
Sam Howison and Mario Steinberg
The problem of calculating the approximate `continuity
correction' for discretely sampled barrier options was solved for
the Black-Scholes model by Broadie, Glasserman and Kou using
probabilistic techniques. (An explicit solution to the full problem
has recently been presented by Abrahams et al. using the Wiener-Hopf
method.) The problem can also be viewed in a PDE framework and the
method of matched asymptotic expansions used. This has the advantage
that it is more flexible and can, for example, be generalised to the
case of local volatility surfaces. I will describe the approach in
general terms and in relation to the specific example of discretely
sampled barrier contracts.
| |
Indifference pricing
and hedging in stochastic volatility models
Tom Hurd and Matheus Grasselli
We introduce "reciprocal affine" stochastic volatility models
whose elegant analytic properties lead to tractable formulas for the
indifference pricing and hedging of pure volatility claims. These
``unhedgable'' claims are not well understood theoretically, and are
difficult to price, but in our model we have the opportunity of
testing the usefulness and practicality of utility based methods for
incomplete markets.
| |
Optimal Static-Dynamic
Hedges for Barrier Options Aytac Ilhan and Ronnie Sircar
In a general semimartingale market model, we study optimal
hedging of barrier options using a combination of a static position
in vanilla options and dynamic trading of the underlying asset. Only
a finite number of types of vanilla options are available to the
investor, who chooses the optimal combination by maximizing her
expected terminal utility. The problem reduces to computing the
Fenchel-Legendre transform of the utility-indifference price as a
function of the number of vanilla options used to hedge, evaluated
at the market price of these options. Using the well-known duality
between exponential utility and relative entropy, we provide a new
characterization of the indifference price in terms of the minimal
entropy martingale measure, and give conditions guaranteeing
differentiability and strict convexity in the hedging quantity, and
hence a unique solution to the hedging problem. We discuss
computational approaches within the context of Markovian stochastic
volatility models.
| |
Evaluating the
Switching Options by Simulation
Junichi Imai
This paper develops the valuation model of a switching option
with the simulation methods. Since the switching option contains a
wide range of contingent claims the development of the valuation
model of the switching option is important. Especially, the idea of
a switching option is useful in evaluating the project value with
real options because many types of real options are regarded as
switching options. This paper extends two existing simulation
methods that can evaluate an American style option to value the
switching option, which are called the low-discrepancy mesh method
and the least square method. The efficiency and the accuracy of
these methods are examined theoretically and in the numerical
experiences.
| |
Futures Trading Model
with Transaction Costs Karel Janecek and Steven E. Shreve
We consider an agent who invests in a futures contract and a
money market and consumes in order to maximize the utility of
consumption over an infinite planning horizon in the presence of a
proportional transaction cost $\lambda>0$. The utility function
is of the form $U(c)=c^{1-p}/(1-p)$ for $p>0$, $p\neq 1$. The
asymptotic analysis for a standard stock price process was
rigorously done by Janecek and Shreve in \emph{'Asymptotic Analysis
for Optimal Investment and Consumption with Transaction Costs'} (to
appear in \textbf{Finance and Stochastics}). The authors use the
technique of super and subsolutions to the corresponding HJB
equation. Unfortunately, similar technique is not readily available
for more general market models. A fundamentally different approach
is based on purely probabilistic arguments. The advantage of this
approach is that, besides being intuitively appealing, it can be
generalized to more general market setup, e.g. stochastic
volatility, which is necessary for a successful practical
implementation.
| |
Measuring default
premium using the Cox process with shot noise intensity
Jiwook Jang
We employ the Cox process with shot noise intensity to model the
default time. The survival probability is derived based on the Cox
process with shot noise intensity that has doubly stochastic
property. As an interest rate process for non-defaultable bond, i.e.
a government bond, we use a generalised Cox-Ingersoll- Ross (CIR)
model (1985). As Lando (1998) has shown that we can combine the
effects of default and of discounting for interest rates, it is used
to obtain the default premium between non-defaultable bond and
defaultable bond (i.e. a corporate bond). Using an equivalent
martingale probability measure obtained via the Esscher transform,
risk-neutral default premium formula is derived. The asymptotic
distribution of the shot noise intensity is used not to have its
initial value. We also assume that the jump size of shot noise
intensity follows an exponential distribution to illustrate the
calculation of default premium. For simplicity, we ignore the
recovery rate.
| |
On valuation before
and after tax in no arbitrage models: Tax neutrality in the continuous
time model Bjarne Jensen
We establish necessary and sufficient conditions for a linear
taxation system to be neutral - within the continuous-time ``no
arbitrage'' model - in the sense that asset valuation is invariant
to the process for tax rates and choice of realization dates as well
as immune to timing options attempting to twist the time profile of
taxable income through wash sale transactions. We also demonstrate
that despite neutrality the portfolio choice can be quite different
across investors subject to different tax rates.
| |
Invariance Tests of
Forward Rate Models Malene Jensen and Bent Jesper Christensen
We introduce a statistical invariance test for the consistency
between the shape of the yield curve and the stochastic process
driving interest rates. The analysis is cast in the
Heath-Jarrow-Morton framework, and utilizes factor analysis of
forward rates estimated from coupon-bearing bond data. The
theoretical properties of the invariance tests are investigated, and
an application to U.S. government bonds is offered. Keywords: factor
analysis; interest rate models; invariant manifolds; JEL Codes: C51;
C52;
| |
Continuous-Time
Mean--Variance Portfolio Selection with Bankruptcy
Hanqing Jin, Tomasz R. Bielecki, Stanley R. Pliska, Xun Yu Zhou
A continuous-time mean--variance portfolio selection problem is
studied where all the market coefficients are random and the wealth
process under any admissible trading strategy is not allowed to be
below zero at any time. The trading strategy under consideration is
defined in terms of the dollar amounts, rather than the proportions
of wealth, allocated in individual stocks. The problem is completely
solved using a decomposition approach. Specifically, a (constrained)
variance minimizing problem is formulated and its feasibility is
characterized. Then, after having solved a system of equations for
two Lagrange multipliers, variance minimizing portfolios are derived
as the replicating portfolios of some contingent claims, and the
variance minimizing frontier is obtained. Finally, the efficient
frontier is identified as an appropriate portion of the variance
minimizing frontier after the monotonicity of the minimum variance
on the expected terminal wealth over this portion is proved, and all
the efficient portfolios are found. In the special case where the
market coefficients are deterministic, efficient portfolios are
explicitly expressed as feedback of the current wealth, and the
efficient frontier is represented by parameterized equations. Our
results indicate that the efficient policy for a mean--variance
investor is simply to purchase a European put option that is chosen,
according to his or her risk preferences, from a particular class of
options.
| |
Malliavin Monte Carlo
Greeks for jump diffusions Martin Johansson and Mark Davis
In recent years efficient methods have been developed for
calculating derivative price sensitivities using Monte Carlo
simulation. Malliavin calculus has been used to transform the
simulation problem in the case where the underlying follows a Markov
diffusion process. In this work, recent developments in the area of
Malliavin calculus for Levy processes are applied and slightly
extended. This permits derivation of stochastic weights, as in the
continuous case, for a certain class of jump diffusion processes.
| |
Bayesian Analysis of
Stochastic Betas Gergana Jostova and Alexander Philipov
This paper proposes a mean-reverting stochastic process for the market beta. In a simulation study, the proposed model generates significantly more precise beta estimates relative to competing GARCH betas, betas scaled by aggregate or firm-level variables, and
betas based on rolling regressions, even when the true betas are generated based on these competing specifications. Applying our model to US industry portfolios, we document significant improvement in out-of-sample hedging effectiveness relative to the traditional OLS beta estimate. In asset-pricing tests, our model provides substantially stronger support for the conditional CAPM relative to competing
beta models. It also helps resolve asset-pricing anomalies such as the size, book-to-market, and idiosyncratic volatility effects in the cross-section of stock returns.
| |
On asymptotic pricing
of securities in a multivariate extension of Scotts stochastic volatility
model Joerg Kampen, Joerg Kampen, Nicolae Surulescu
We consider a natural multivariate extension of Scotts
stochastic volatility model and analyse the pricing and its
asymptotics in the sense of \cite{papanic1}. The pricing is done via
the mean-variance hedging approach, where the market risk function
of volatility is determined by a fully nonlinear PDE, and where we
point out that explicit stationary solutions can be constructed in
case of non-leverage. We provide explicit formulas for the effective
volatility matrix and the correction constants by an explicit
solution of a multivariate Poisson equation. From the computational
point of view the high dimensional asymptotic option pricing to
lower order reduces to solving Gaussian (explictly given) integrals
which can be done very effectively by using sparse grids. We show
that explicit asymptotic formulas (in terms of fast converging
univariate power series) can be obtained for standard bivariate
options such as the call on the maximum of two assets.
| |
Computational Solution
of the American Put using the Moving Free Boundary Method
Michael Kelly
Computational Solution of the American Put using the Moving Free
Boundary Method: The solution of the American Put option has been
shown by Jamshidian, Carr, Buchen and Kelly to be theoretically
expressible in terms of an integro-differential Fredholm equation.
Unfortunately there is no known method for the general solution of
such equations. However it is shown in this paper that such
equations can be described symbolically and solved numerically using
recent mathematical software such as Mathematica. These solutions
are also compared with other approximate results.
| |
Optimal bank capital
with costly recapitalization Jussi Keppo and Samu Peura
We study optimal bank capital holding behavior as a dynamic
tradeoff between the opportunity cost of equity, the loss of
franchise value following a regulatory minimum capital violation,
and the cost of recapitalization. Our model indicates that a
recapitalization option may be valuable despite substantial delays
and fixed costs in capital issuance, and that a significant fraction
of the value of low capitalized banks may be attributable to the
option to recapitalize. Due to the downward bias in banks'
accounting return volatility, we operate the model with implied bank
return volatilities that correctly replicate observed bank capital
ratios. We argue that the option to recapitalize (at a lower cost)
can explain the lower capital ratios of large banks relative to
small banks.
| |
A Parallel time
stepping approach using meshfree approximations for pricing options with
non-smooth payoffs Greg Fasshauer, A. Q. M. Khaliq, D. A. Voss
We consider a meshfree radial basis function approach for the
valuation of pricing options with non-smooth payoffs. By taking
advantage of parallel architecture, a strongly stable and highly
accurate time stepping method is developed with computational
complexity comparable to the implicit Euler method implemented
concurrently on each processor. This, in collusion with radial
valuation of exotic options, such as American digital options.
| |
Valuation and Hedging
of Power-Sensitive Contingent Claims for Power with Spikes: a
Non-Markovian Approach Valery Kholodnyi
We present a new approach to modeling spikes in power prices
proposed earlier by the author. In contrast to the standard
approaches, we model power prices with spikes as a non-Markovian
stochastic process. This allows for modeling spikes directly as
self-reversing jumps. This also allows for the analytical valuation
of European contingent claims on power with spikes as well as for
the analytical valuation and dynamic hedging of European contingent
claims on forwards on power for power with spikes. Moreover, the
proposed non-Markovian process provides a natural mechanism to
explain the absence of spikes in the values of European contingent
claims on power far from their expiration time while power prices
exhibit spikes. This mechanism also explains why power forward
prices, far from the maturity time of the forward contracts on
power, do not exhibit spikes while power prices do. Finally, we
obtain a linear evolution equation for European contingent claims on
power with spikes. We also obtain a semilinear evolution equation
for universal contingent claims, including Bermudan and American
options, on power with spikes. Universal contingent claims and the
semilinear evolution equation for universal contingent claims were
introduced earlier by the author.
| |
Market Price of Risk
Specifications for Affine Models: Theory and Evidence
Robert Kimmel, Patrick Cheridito, Damir Filipovic
We extend the standard specification of the market price of risk
for affine yield models of the term structure of interest rates, and
estimate several models using the extended specification. For most
models, the extended specification fits US data better than standard
specifications, often with extremely high statistical significance.
Our specification yields models that are affine under both objective
and risk-neutral probability measures, but is never used in
financial applications, probably because of the difficulty of
applying traditional methods for proving the absence of arbitrage.
Using an alternate method, we show that the extended specification
does not permit arbitrage opportunities, provided that under both
measures the state variables cannot achieve their boundary values.
Likelihood ratio tests show our extension is statistically
significant for four of the models considered at the conventional
95% confidence level, and at far higher levels for three of the
models. The results are particularly strong for affine diffusions
with multiple square-root type variables. Although we focus on
affine yield models, our extended market price of risk specification
also applies to any model in which Feller's square-root process or a
multivariate extension is used to model asset prices.
| |
Pricing Options in
Electricity Markets Tino Kluge
In this working paper we examine a particular stochastic process
describing the spot price development in electricity markets. The
process under investigation is mean-reverting with a jump component
and two mean-reverting rates: one for the diffusion part and one for
the jump part. Setting the mean-reverting rate for the jump
component to a high value allows for a realistic modelling of
spikes, a phenomenon peculiar to electricity markets. This model is
highly incomplete, not only because it allows for jumps but also
because of the inability to store electricity efficiently, which
rules the underlying out to be used as a hedging instrument. We
describe a (subset) of risk neutral measures Q and show how Q can be
determined based on a continuous forward curve observed in the
market. Furthermore we give a semi-analytic formula for call
options.
| |
Optimal Portfolios
with Fixed Consumption or Income Streams
Ralf Korn and Martin Krekel
We consider some portfolio optimisation problems where either
the investor has a desire for an a priori specified consumption
stream or/and follows a deterministic pay in scheme while also
trying to maximize expected utility from final wealth. We derive
explicit closed form solutions for continuous and discrete monetary
streams. The mathematical method used is classical stochastic
control theory.
| |
Neutral Derivative
Pricing in Incomplete Markets
Christoph Kühn and Jan Kallsen
This talk is about some recent developments in neutral
derivative pricing. Neutral prices occur if investors are utility
maximizers and if derivative supply and demand are balanced.
Provided that the dual pricing measure exists unique price processes
can be derived for derivatives of European, American, and game type.
More delicate is the case that the dual pricing measure does not
exist. As stated in Hugonnier, Kramkov, and Schachermayer (2002)
there can be a whole interval of neutral prices. In addition it
turns out that in this case some initial prices which are neutral in
a model allowing only for buy-and-hold strategies in the derivative
cannot be extended to neutral price processes in the same model
allowing also for intermediate trades in the derivative. We prove an
existence theorem for neutral price processes and characterize them
as martingales under a special set of finitely additive set
functions (joint work with Jan Kallsen).
| |
Linkage between
lookback and reset features Yue Kuen Kwok and Hoi Ying Wong
The lookback feature in an option contract refers to the payoff
structure where the terminal payoff depends on the realized extreme
value of the underlying state variables. The reset feature is the
privilege given to the holder to reset certain terms in the contract
according to specified rules at the moment of reset. The reload
provision in an employee stock option entitles its holder to receive
one new (reload) option from the employer for each share tendered as
payment of strike upon the exercise of the stock option. The dynamic
protection feature in equity-indexed annuities entitles the investor
the right to reset the fund value to that of the reference stock
index. When the number of reset or reload rights becomes infinite,
the non-deterministic free boundary associated with the optimal
reset or reload decision becomes deterministic. The optimal stopping
problems associated with multiple reset or reload rights become
derivative models with an additional lookback variable. The
non-linear free boundary value problem of optimal reset becomes
linear, though the model formulation involves an additional path
dependent lookback variable.
| |
Optimal Portfolio
Delegation when Parties have different Coefficients of Risk
Aversion Kasper Larsen
We consider the problem of delegated portfolio management when
the involved parties are risk-averse. The agent invests the
principal's money in the financial market, and in return he receives
a compensation, depending on the value that he generates over some
period of time. We use a dual approach to explicitly solve the
agent's problem and subsequently use this solution to solve the
principal's problem numerically. The interaction between the risk
coefficients and the optimal compensation scheme is studied for
different coefficients. E.g.\ in the case of a more risk averse
agent the principal should according to common folklore optimally
choose a fee schedule such that the agent's derived risk aversion
decreased. We illustrate that this is not always the case.
| |
Generating Functions
for Stochastic Integrals Stephan Lawi and C. Albanese
Generating functions for stochastic integrals have been known in
analytically closed form for just a handful of stochastic processes:
namely, the Ornstein-Uhlenbeck, the Cox-Ingerssol-Ross (CIR) process
and the exponential of Brownian motion. In virtue of their
analytical tractability, these processes are extensively used in
modelling applications. In this paper, we construct broad extensions
of these process classes. We show how the known models fit into a
classification scheme for diffusion processes for which generating
functions for stochastic integrals and transition probability
densities can be evaluated as integrals of hypergeometric functions
against the spectral measure for certain self-adjoint operators. We
also extend this scheme to a class of finite-state Markov processes
related to hypergeometric polynomials in the discrete series of the
Askey classification tree.
| |
Sharpe Ratio as a
Performance Measure in a Multi-Period Setting
Ali Lazrak, Jaksa Cvitanic, Tan Wang
We study Sharpe ratio as a performance measure in a multi-period
setting. We show that the typical mean-variance efficiency
justification for using Sharpe ratio, valid in a static setting,
typically fails in a multi-period setting. To focus on the contrast
between static vs multi-period settings, we maintain the
mean-variance utility assumption of the static model. We show that
the Sharpe ratio maximization criterion is subject to the horizon
problem noted by Jensen (1969). That is, the trading strategy that
leads to the most desirable portfolio for, say, each quarter and for
four consecutive quarters is not same as the strategy that gives the
highest Sharpe ratio for a year. As a consequence, unless the
investor's investment horizon matches exactly the performance
measurement period of the portfolio manager, the portfolio with the
highest Sharpe ratio is not necessarily the most desirable from the
investor's point of view. We also show that quantitatively the
difference is significant even when the investor's investment
horizon is only six months. The difference becomes larger has the
investor's horizon becomes longer.
| |
Robust Replication of
Volatility Derivatives Roger Lee, Peter Carr
Suppose the positive asset price S follows a diffusion process.
Define realized variance to be the quadratic variation of log(St) on
[0; T]. (In practice, one typically uses the sample variance of the
daily or weekly logarithmic returns of S.) Without knowing the
dynamics of S, one can replicate a contract which pays realized
variance, as follows: hold a particular static position in
European-style options expiring at T, and trade S according to a
particular dynamic strategy. Does this known result extend to
general functions of realized variance? For example, it is of
practical interest to replicate contracts which pay realized
volatility, the square root of realized variance. Previous
replication efforts sacrificed robustness, by imposing specific models
on the S-dynamics. We show that, with trading in options, we can
replicate contracts which pay general functions of realized
variance. Instead of imposing a volatility model, we make only a
correlation assumption. We also show how to correct for deviations
from the correlation condition.
| |
The American put and
European options near expiry, under Levy processes
Sergey Levendorskiy
We derive explicit formulas for time decay for the European call
and put options at expiry, and use them to calculate analytical
approximations to the price of the American put and early exercise
boundary near expiry. We show that for many families of non-Gaussian
processes used in empirical studies of financial markets, the early
exercise boundary for the American put without dividends is
separated from the strike price by a non-vanishing margin. As the
riskless rate vanishes and the drift decreases accordingly so that
the stock remains a martingale, the boundary goes to zero uniformly
over the interval.
| |
Mean-Reverting and
Co-Integrated Energy Futures Curve Models for Pricing and Risk
Management Alex Levin
We present a broad class of multifactor no-arbitrage
mean-reverting and co-integrated HJM-type energy futures curve
models. They are equally suited for the long-term simulation in
Credit Risk systems and short-term simulation in Market Risk systems
in the real-world measure or pricing in the risk-neutral measure of
commodity derivatives. This approach generalizes a PCA methodology
for a class of no-arbitrage entire futures curve models represented
as a static non-linear transformation of linear combinations of the
time-homogeneous volatility functions driven by Gaussian
mean-reverting or co-integrated stochastic systems with constant
coefficients. A full characterization of such family is obtained. It
is proven that only four following types of models with the
corresponding types of static transformations and volatility
functions are consistent with the no-arbitrage dynamics: Generalized
Vasicek and Generalized Black-Karasinski model corresponding to
linear and exponential transformations and exponential-polynomial
volatility functions; a new no-arbitrage model with capped (bounded)
futures and spot prices corresponding to Normal CDF transformation
and volatility functions derived in closed analytical form from the
non-linear dynamic system; and a new "explosive" model corresponding
to some integral transformation. Presented results are generalized
for the state variables driven by jump-diffusion processes and the
real world futures curve dynamics. A new "historical" multifactor
parameter calibration procedure based on Lyapunov equations and
introduced Modified Principal Components is developed for
mean-reverting and multi-curve cross-mean-reverting models.
Practical examples of effective long-term simulation for the NYMEX
Crude Oil and Natural Gas futures curves are considered.
| |
Mean-variance hedging
when there are jumps Andrew Lim and Thaisiri Watewai
This paper concerns the problem of hedging a random liability in
a market when the uncertainty is modelled by Brownian motion as well
as a jump diffusion. Explicit representations of the optimal hedging
portfolio are obtained using the theory of backward stochastic
differential equations.
| |
The Spectral
Decomposition of the Option Value Vadim Linetsky
This paper develops a spectral expansion approach to the
valuation of contingent claims when the underlying state variable
follows a one-dimensional diffusion with the infinitesimal variance
$a^2(x)$, drift $b(x)$ and instantaneous discount (killing) rate
$r(x)$. The Spectral Theorem for self-adjoint operators in Hilbert
space yields the spectral decomposition of the contingent claim
value function. Based on the Sturm-Liouville (SL) theory, we
classify Feller's natural boundaries into two further subcategories:
non-oscillatory and oscillatory/non-oscillatory with cutoff
$\Lambda\geq 0$ (this classification is based on the oscillation of
solutions of the associated SL equation) and establish additional
assumptions (satisfied in nearly all financial applications) that
allow us to completely characterize the qualitative nature of the
spectrum from the behavior of $a$, $b$ and $r$ near the boundaries,
classify all diffusions satisfying these assumptions into the three
spectral categories, and present simplified forms of the spectral
expansion for each category. To obtain explicit expressions, we
observe that the Liouville transformation reduces the SL equation to
the one-dimensional Schr\"{o}dinger equation with a potential
function constructed from $a$, $b$ and $r$. If analytical solutions
are available for the Schr\"{o}dinger equation, inverting the
Liouville transformation yields analytical solutions for the
original SL equation, and the spectral representation for the
diffusion process can be constructed explicitly. This produces an
explicit spectral decomposition of the contingent claim value
function.
| |
Pricing Vulnerable
European Options with Stochastic Default Barriers
Chi-Fai Lo, C.H. Hui, K.C. Ku
This paper develops a valuation model of European options
incorporating a stochastic default barrier, which extends a constant
default barrier proposed in the Hull and White model. The default
barrier is considered as an option writer's liability. Closed-form
solutions of vulnerable European option values based on the model
are derived to study the impact of the stochastic default barriers
on option values. The numerical results show that negative
correlation between the firm values and stochastic default barriers
of option writers gives material reductions in option values where
the options are written by firms with leverage ratios corresponding
to BBB or BB ratings.
| |
Geometric Brownian
Motion of Skorohod Type as a Canonical Model for Assets with Correlated
Returns and Heavy Tails Andrew Lyasoff
Essentially all statistical studies of market data indicate that
returns from stock are slightly negatively autocorrelated. Usually,
this finding is interpreted as an evidence that returns from stocks
are very close to random walks. I will demonstrate with a concrete
example that a very small — in fact much smaller than the one
observed in market data — autocorrelation may reveal a strong
self-regulatory pattern which is inconsistent with the random walk
hypothesis. I will then consider a model for stock prices that
differs from the standard geometric Brownian motion only in that its
starting point depends on the driving Brownian motion. Somewhat
surprisingly, this model leads to probability distributions with
much heavier tails than the tails of the standard geometric Brownian
motion.
| |
Risk Control of
Dynamic Investment Models William Ziemba, Leonard MacLean, Y. Zhao
The risk inherent in the accumulation of investment capital
depends on the true return distributions of risky assets, the
accuracy of estimated returns, and the investment strategy. This
paper considers risk control with Value-at-Risk and Conditional
Value-at-Risk constraints, using control limits to determine times
for rebalancing the portfolio. Optimal strategies and control limits
are determined for a geometric Brownian motion asset pricing model,
with random parameters. The approaches to risk control are applied
to the fundamental problem of investment in stocks, bonds, and cash
over time. The advantages of portfolio rebalancing at random times
determined from control limits are illustrated.
| |
A Multinomial
Approximation of American Option Prices in a Levy Process Model
Ross Maller, David Solomon, Alex Szimayer
Models in which the the logarithm of the stock price process
follows a Levy process (a continuous time process with independent
stationary increments which generalises Brownian motion) are
presently under vigorous investigation. Of particular interest and
difficulty is the pricing of American options written on stocks
which follow these models. A couple of existing schemes are able to
deal with some but not all aspects of this. This paper proposes a
multinomial tree setup which can be viewed as generalising the
binomial model of Cox et al.(1979) for Brownian motion. Under very
mild conditions, we show that the American option prices obtained
under the multinomial model converge to the corresponding prices
under the continuous time Levy process model. Our procedure is very
general and can handle infinite activity processes such as the
variance gamma (VG) model straightforwardly. It also overcomes some
practical difficulties that have previously been encountered.
Explicit illustrations are given for jump diffusion models and for
the VG model.
| |
Detecting the presence
of a diffusion in asset prices
Cecilia Mancini
We consider a process evolving by $$dY_t = a_t dt+sigma_t
dW_{t}+dJ_t, t \leq T,$$ where $a$, $\sigma$ are progressively
measurable stochastic processes, $W$ is a standard Brownian motion,
and $J_t$ is a pure jump Lévy kind process. Such a process can be
used to model the evolution of the logarithm of the price of a
stock, of an index or of a commodity. Given a discrete record of
equally spaced observations
$\{Y_{0},Y_{t_1},...,Y_{t_{n-1}},Y_{t_n}\}$, with $t_i=ih$, $hn=T$,
we are going to manipulate them to deteremine whether the diffusion
component is zero or not. This is to make a comparison between the
widely used jump-diffusion processes and the pure jump infinite
activity processes used for instance y Madan (1999), Geman&al,
for financial stock prices or indexes. At this aim we construct
quadratic variation based estimators of the integrated volatility
and of the jump part of $Y$, which separate the two parts of $Y$. We
check the performance of our estimators on a variety of simulated
models, both in the finite jump activity case and in the infinite
activity (both finite and infinite variation) for any type of
dependence of sigma on W (cfr Barndorff-Nielsen and Shephard,
Ait-Sahalia, Woerner). We then apply our estimators to financial
high frequency data.\\
| |
Harmonic analysis methods for volatility computation
Emilio Barucci, Paul Malliavin, Maria Elvira Mancino
We provide two nonparametric methods based on harmonic analysis to
compute instantaneous and realized volatility of a diffusion process.
The methods are well suited to use high frequency data. We reconstruct
the volatility of a diffusion process as a function of time by
establishing a connection between the Fourier transform of the price
process and that of the volatility function. Then the volatility is
obtained as a series of trigonometric polynomials. A relationship is
also established between the Laplace transform of the price process and
of the volatility function.
| |
Continuous time option
valuation with discrete hedging subject to transaction costs and trading
delays Michael Marcozzi
We consider a model for the writing price of a European option
based on the utility maximization formulation of Hodges and
Neuberger (1989) and Davis et al. (1993). Transaction costs, block
trading, and transaction delays within the model are implemented
through the velocity of the wealth process, representing holding a
portfolio of securities. In particular, the value function is seen
to satisfy a temporally weak ultraparabolic Hamilton-Jacobi-Bellman
equation.
| |
Hedging under the
Minimal Potential Measures Massimo Masetti
In this paper we show a new characterization of the Minimal Mar-
tingale Measure bP, used for the local risk minimization hedging ap-
proach. First we derive the Föllmer-Schweizer decomposition under
an absolutely continuous change of measure from bP to another equiv-
alent martingale measure Q 2 Me(S). Second we derive the optional
decomposition of a Q-submartingale under an absolutely continuous
change of measure from the optimal measure Q to another Q 2 Me(S).
We show that bP is the unique Minimal Potential Measure such that,
the potential generated by the Riesz optional decomposition of the
value process bV under Q, is reduced to zero bP-a.s. and P-a.s.
Finally, we extend the interpretation of E bP [H] in terms of the
arbitrage-free prices range.
| |
A simple calibration
procedure of stochastic volatility models with jumps by short term
asymptotics Alexey Medvedev and Olivier Scaillet
In this paper we propose a simple calibration procedure to infer
key model relationships from short term option prices. We develop
and make use of an approximating formula for European options prices
based on short maturity asymptotics, i.e. when time-to-maturity
tends to zero. The analysis is performed in a general setting where
stochastic volatility and jumps drive stock price dynamics. In a
numerical study we show that the calibration procedure is accurate
for a reasonable set of model parameters. An empirical application
illustrates its application to S&P 500 index options.
| |
Efficient Hedging and
Equity-Linked Life Insurance
Alexander Melnikov
The talk is devoted to how hedging methodologies developed in
the modern financial mathematics can be exploited to price
equity-linked life insurance contracts. We study pure endowment life
insurance contracts with fixed and flexible guarantees. In our
setting, these insurance instruments are based on two risky assets
of the market controlled by Black-Scholes model during a contract
period. The first asset is responsible for the maximal size of a
future profit while the second, more reliable, asset provides a
flexible guarantee for the insured. The insurance company is
considered as a hedger of a maximum of these assets conditioned by
remaining life time of a client in the framework of this market. The
main attention is paid to new types of hedging (quantile hedging and
efficient hedging with power loss function), which, together with
Black-Scholes (fixed guarantee) and Margrabe (flexible guarantee)
formulae, creates effective actuarial analysis of such contracts. We
show also how this approach is extended to a jump-diffusion scheme
and discuss some connections with the pricing of credit risks and
defaultable derivative securities. Finally, we give numerical
examples based on financial indices the Dow Johns Industrial Average
and the Russell 2000 to demonstrate how our results can be applied
to actuarial practice.
| |
A Note on Esscher
Transformed Martingale Measures for Geometric Levy Processes
Yoshio Miyahara
The Esscher transform is one of the very useful methods to
obtain a reasonable equivalent martingale measure, and it is defined
with relation to the corresponding risk process. In this article we
consider two kinds of risk processes (compound return process and
simple return process). Then we obtain two kinds of Esscher
transformed martingale measures. The first one is the one which was
introduced by Gerber and Shiu, and the second one is identified with
the MEMM (minimal entropy martingale measure). We set up the
economical characterization of these two kinds of Esscher
transforms, and then we study the properties of the two kinds of
Esscher transformed martingale measures, comparing each others. Key
words: incomplete market, geometric Levy process, equivalent
martingale measure, Esscher transform, minimal entropy martingale
measure. JEL Classification: G12, G13
| |
On the Starting and
Stopping Problem: Application in reversible investment
Jeanblanc Monique and Hamademe Said
In this work we solve completely the starting and stopping
problem when the dynamics of the system are a general adapted
stochastic process. We use backward stochastic differential
equations and Snell envelopes. A power station produces electricity
whose selling price fluctuates. We suppose that electricity is
produced only when its profitability is satisfactory. Otherwise the
power station is closed up to time when the profitability is coming
back. So for the power station there are two modes: operating and
closed. At the initial time, we assume it is in its operating mode.
On the other hand, like every economic unit, there are expenditures
when the station is in its operating mode as well as in the closed
one. In addition, switching from a mode to another is not free and
generates costs. The problem we are interested in is to find the
sequence of times where one should make decisions to stop the
production and to start it again successively in order to maximize
the profitability of the station and then to determine the maximum
profit.
| |
Optimal Asset
Allocation and Ruin-Minimization Annuitization Strategies
Kristen Moore, Virginia R. Young, Moshe A. Milevsky
In this paper, we derive the optimal investment and
annuitization strategy for a retiree whose objective is to minimize
the probability of lifetime ruin, namely the probability that a
fixed consumption strategy will lead to zero wealth while the
individual is still alive. Recent papers in the insurance economics
literature have examined utility-maximizing annuitization
strategies. Others in the probability, finance, and risk management
literature have derived shortfall-minimizing investment and hedging
strategies given a limited amount of initial capital. This paper
brings the two strands of research together. Our model pre-supposes
a retiree who does not currently have sufficient wealth to purchase
a life annuity that will yield her exogenously desired fixed
consumption level. Therefore, she decides to self-annuitize, while
dynamically managing her investment portfolio and possibly
purchasing some annuities to minimize the probability of lifetime
ruin. However, it turns out that she will not annuitize any of her
wealth until she can fully cover her desired consumption by
annuities. We derive a variational inequality that governs the ruin
probability and optimal strategies and demonstrate that the problem
can be recast as a related optimal stopping problem that yields a
free-boundary problem that is more mathematically tractable. We
numerically approximate the ruin probability and optimal strategies
and examine how they change as we vary the mortality assumption and
parameters of the financial model. Moreover, we solve the problem
implicitly for exponential future lifetime. As a byproduct, we are
able to quantify the reduction in lifetime ruin probability that
comes from being able to dynamically manage the investment
portfolio.
| |
Pricing American
Options: a Variance Reduction Technique for the Longstaff-Schwartz
Algorithm Nicola Moreni
We propose a drift-based variance reduction technique that
applies to the computation of American option prices via the
Longstaff-Schwartz algorithm. It is well known that pricing American
options is not an easy task and that difficulties arise as we increase
the dimension of the underlying (stock) asset. In the unidimensional
case, we dispose of efficient numerical techniques that approximate
prices and that unfortunately cannot be extended to high-dimensional
cases. A possible alternative consists in using Monte Carlo
simulation and in the last decade, many authors proposed algorithms
to price American options on multidimensional assets. We analyse the
one introduced by Longstaff and Schwartz and which is based on a
least squares approach, obtaining an efficient importance sampling
variance reduction technique by means of Girsanov's theorem. The
main idea is to jointly change the drift of the driving Brownian
motion and the form of the payoff function, in order to obtain a set
of equivalent option pricing problems. These pricing problems return
the same price but have different convergence rates and among them,
we search for the one which guarantees the fastest convergence. It
is possible to prove the convergence of the modified algorithm as
well as the existence of an optimal estimator. As it is not
straightforward to locate the optimal estimator, we suggest some
euristic approximations which make the algorithm more time
performing. We analyse in details the numerical implementation and
the results obtained by chosing different payoff functions and
evolution models for the underlying.
| |
An empirically
efficient cascade calibration of the LIBOR Market Model based only on
directly quoted swaption data
Damiano Brigo and Massimo Morini
This work focuses on the swaptions automatic cascade calibration
algorithm (CCA) for the LIBOR Market Model (LMM) first appeared in
Brigo and Mercurio (2001). This method induces a direct analytical
correspondence between market swaption volatilities and LMM
parameters, and allows for a perfect recovery of market quoted
swaption volatilities if a common industry swaptions approximation
is used. We present explicitly an extension of the CCA to calibrate
the entire swaption matrix rather than its upper diagonal part.
Then, while previous tests on earlier data showed the appearance of
numerical problems, we present here different calibration cases
leading to acceptable results. We analyze the characteristics of the
configurations used and concentrate on the effects of different
exogenous instantaneous historical or parametric correlation
matrices. We also investigate the influence of manipulations in
input swaptions data for missing quotes, and devise a new algorithm
maintaining all the positive characteristics of the CCA while
relying only on directly quoted market data. Empirical results on a
larger range of market situation and instantaneous covariance
assumptions show this algorithm to be more robust and efficient than
the previous version. Calibrated sigma parameters are in general
regular and financially satisfactory, as confirmed by the analysis
of various diagnostics implied structures. Finally we Monte Carlo
investigate the reliability of the underlying LMM swaption
analytical approximation in the new context, and present some
possibilities to include information coming from the semi-annual
tenor cap market.
| |
Square-root process
and Asian options
Jayalaxshmi Nagaradjasarma, Angelos Dassios
Although the square-root process has long been used as an
alternative to the Black-Scholes geometric Brownian motion model for
option valuation, the pricing of Asian options on this diffusion
model has never been studied analytically. However, the additivity
property of the square-root process makes it a very suitable model
for the analysis of Asian options. We here develop explicit prices
for digital and regular Asian options. We also obtain a number of
distributional results concerning the square-root process and its
average over time, including analytic formulae for their joint
density and moments. We also show that the distribution is actually
determined by those moments.
| |
Valuation of
Mortgage-Backed Securities Based on Unobservable Prepayment Cost
Processes Hidetoshi Nakagawa and Tomoaki Shouda
We propose a new prepayment model of mortgage in order to
discuss the analysis of mortgage-backed securities (MBS) market. In
our model, it is assumed that each mortgager in the underlying loan
pool shall prepay rationally at the first time when her or his
prepayment cost process falls below zero. Prepayment cost is defined
as a stochastic process that consists of common cost and
idiosyncratic cost. Common cost, called fundamental cost, depends on
refinance rate and is observable at each time, while idiosyncratic
cost is impossible to directly observe in the MBS market (hence so
is the whole prepayment cost). Besides, there are several classes
for mortgagers and the dynamics of idiosyncratic cost is specified
according to the class. MBS investor can only the probability of
which class each mortgager belongs to, but the probability depends
upon an unobservable parameter that is recognized as state variable
peculiar to MBS market. We calculate the conditional distributions
of prepayment cost and the state variable given discrete-time
observation to estimate parameters in the model and to deduce
expected prepayment probability and hazard rate. Moreover we argue
risk-neutrality in MBS market and a fair value of MBS.
| |
Fractional Volatility
Models and Malliavin Calculus Chi Tim Ng and Ngai Hang Chan
The purpose of this thesis is to develop European option pricing
formulae for fractional market models. Although there have been
previous papers discussing the option pricing problem for a
fractional Black Scholes model using Wick calculus and Malliavin
calculus, the formula obtained is similar to the classical Black
Scholes formula, which cannot explain the volatility smile pattern
that is observed in the market. In this thesis, a fractional version
of the Constant Elasticity of Volatility (CEV) model is developed.
An European option pricing formula similar to that of the classical
CEV model is also obtained and a volatility skew pattern is
revealed.
| |
The Futures Market
Model and No-Arbitrage Conditions on the Volatility
Jorgen Aase Nielsen, Kristian R. Miltersen, Klaus Sandmann
Interest rate futures are basic securities and at the same time
highly liquid traded objects. Despite this observation, most models
of the term structure of interest rate assume forward rates as
primary elements. The processes of futures prices are therefore
endogenously determined in these models. In addition, in these
models hedging strategies are based on forward and/or spot contracts
and only to a limited extent on futures contracts. Inspired by the
market model approach of forward rates by Miltersen, Sandmann, and
Sondermann (1997), the starting point of this paper is a model of
futures prices. Using the prices of futures on interest related
assets as the input to the model, new no-arbitrage restricions on
the volatility structure are derived. Moreover, these restrictions
turn out to prevent an application of a market model based on
futures prices.
| |
Higher order numerical algorithms for the solution of some path dependent options pricing problems
Maria R Nogueiras, Alfredo Bermudez, Carlos Vazquez
In this paper we deal with the numerical solution of some degenerate partial differential equations (PDE) or inequalities (PDI). These problems arise, for instance,
when pricing path dependent options, as fixed-strike Amerasian options.
We propose a general methodology that can be applied to one or several factors models.
When there are restrictions on the solution, as early exercise features, the problem is formulated as a mixed problem (including a Lagrange multiplier) and then solved by a numerical iterative algorithm.This algorithm also provides the optimal exercise boundary. For numerical solution of PDE, we propose Lagrange-Galerkin methods of order
two both in time and space. These are combinations of the characteristics method with the finite element method. A rigourous analysis of the stability and convergence
properties of these methods has been developed.
Finally, we will show an application to pricing (two-factors) Asian options, and compare our results with others existing in the bibliography.
| |
An agent market model
using evolutionary game theory Craig Nolder and Benoit Montin
We develope a simple agent based economic model. Stock prices
evolve by trades between agents. Agents learn from the past by
replicator dynamics of evolutionary game theory. By casting the
model as a dynamical game we show from previous results the
existence of a unique limiting distribution which serves as a
stochastic equilibrium. We present a numerical study of this
limiting distribution. It displays features similar to those of
actual financial data.
| |
On utility based super
replication prices Mark Owen
We consider a financial market in which an agent can trade with
only utility-induced restrictions on negative wealth. For a
sufficiently integrable (but possibly unbounded) contingent claim,
we give a representation of the utility-based super-replication
price of the claim as the supremum of its discounted expectations
under pricing measures with finite generalised entropy. Central to
our proof is a bipolar relation between the cone of super replicable
contingent claims with zero initial endowment, and the cone
generated by pricing measures with finite loss-entropy. The cone of
super replicable claims is shown to be the closure, under a relevant
weak topology, of the cone of claims which are super replicable
using only admissible strategies. If the agent has a utility
function which is unbounded above, the set of pricing measures with
finite loss-entropy can be slightly larger than the set of pricing
measures with finite entropy. Indeed, the former set is also the
closure of the latter under a weak topology.
| |
Symmetries and Pricing
of exotic options in Levy models
Antonis Papapantoleon and Ernst Eberlein
This talk surveys symmetries and pricing methods for vanilla and
exotic options in models driven by Lévy processes. Properties of
time-inhomogeneous Lévy processes, more specifically processes with
indepedent increments and absolutely continuous characteristics
(PIIAC), are presented. Empirical research, using data from equity
and bond markets, supports their use for financial modelling
alongside Lévy processes. A method to explore symmetries between different payoffs –which involves a choice of numéraire, a subsequent change of measure and the characteristic triplet of a dual PIIAC– is
described. This allows one to reduce the complexity of problems in
option pricing. Firstly, symmetries between European (plain vanilla)
call and put options are derived and a valuation method is outlined.
Secondly, symmetries between floating and fixed strike Asian and
Lookback options are described and some valuation methods are
discussed. Finally, symmetries for payoffs involving two assets, such
as Margrabe options, are explored.
| |
Satisfying convex risk
limits by trading Traian Pirvu, Kasper Larsen, Steven E Shreve, Reha Tutuncu
A random variable, representing the final position of a
strategy, is deemed acceptable if under each of a variety of
probability measures its expectation dominates a floor associated
with the measure. The set of random variables representing pre-final
positions from which it is possible to trade to final acceptability
is characterized. In particular, the set of initial capitals from
which one can trade to final acceptability is shown to be a closed
half-line. Methods for computing it , and the application of these
ideas to derivative security pricing is developed.
| |
Modeling the
volatility and expected value of a diversified world index
Eckhard Platen
This paper considers a diversified world stock index in a
continuous financial market with the growth optimal portfolio (GOP)
as reference unit or benchmark. Diversified broadly based indices
and portfolios, which include major world stock market indices, are
shown to approximate the GOP. It is demonstrated that a key
financial quantity is the trend of a world index. It turns out that
it can be directly observed since the expected increments of the
index equal four times those of the quadratic variation of its
square root. Using a world stock index as approximation of the
discounted GOP it is shown that, in reality, the trend of the
discounted GOP does not vary greatly in the long term. This leads
for a diversified world index to a natural model, where the index is
a transformed square root process of dimension four. The squared
index volatility appears then as the inverse of the square root
process. This feature explains most of the properties of a stock
index and its volatility.
| |
Optimal Mortgage
Refinancing with Endogenous Mortgage Rates
Stanley Pliska
This paper is based upon earlier work with and by Goncharov,
where risk neutral martingale methods together with the
intensity-based approach borrowed from credit risk modeling were
used to derive new, continuous time results for the value of a
mortgage contract that is subject to prepayment risk. In particular,
with the mortgagor's prepayment behavior suitably modeled by an
intensity process, in the absence of arbitrage opportunities the
interest rate for a fixed rate mortgage was shown to be an
endogenous variable that can be computed. This paper extends these
results by presenting three new developments. First, the earlier
results are derived for a discrete time environment. Second, a
complementary problem is solved, where the mortgage market is fixed,
refinancing incurs transaction costs, and the mortgagor seeks to
choose the refinancing schedule in an optimal manner. The solution
is obtained via an infinite horizon, but nonstandard, Markov
decision chain. Third, this paper addresses and solves an economic
equilibrium problem where the mortgagor is a representative agent,
he/she refinances in an optimal fashion, and the value of the
mortgage rate is such that the mortgage market is free of arbitrage
opportunities. In other words, the resulting solution provides both
the optimal refinancing schedule as well as the endogenous mortgage
rate.
| |
The Critical Kurtosis
Value and Skewness Correction
Vassilis Polimenis
In the empirical option pricing literature, it is generally
agreed that the pronounced volatility smirks are signs of a strongly
negative risk neutral skewness. The paper provides exact conditions
under which skewness in a Lévy process is corrected. The initial
observation is that, unable to dynamically hedge pure jump gamma
risk, agents scale risk neutral volatility. Such scaling violates
arguments in the literature that leptokurtosis always leads to
skewness correction. Actually, it is shown that such arguments are
correct only for symmetric Lévy processes. The general formula for
skewness correction for homogeneous Lévy processes is developed.
Jump induced skewness is corrected only when kurtosis is greater
than a critical value.
| |
Efficient trading
strategies with transactions costs
Vincent Porte and Elyès Jouini
In this article, we characterize efficient contingent claims in
a context of transaction costs and multidimensional utility
functions. The dual formulation of utility maximization helps us
outline the key notion of cyclic anticomonotonicity. Moreover, after
defining a utility price in this multidimensional setting, we
provide a measure of strategies inefficiency and a tool allowing to
effectively compute this measure with the help of cyclic
anticomonotonicity.
| |
Exotic Options: Proofs
Without Formulas Rolf Poulsen
We review how reflection results can be used to give simple
proofs of price formulas and derivations of static hedge portfolios
for barrier and lookback options in the Black-Scholes model.
| |
A Theory of stochastic
integration for bond markets.
Maurizio Pratelli and Marzia DE DONNO
(Based on a joint paper with Marzia De Donno). In the Bond
Market model there is a continuum of securities, and this gives rise
to the problem of what exactly should be meant by the word
"portfolio" in this setting: for this reason, Bjork-DiMasi- Kabanov
and Runggaldier (1997) considered the bond market as a stochastic
process with values in the space of continuous functions C([0,T]),
and introduced a construction of a stochastic integral where the
integrand process (i.e. the mathematical model for a self-financing
strategy) is a process with values in the dual on C([0,T]), i.e. the
set of Radon measures on [0,T]. But this approach, which seems to be
the natural one, has some drawbacks: for instance the "uniqueness"
of martingale probability is equivalent to the "approximate
completeness". We introduce a theory of stochastic integration,
starting from "elementary integrands" (the mathematical
representation of portfolios based on a finite number of bonds) and
going to the limit with "generalized integrands": with our theory,
the stochastic integral is "isometric" and we prove an extension of
a result due to Memin (limit of stochastic integrals, for the
semimartingale topology, is still a stochastic integral). This gives
a good answer to the problem of completeness, and with an
appropriate definition of "admissible" strategy it is possible to
extend to this infinite dimensional setting the method of convex
duality for the problem of "utility maximization". But the
"generalized integrands" cannot be characterized in an explicit way:
with further regularity assumptions we can state explicitly which
kind of processes are integrable. Finally, we give some applications
of these results to the Bond Market models.
| |
Weak Convergence of
Option Quantile Hedging Strategies
Jean-Luc Prigent
Option hedging of contingent claims is a prominent problem in
finance. This problem is straightforward when dealing with complete
markets. However, most of financial markets are incomplete. In that
case, the risk-neutral probability is no longer unique and
contingent claims are not all attainable. The problem of hedging can
be seen by another point of view: what should an investor do if he
is unwilling to invest all the amount needed for a perfect hedging
or a superhedging strategy ? Alternatively, we can ask: which
initial amount the investor can save by accepting a certain fixed
``shortfall'' probability? Two seminal papers of Follmer and Leukert
(1999), (2000) deal with quantile hedging. In Prigent (1999) and
Prigent and Scaillet (2002), it is proved that both prices and
hedging strategies associated to the locally risk-minimizing
criteria are stable under weak convergence. The goal of this paper
is to examine the same problem for the quantile approach. It is
proved that the stability under convergence is generally satisfied
in the complete case. Nevertheless, for the incomplete case, usually
there is no longer stability under weak convergence. This property
has potential applications since it indicates that we have to take
care when using quantile approach from the convergence point of
view. AMS 2000 classification : 60 F05, 91 B 28.
| |
A Chaotic Approach to
Interest Rate Modelling Avraam Rafailidis and Lane Hughston
This paper presents a new approach to interest rate dynamics. We
consider the general family of arbitrage-free positive interest rate
models, valid on all time horizons, in the case of a discount bond
system driven by a Brownian motion of one or more dimensions. We
show that the space of such models admits a canonical mapping to the
space of square-integrable Wiener functionals. This is achieved by
means of a conditional variance representation for the state price
density. The Wiener chaos expansion technique is then used to
formulate a systematic analysis of the structure and classification
of interest rate models. We show that the specification of a
first-chaos model is equivalent to the specification of an
admissible initial yield curve. A comprehensive development of the
second-chaos interest rate theory is presented in the case of a
single Brownian factor, and we show that there is a natural
methodology for calibrating the model to at-the-money-forward caplet
prices. The factorisable second-chaos models are particularly
tractable, and lead to closed-form expressions for options on bonds
and for swaptions. In conclusion we outline a general
"international" model for interest rates and foreign exchange, for
which each currency admits an associated family of discount bonds,
and show that the entire system can be generated by a vector of
Wiener functionals.
| |
Nonparametric
estimation of the diffusion coefficient via Fourier analysis, with an
application to short interest rates
Roberto Reno'
We introduce an original fully nonparametric estimator of the
diffusion coefficient of an univariate diffusion, which makes use of
discrete observations. Infill asymptotic properties of the estimator
are fully derived: the estimator is proven to be consistent and
asymptotically normally distributed. Monte Carlo simulations are
conducted to explore the small-sample properties of the estimator
and compare it to already known estimators. Finally, the estimator
is implemented on short rate time series and the diffusion
coefficient is estimated.
| |
A Two-Factor Model for
Commodity Prices and Futures Valuation
Diana Ribeiro and Stewart Hodges
This paper develops a reduced form two-factor model for
commodity spot prices and futures valuation. This model extends
Schwartz's (1997) two-factor model by adding two new features. First
the Ornstein-Uhlenbeck process for the convenience yield is replaced
by a Cox-Ingersoll-Ross (CIR) process. This ensures that our model
is arbitrage-free. Second, spot price's volatility is proportional
to the square root of the convenience yield level. We empirically
test both models using weekly crude oil futures data from 17th of
March 1999 to the 24th of December 2003. In both cases, we estimate
the model parameters using the Kalman filter.
| |
Correcting for
Simulation Bias in Monte Carlo methods to Value Exotic Options in Models
Driven by Lévy Processes Claudia Ribeiro and Nick Webber
Lévy processes can be used to model asset return's
distributions. Monte Carlo methods must frequently be used to value
path dependent options in these models, but Monte Carlo methods can
be prone to considerable simulation bias when valuing options with
continuous reset conditions. In this paper we show how to correct
for this bias for a range of options by generating a sample from the
extremes distribution of the Lévy process on subintervals. We work
with the variance-gamma and normal inverse Gaussian processes. We
find the method gives considerable reductions in bias, so that it
becomes feasible to apply variance reduction methods. The method
seems to be a very fruitful approach in a framework in which many
options do not have analytical solutions.
| |
A synthetic measure of
multivariate risk and its empirical implications for portfolio risk
management Andrea Roncoroni and Stefano Galluccio
We define a new measure of multivariate risk based on the
typical shapes displayed by term structure movements. The resulting
cross-shape covariance is decomposed into uncorrelated shape
factors. Empirical tests conducted on a U.S. database over a wide
range of hedging scenarios suggest that these factors represent more
accurately the yield curve risk than those stemming from the
classical principal components analysis of cross-yield covariance.
| |
Estimating the
Commodity Market Price of Risk for Energy Prices
Ehud Ronn and Sergey P. Kolos
The purpose of this paper is to determine the magnitude and sign
of the commodity "market price of risk" (MPR) for electricity and
natural-gas prices. This MPR variable determines whether forward
prices in energy are upward- or downward-biased predictors of future
expected spot prices. We evaluate that risk premium by estimating
the drift term in spot and forward prices. In futures prices, we
explicitly account for the Samuelson hypothesis "term structure of
volatility." In spot prices of electricity, we examine the
relationship between Day-Ahead Prices and Real-Time Prices. The
results have implications for understanding the relationship between
energy markets and other physical and financial markets, for
incorporating the risk premium in making informed hedging decisions
in industry, and for relating futures prices to the forecast prices
produced by industry.
| |
Pathwise Optimality
for Benchmark Tracking Wolfgang Runggaldier, Paolo Dai Pra, Marco Tolotti
We consider the problem of investing in a portfolio in order to
track a given benchmark. We study this problem from the point of
view of almost sure/pathwise optimality. We first obtain a control
that is optimal in the mean and this control is then shown to be
also pathwise optimal. The standard Merton model leads to
lognormality of the value process so that it does not possess the
ergodic properties required for pathwise optimality. We obtain
ergodicity by transforming the process so that it remains bounded
thereby using a method that can be related to random time change. We
furthermore describe a general approach to solve the
Hamilton-Jacobi-Bellman equation corresponding to the given problem
setup.
| |
A numerical study of
the smile effect in implied volatilities induced by a nonlinear feedback
model Simona Sanfelici, Maria Elvira Mancino, Shigeyoshi Ogawa
The Black-Scholes model requires that the volatility is
constant. Nevertheless this hypothesis is recognized to cause option
price distortions. In order to give explanation to this behavior the
literature has proposed both of modelling directly volatility as a
stochastic process either of relaxing the hypothesis on which
Black-Scholes model is based, such as the facts that the market is
complete, frictionless and that the agents act as price takers. We
propose a model where the price is defined in equilibrium with some
traders employing a trading strategy to hedge a contingent claim. As
a consequence we get a fully nonlinear partial differential equation
which implicitly incorporates the demand induced by hedging. In this
paper we present numerical results which show the impact of the
hedging strategies for derivative asset analysis. In particular we
obtain the smile pattern and term structure of implied volatility.
The numerical results show that in our model the implied volatility
smile can be reproduced as a consequence of dynamical hedging. The
simulations are performed using the Infinite Elements Method.
| |
Utility maximization
for unbounded processes Sara Biagini and Frittelli Marco
When the price processes of the financial assets are described
by possibly unbounded semimartingales, the classical concept of
admissible trading strategies may lead to a trivial utility
maximization problem because the set of bounded from below
stochastic integrals may be reduced to the zero process. However, it
could happen that the investor is willing to trade in such a risky
market, where potential losses are unlimited, in order to increase
his/her expected utility. We translate this attitude into
mathematical terms by employing a class $\mathcal{H}^{W}$ of
$W-$admissible trading strategies which depend on a loss random
variable $W$. These strategies enjoy good mathematical properties
and the losses they could give rise to in the trading are compatible
with the preferences of the agent. We formulate and analyze by
duality methods the utility maximization problem on the new domain
${H}^{W}$. We show that, for all loss variables W contained in a
properly identified set $\mathcal{W}$, the optimal value on the
class $\mathcal{H}^{W}$ is constant and coincides with the optimal
value of the maximization problem over a larger domain ${K}_{\Phi
}.$ The class ${K}_{\Phi }$ doesn't depend on the single
$W\in\mathcal{W},$ but it depends on the utility function $u$
through its conjugate function $\Phi $. By duality methods we show
that the optimal solution exists in ${K}% _{\Phi }$ and it can be
represented as a stochastic integral that is a uniformly integrable
martingale under the minimax measure. We provide the economic
interpretation of the larger class ${K}%_{\Phi }$ and we analyze
some examples that show that this enlargement of the class of
trading strategies is indeed necessary.
| |
Asymptotic Analysis
for American Options on Alternative Stochastic Processes
David Saunders and John Chadam
We study the free-boundary problem arising from the pricing of
the finite horizon American put option. We extend known results on
the asymptotic behaviour of the optimal exercise boundary (critical
stock price) to the case where the underlying follows a process
other than geometric Brownian motion. In particular, we derive
higher order asymptotic expansions of the boundary near expiry, as
well as approximations for larger times. Our results cover both the
case of a diffusion process with a volatility surface (e.g. the
constant elasticity of variance, or CEV model) and a class of jump
diffusion processes.
| |
The Valuation of
Callable Contingent Claims with Applications
Katsushige Sawaki and Susumu Seko
The purpose of this paper is to introduce and analyze callable
contingent claims with an application to callable American options.
These contingent claims are bilateral contracts which entitle both a
seller(issuer) and a buyer(holder) to exercise the rights
respectively. The seller can cancel(call) the contract by paying
some penalty to the buyer and the buyer can exercise(put) the right
at any time before the maturity of the contract. In this paper we
establish a pricing formula for the callable contingent claims that
can be decomposed into the difference of the value of an American
option and the callable discount. Moreover, following our
decomposition of the callable contingent claim, the value of the
claim can be solved only with the numerical technique based upon the
discrete binomial model, which provides a visual understanding of
the analytical properties of the value.
| |
Analytic American
Option Pricing: The Flat-Barrier Lower Bound
Alessandro Sbuelz
I study the pricing performance of a closed-form lower bound to
American option values based on an exercise strategy corresponding
to a flat exercise boundary. The lower bound has a simple two-step
implementation akin to the Barone-Adesi and Whaley (1987) formula
and shows superior pricing performance in the Out-of-The-Money
region and for long maturities. JEL-Classification: G12, G13.
Keywords: American Options, Flat Barrier Options.
| |
Risk measures and
capital requirements for processes
Giacomo Scandolo
In this paper we propose a generalization of the concepts of
convex and coherent risk measures to a multi-period setting, in
which payoffs are spread over different dates. To this end, a
careful examination of the axiom of translation invariance and the
related concept of capital requirement in the one-period date is
performed. These two issues are then suitably extended to the
multi-period case, in a way that makes their operative financial
meaning clear. A characterization in terms of expected values is
derived for this class of risk measures and some examples are
presented.
| |
Evolutionary Stable
Stock Markets Klaus Reiner Schenk-Hoppé, Igor V. Evstigneev, Thorsten Hens
This paper shows that a stock market is evolutionary stable if
and only if stocks are evaluated by expected relative dividends. Any
other market can be invaded by portfolio rules that will gain market
wealth and hence change the valuation. In the model the valuation of
assets is given by the wealth average of the portfolio rules in the
market. The wealth dynamics is modelled as a random dynamical
system. Necessary and sufficient conditions are derived for the
evolutionary stability of portfolio rules when (relative) dividend
payoffs form a stationary Markov process. These local stability
conditions lead to a unique evolutionary stable strategy according
to which assets are evaluated by expected relative dividends.
| |
Optimal investments
for robust utility functionals
Alexander Schied
We introduce a systematic approach to the problem of maximizing
the robust utility of the terminal wealth in a general market model,
where the robust utility functional is defined via a lower
expectation of a set Q of probability measures. In a complete model,
this problem can be reduced to determining a least favorable measure
in Q. This least favorable measure often is universal in the sense
that it does not depend on the particular utility function and can
be constructed by using the Huber-Strassen theory from robust
statistics. Applications include the problem of robust utility
maximization with uncertain drift and optimal investment under weak
information. Building on results by Kramkov and Schachermayer, we
also discuss the duality theory for general incomplete market
models.
| |
Infinite Factor Models
for Credit Risk Thorsten Schmidt
The defaultable term structure is modeled in two ways, both in
an infinite dimensional framework. First, we use stochastic
differential equations in Hilbert spaces, which allows for
incorporating different recovery scenarios as well as ratings.
Second, a special case is described via transformations of Gaussian
random fields. This approach concentrates on deriving explicit
pricing formulas and calibration of the model. In the calibration
procedure we account for the fact, that credit derivatives data is
scarce, so the model can be calibrated using only a small set of
credit derivatives. Hedging issues are also discussed.
| |
Information-Driven
Default Contagion Philipp Schoenbucher
Much of the existing literature on default contagion assumes a
direct causal relationships between two obligors' defaults. In this
paper we present a model in which default contagion arises without
causal links solely from information effects if investors are
imperfectly informed about some common factors affecting the true
riskiness of the obligors. We model this effect in a simple
extension of the intensity-based modelling framework using
unobserved frailty variables. This allows the modelling of much
higher (and more realistic) levels of default dependence between the
obligors than what purely diffusion-based intensity models were able
to capture previously, without adding too much additional
complexity. The parameters of the dependence can be implied directly
from spread jumps observed in the market, thus enabling a full
specification of the model under pricing probabilities without
recourse to historical default correlations. We furthermore present
an extension of the model in which the size of the contagion effect
can depend on the reason for the default and not just the identity
of the defaulted obligor, thus introducing stochastic dependency.
| |
Pricing Swaptions in
Affine Term Structure Models David Schrager and A. A. J. Pelsser
We propose an approach to find an approximate price of a
swaption in Affine Term Structure Models. Our approach is based on
the derivation of approximate dynamics in which the volatility of
the Forward Swap Rate is itself an affine function of the factors.
Hence we remain in the Affine framework and well known results on
transforms and transform inversion can be used to obtain swaption
prices in ways similar to bond options (i.e. caplets). We
demonstrate that we can even obtain a closed form formula for the
approximate price which is based on square-root dynamics for the
swap rate. The latter approximation is extremely fast to compute
while remaining accurate. Computational time compares favorably with
other known approximation methods. The results on the quality of the
approximation are excellent. Our results show that, analogously to
the LIBOR Market Model, LIBOR and Swap rates are driven by
approximately the same type of (in this case affine) dynamics.
| |
On the Martingale
Measures in Exponential Levy Models
Andrey Selivanov
We study the existence and the uniqueness of martingale measures
in 1) exponential Levy models and 2) time-changed exponential Levy
models. We consider models with finite and infinite time horizon. By
the First Fundamental Theorem of Asset Pricing our problem is
equivalent to the problem of absence of arbitrage (and completeness)
of the corresponding model. We discuss two arbitrage concepts:
traditional Free Lunch with Vanishing Risk and so-called Generalized
Arbitrage. We obtain criteria for the absence of arbitrage for the
first model both with finite and infinite time horizon and for the
second model with finite time horizon.
| |
Bermudan Guaranteed
Return Contracts: Analysis and Valuation
Steven Simon
A guaranteed or minimum return can be found in different
financial products, e.g. 'guaranteed investment contracts' (GIC's)
issued by investment banks and life-insurance contracts. We consider
the so-called multi-period or compounding guaranteed return
contracts. With such a contract a minimum return is guaranteed over
a series of sub-periods, making the pay-off at maturity
path-dependent. We first derive some important features of the
European style contract in a general Heath-Jarrow-Morton framework.
Secondly, we analyze the effect of adding a Bermudan put feature to
this type contract and we derive some interesting properties about
the optimal exercise behavior. More precisely, we derive the
sufficient and necessary conditions on the dynamics of the
underlying asset for the optimal exercise decision to be a stopping
time with respect to the term structure of interest rates.
| |
The Merton Problem in
an Illiquid Financial Market
Surbjeet Singh and L.C.G. Rogers
Liquidity is an important effect in the markets, yet it is hard
to come up with a definition which not only has some economic
explanation but also retains a reasonable degree of tractability. In
this paper, we propose a simple microeconomic model in discrete time
which carries over to the continuous-time setting; this results in a
modification of the usual dynamics of portfolio wealth, which
appears to be impossible to analyse exactly. We investigate the
Merton problem numerically, and through some asymptotic analysis.
| |
A Two-Person Game for
Pricing Convertible Bonds Mihai Sirbu and Steven E. Shreve
A firm issues a convertible bond. At each subsequent time, the
bondholder must decide whether to keep the bond, thereby collecting
coupons, or to convert it to stock. The bondholder wishes to choose
a conversion strategy to maximize the bond value. Subject to some
restrictions, a convertible bond can be called by the issuing firm,
which presumably acts to maximize equity value and thus to minimize
the bond value. This creates a two-person game, and we model the
bond price as the value of this game. We show, however, that under
our standing assumption (dividends are paid at a lower rate than the
money market rate) this game reduces to one of two optimal stopping
problems, and the relevant stopping problem can be determined a
priori, i.e., without first solving the convertible bond pricing
problem. Because of dividend payments, the partial differential
equation describing the pricing function becomes nonlinear. This
means that our analysis involves a fixed point problem. We also
prove that for large time to maturity the value of the convertible
bond approaches the value of the perpetual convertible bond. The
presentation is based on joint work with Steven E. Shreve.
| |
Asymptotic Option
Pricing under a Pure Jump Process
Seongjoo Song
This paper studies the problem of option pricing in an
incomplete market. Under the market incompleteness from the
discontinuity of the asset price process, we try to find a
reasonable price for a European contingent claim by adopting an
asymptotic approach. First, we find the unique minimal martingale
measure and get a price by taking an expectation of the payoff under
this measure. We also show that it converges weakly to the
equivalent martingale measure in the limit. To get a closed-form
price, we use an asymptotic expansion. In case where the minimal
martingale measure is a signed measure, we use a sequence of
martingale measures that converges to the equivalent martingale
measure in the limit to compute the price of an option.
| |
Good Deal Bounds for
Valuation of Real and Financial Options
Jeremy Staum
We explore an approach to constructing good deal bounds for
pricing contingent claims in incomplete markets. The approach
applies to trades in financial options and to real investment
decisions, which may involve real options. It is grounded in
equilibrium theories (such as the capital asset pricing model) which
support the decision tools (such as net present value and real
options analyses) currently used in practice to make real investment
decisions. An equilibrium model yields a specification of the
acceptance set used to construct good deal bounds. Good deal bounds
enable trading and investment decisions to incorporate robustness or
aversion to ambiguity about stochastic models, for instance,
regarding market risk premia.
| |
Optimal Investments in
Markets with Stochastic Opportunity Sets
Sasha Stoikov and Thaleia Zariphopoulou
A class of optimal investment and consumption models in
incomplete market environments will be analyzed. The focus will be
on a universal characterization of the optimal portfolios (myopic
and excess risky demand) in terms of hedging strategies of
supporting pseudoclaims. These claims are written on the market
price of risk and are priced by indifference. Recent results on
indifference prices will be used for the sensitivity and robustness
analysis of the optimal investments. Issues related to model
specification and to the interplay between market incompleteness and
risk preferences will also be discussed.
| |
Optimal Statistical
Decisions About Some Alternative Financial Models
Wolfgang Stummer and Igor Vajda
We deal with Bayesian decisions between the following two models
for the price dynamics X(t) of a financial asset X: (i) the
prominent geometric Brownian motion, which is a special diffusion
process with constant growth rate and constant volatility, and (ii)
a (generally non-lognormally-distributed) diffusion process with
non-constant growth rate but with the same volatility. We derive
some bounds on the corresponding decision-theoretic quantities such
as e.g. the Bayes factor, the Bayes loss (minimal mean loss) and the
Bayes probability of (discrimination) error. This is achieved with
the help of obtained bounds on generalized relative entropies
between (i) and (ii). Furthermore, we also investigate the total
variation distance between the two models.
| |
Endogenous Risk
Aversion and Ockham's Razor
Michael Stutzer
Influential early articles by Paul Samuelson advocated use of
expected concave utility of wealth criteria in T-repeated betting
and investment problems. He and other founders of modern decision
theory viewed their work as normative prescriptions for choice under
uncertainty; not just as predictive theories of pre-existing
behavior. Results of Rabin (Econometrica, 2000), exposited and
applied in Rabin and Thaler (Journal of Economic Perspectives, 2001,
pp. 219-232), directly challenged both the prescriptive and
predictive usefulness of ANY expected concave utility criterion in
these settings. As a predictive alternative, they advocated the use
of loss averse preferences as a substitute for expected concave
utility. While different systems of preference axioms have been
found that imply the use of specific alternatives to concave
utility, they are not normatively convincing. Moreover, none of them
do anything to solve a problem that plagues both the prescriptive
and predictive use of expected utility: results critically depend on
adjustable preference parameters that are difficult to directly
measure, and hence must be "fit" to the observable data that they
attempt to explain. As a simpler alternative criterion, this paper
proposes the probability of outperforming an observable benchmark
the agent wants to beat. This criterion does not suffer from the
possible ills of some other probabilistic criteria that were
(influentially) critiqued by Samuelson. Large deviations theory is
used to show that for suitably large T, this criterion is equivalent
to maximizing an expected CRRA (power) habit-formation utility, but
with a coefficient of risk aversion that varies endogenously with
the alternative evaluated. This eliminates the adjustable curvature
parameter used in other expected and non-expected utility (e.g. loss
aversion) preference theories, in accord with the scientific
principle of parsimonious parameterization called Ockham's Razor. A
Bayesian formulation of Ockham's Razor is used to illustrate the
advantages of this parsimony.
| |
A Bayesian Learning
Model of Risk Taking by Fund Managers
Ajay Subramanian, Ping Hu, Jayant Kale 
We propose a multi-period bayesian learning model to examine the
impact of explicit and implicit incentives on the risk taking
behavior of fund managers. We show that implicit incentives arising
from career concerns and, especially, employment risk lead to a
non-monotonic U-shaped relation between a fund manager's relative
risk choices and her prior performance relative to her peers.
Moreover, the incentive to increase relative risk, ceteris paribus,
declines with the manager's experience. Our empirical analysis of
fund managers' risk taking behavior documents significant support
for our theory. Our study therefore demonstrates the importance of
both explicit and implicit incentives in driving the risk-taking
behavior of fund managers.
| |
On an Alternative
Approach to Pricing General Barrier Options
Michael Suchanecki
In this paper, an alternative approach to pricing barrier
options is presented that relies on the use of the first hitting
time density to the barrier. Laplace transforms with respect to time
are used in order to determine option prices. It turns out that this
approach allows for pricing more general barrier options (at least
numerically). As an example, a simple down-and-in call option is
considered. In this case, all Laplace transforms (even the finite
Laplace transforms) can be inverted analytically and the well-known
closed form pricing formula is obtained.
| |
Environment
&Financial Markets
Wojciech Szatzschneider and Monique Jeanblanc
We argue that practical solutions for the environmental
degradation are in a short supply. Most of the increasingly complex
models set off different opinions about their applicability. Models
should be well specified. This requirement is hard to meet in
environmental studies. Thus, the efficient global environmental
decision--making becomes very difficult. Moreover politicians often
tend to justify their decisions by inappropriate theories. This
situation leads to proliferation of ineffective studies and waste of
resources. We shall propose to apply the market approach in the
solutions of several environmental problems. It could result in more
transparent transfer of funds and the involvement of everybody
concerned. Also we can expect that the transparency could stem in an
increment of these funds. We shall focus on the issue of
deforestation due to its importance for the global well--being, and
the possibility to assess the number of trees. Our approach is based
on a positive involvement of holders of "good" options bought or, in
the first stage, obtained for free. In the case of the forest "good"
means a kind of Asian call option. We will show that, in a natural
way, three kinds of optimization problems crop up: 1)Individual
agent problem. 2)Local optimization problem. 3)Global optimization
problem. The first one is how the holder of a good option could
eventually contribute to reforestation. The second one is how to
choose prices of options, to maximize the space mean of the temporal
mean of the "asset" in given place. The last one is how to
distribute funds into particular projects. In the final part we
analyze the dynamical control for bounded processes and awards
partially based on the mean of the underlying value.
| |
A new fast and
accurate method to calculate Value-at-Risk and other tail risk measures
Tanya Tamarchenko and Rabi De
We present a highly efficient methodology for calculating
Value-at-Risk (VAR) and other tail-risk measures for a portfolio of
derivative securities. This new methodology, named "Reliability-VAR"
provides an answer to the question that is important to risk
managers: what are the most likely values of underlying risk factors
that cause a loss of a certain size? We show that the ability to
answer this question enables us to construct highly efficient hybrid
numerical procedures to calculate the tail of distribution for
portfolio losses. We present numerical examples that demonstrate the
superior performance of Reliability VAR in comparison to commonly
used VAR methods.
| |
Pricing CEV moving
barrier options with time-dependent parameters — Lie algebraic
approach Hoi-man Tang, C.F. Lo, C.H. Hui
In this paper we apply the Lie-algebraic technique for the
valuation of moving barrier options with time-dependent parameters.
The value of the underlying asset is assumed to follow the constant
elasticity of variance (CEV) process. By exploiting the dynamical
symmetry of the pricing partial differential equations, the new
approach enables us to derive the analytical kernels of the pricing
formulae straightforwardly, and thus provides an efficient way for
computing the prices of the moving barrier options. The method is
also able to provide tight upper and lower bounds for the exact
prices of CEV barrier options with fixed barriers. In view of the
CEV model being empirically considered to be a better candidate in
equity option pricing than the traditional Black-Scholes model, our
new approach could facilitate more efficient comparative pricing and
precise risk management in equity derivatives with barriers by
incorporating term-structures of interest rates, volatility and
dividend into the CEV option valuation model.
| |
Solvable affine term
structure models Claudio Tebaldi and Martino Grasselli
Pricing of contingent claims in the Affine Term Structure Models
(ATSM) can be reduced to the solution of a set of Riccati-type
Ordinary Differential Equations (ODE), as shown in Duffie, Pan and
Singleton (2000) and in Duffie, Filipovic and Schachermayer (2003).
We discuss the solvability of these equations. While admissibility
is a necessary and sufficient condition in order to express their
general solution as an analytic series expansion, we prove that,
when the factors are restricted to have continuous paths, these ODE
admit a fundamental system of solutions if and only if all the
positive factors are independent. Finally, we classify and solve all
the consistent polynomial term structure models admitting a
fundamental system of solutions.
| |
Optimal Portfolio
Strategies with Different Constraints: A Unified Treatment
Phelim Boyle and Weidong Tian
The traditional portfolio selection problem concerns an agent
who wishes to maximize expected utility of consumption and or
terminal wealth during some planning horizon. This basic problem can
be modified by adding constraints and portfolio insurance provides
one example. In this paper we investigate the portfolio selection
problem under different constraints within a unified framework.
These constraints reflect differing objectives and in this paper we
analyze three types of objectives. First we consider an investment
strategy which aims to maximize expected utility subject to a
guaranteed (or minimal performance) wealth level. Under this
strategy a benchmark index is chosen to reflect the particular
objective. The benchmark index can be either deterministic or
stochastic. The second strategy is designed to maximize expected
utility with the added objective of taking on investment risk with
the aim of realizing higher returns. In this case the constraint is
expressed in probabilistic terms as for example in a conventional
Value at Risk framework. Once again the benchmark index can be
either deterministic or stochastic. Under the third strategy a
portfolio manager follows a mixed strategy, by combining the
previous two investment objectives. In this case the investor
maximizes expected utility while achieving a guaranteed return and
also maintaining features of the second strategy as well. We analyze
these three investment problems and provide explicit solutions and
discuss several applications.
| |
Efficient Computation
of Hedging Parameters for Discretely Exercisable Options
Stathis Tompaidis, Ron Kaniel, Alexander Zemlianov
We describe a method to obtain bounds on the hedging parameters
of discretely exercisable options using Monte-Carlo simulation. The
method is based on a combination of the duality formulation of the
optimal stoping problem for pricing discretely exercisable options
and Monte-Carlo estimation of hedging parameters for European
options. For a given computer budget and exercise strategy we
provide an algorithm that achieves the tightest bounds. The method
can handle arbitrary payoff functions, general diffusion processes,
and a large number of random factors. We also present a fast,
heuristic, alternative method and use our method to evaluate its
accuracy.
| |
Flexible Complete
Models with Stochastic Volatility: Generalising Hobson & Rogers (1998)
Robert Tompkins, Friedrich Hubalek, Josef Teichmann
Hobson and Rogers (1998) propose an option pricing model where
the volatility is a deterministic function of the moving average of
past (logarithm of) underlying prices. They show that such a model
can also generate implied volatilities that vary across striking
price and term to expiration. In this research, this model is tested
on actual option markets. While the Hobson and Rogers (1998) model
produces divergences from Black-Scholes (1973) prices on a
microscopic scale, we have not been able to replicate actual option
prices with this model. To determine prices from this model we
develop a robust analytic approximation. To better fit observed
options prices, we generalise the model Hobson and Rogers (1998) by
the addition of two additional parameters. This model is able to
match option prices on the British Pound/US Dollar across both the
striking price dimension (smiles) and across different maturities
(the term structure of implied volatility). By use of Mavillian
calculus, we are able to determine partial derivatives of the
generalised model and compute hedging ratios.
| |
Superreplication of
options on several underlying assets
Johan Tysk, Erik Ekström, Svante Janson
We investigate when a hedger who over-estimates the volatility
with a time- and level-dependent volatility model will
superreplicate a convex claim on several underlying assets. It is
shown that the classical Black-Scholes model is the only model,
within a large class, for which over-estimation of the volatility
yields the desired superreplication property for any convex claim.
This is in contrast to the one-dimensional case, in which it is
known that over-estimation of the volatility with any model
guarantees superreplication of convex claims. The proof is based on
the fact that preservation of convexity of solutions to parabolic
partial differential equations, with no lower order terms, is a
generic property in one spatial dimension but is very rare in higher
dimensions.
| |
A new jump-diffusion
model and performances of affine stochastic volatility models for equity
emerging markets Rosanna Pezzo and Mariacristina Uberti
Most traditional models fail short of completely capturing several volatility properties, such as time variation, clustering, mean reversion and volatility smile. A good volatility model should be able to reflect the above properties as well as take into account the fact that positive and negative shocks in returns have rather different impact on the volatility. This leads us to involve affine stochastic volatility models with jumps. While in literature similar models are usually applied to option pricing, Pezzo-Uberti (2002) showed how Duffie-Pan-Singleton's affine jump-diffusion model with jumps in returns and volatility outperforms the stochastic volatility model without jumps in volatility forecasting.
In this paper we propose a new affine jump-diffusion model for volatility, in which returns and volatility are generated by two independent stochastic processes and shocks are driven by separate jump processes. We model the volatility jumps with a different kind of positive distribution: an Inverted-Gamma. Some interesting theoretical results involve the jump joint process that turns out to follow a Cauchy distribution. We calibrate and use the model to simulate the volatility behaviour of assets in very volatile markets. A good match with real data is pointed out and in several occasions the model performs better than the existing models.
| |
Criterions for
absolute continuity and singularity of measures via separating
times Mikhail Urusov and A. Cherny
We introduce a notion of separating time for a pair of measures
P and Q on a filtered space. This notion is convenient for
describing the mutual arrangement of P and Q from the viewpoint of
absolute continuity and singularity. Furthermore, we find an
explicit form of the separating time for the case, where P and Q are
distributions of Levy processes, solutions of stochastic
differential equations, and distributions of Bessel processes. The
obtained results yield, in particular, criteria for local absolute
continuity, absolute continuity, and singularity of P and Q.
| |
Portfolio Analysis
with General Deviation Measures
Michael Zabarankin, R. Tyrrell Rockafellar, Stan Uryasev
Generalized measures of deviation, as substitutes for standard
deviation, are considered in a framework like that of classical
portfolio theory for coping with the uncertainty inherent in
achieving rates of return beyond the risk-free rate. Such measures,
associated for example with conditional value-at-risk and its
variants, can reflect the different attitudes of different classes
of investors. They lead nonetheless to generalized one-fund theorems
as well as to covariance relations which resemble those commonly
used in capital asset pricing models (CAPM), but have wider
interpretations. A more customized version of portfolio optimization
is the aim, rather than the idea that a single "master fund" might
arise from market equilibrium and serve the interests of all
investors. Through techniques of convex analysis, they deal
rigorously with a number of features that have not been given much
attention in this subject, such as solution nonuniqueness, or
nonexistence, and a potential lack of differentiability of the
deviation expression with respect to the portfolio weights. Moreover
they address in detail the previously neglected phenomenon that, if
the risk-free rate lies above a certain threshold, a master fund of
the usual type will fail to exist and need to be replaced by one of
an alternative type, representing a "net short position" instead of
a "net long position" in the risky instruments.
| |
Pricing of arithmetic
Asian options and basket options by conditioning on more than one
variable Michèle Vanmaele, Deelstra Griselda, Liinev Jan
Pricing of arithmetic Asian options and basket options where the
underlying is modelled in a Black-Scholes setting, boils down to
computing stop-loss premia of sums of dependent lognormal random
variables. In earlier work we derived lower and upper bounds for
these types of options based on comonotonicity results and by
conditioning upon one variable. A natural extension is to condition
on more than one variable, because one intuitively expects in this
way to improve those bounds. We derive analytical expressions for
the comonotonic bounds of stop-loss premia of sums of dependent
random variables when conditioning on more variables. As expected,
the dimension of the integrals in the comonotonic bounds increases
with the number of conditioning variables. However in case of two
conditioning variables, the lower bound uses only one integration if
the conditional density function is known. We specify this lower
bound in the case of a sum of lognormal variables. Numerical results
show that conditioning on two variables leads to very sharp lower
bounds.
| |
Comparison Theorem and
Option Pricing in the Presence of Jumps Jan
Vecer and Mingxin Xu
It is well known that in diffusional framework, the process with
the largest volatility term in absolute value has the largest
expectation if evaluated in some convex function (under certain mild
technical conditions). This result has immediate applications to
option pricing and stochastic optimal control problems. Since option
prices are computed as expectations of convex payoffs corresponding
to the stock price process, one expects that the price will be
higher for stocks with higher volatility. The aim of this talk is to
explore possibilities of generalizing this result to processes which
exhibit jumps. One might expect that the larger is the jump size,
the larger would be the convex payoff, but this turns out to be not
true in general. This talk will discuss these issues and the
consequences for option pricing and other areas of stochastic
analysis. (Joint work with Mingxin Xu).
| |
Expanding the Universe
of Exotic Options Closed Pricing Formulas in the Black Sholes
Framework Carlos Veiga
A pricing method resulting in a closed formula is proposed for a
large class of options such as Best Of and Rainbow based on the
analysis of the return profile of the option. We assume that returns
follow a brownian motion and the usual hypotheses of the
Black-Scholes model extended for the multi-underlying/multi-currency
case. The result states that, if the pay-of of an option is a linear
combination of the prices at maturity of traded assets multiplied by
an indicator function generated by an exercise condition, then, the
pricing formula is also a linear combination of the current market
prices of the traded assets multiplied by a probability expressed in
the risk-neutral measure where the asset is the risk-free asset. The
proof of the result follows the self-financing portfolio reasoning
for the multi-underlying/multi-currency case and uses the change of
numeraire technique. We show that the well-known results by Black
and Scholes (1973), Magrabe (1978) and Johnson (1987) follow as a
particular case of our result and we show how to price certain new
options. Comparison results of simulation are also presented.
| |
Pricing Power
Derivatives: a two-factor jump-diffusion approach
Pablo Villaplana
We propose a two-factor jump-diffusion model with seasonality
for the valuation of electricity future contracts. The model we
propose is an extension of Schwartz and Smith (Management Science,
2000) long-term / short-term model. One of the main contributions of
the paper is the inclusion of a jump component, with a non-constant
intensity process (probability of occurrence of jumps), in the
short-term factor. We model the stochastic behaviour of the
underlying (unobservable) state variables by Affine Diffusions (AD)
and Affine Jump Diffusions (AJD). We obtain closed form formulas for
the price of futures contracts using the results by Duffie, Pan and
Singleton (Econometrica, 2000). We provide empirical evidence on the
observed seasonality in risk premium, that has been documented in
the PJM market. This paper also complements the results provided by
the equilibrium model of Bessembinder and Lemmon (Journal of
Finance, 2002), and provides an easy methodology to extract
risk-neutral parameters from forward data.
| |
Liquidity Risk and
Corporate Demand for Hedging and Insurance
Stephane Villeneuve and Rochet Jean-Charles
We analyze the demand for hedging and insurance by a corporation
that faces liquidity risk. Namely, we consider a firm that is
solvent (i.e. exploits a technology with positive expected net
present value) but potentially illiquid (i.e. that may face a
borrowing constraint). As a result, the firm's optimal liquidity
management policy consists in accumulating reserves up to some
threshold and distribute dividends to its shareholders whenever its
reserves exceed this threshold. We study how this liquidity
management policy interacts with two types of risk: a Brownian risk
that can be hedged through a financial derivative, and a Poisson
risk that can be insured by an insurance contract. We derive
individual demand functions for hedging and insurance by
corporations. We show that there is a finite price above which both
demand functions are zero. Surprisingly we find that the patterns of
insurance and hedging decisions as a function of liquidity are pole
apart: cash poor firms should hedge but not insure, whereas the
opposite is true for cash rich firms. We also find non monotonic
effects of profitability and leverage. This may explain the mixed
findings of empirical studies on corporate demand for hedging and
insurance: linear specifications are bound to miss the impact of
profitability and leverage on risk management decisions.
| |
Monte Carlo Static
Replication of Barrier Options
Antonio Vulcano
The static hedge for barrier options, initially proposed by
Derman.et.al (1995), is theoretically very appealing, since no
adjustment is required once the replicating portfolio is in place.
However, their procedure is practically flawed in assuming a perfect
knowledge of the future implied volatility surface. As a possible
solution, in this paper we propose a statistical static hedge, also
called Monte Carlo static replication, consisting in generating
risk-neutral dynamics for the volatility surface and in determining
the portfolio minimizing the tracking error, i.e. the difference
between the portfolio value and the barrier option price. This
minimization is accomplished by a least square approach along the
barrier level, where the dependent variable is the barrier option
price, while the independent variable is the replicating portfolio
value. Through our statistical approach, we can use the R²
coefficient in order to have a preliminary measure of the goodness
of the Monte Carlo static hedge; check the convergence of the
statistical replicating portfolio to the barrier option price
obtained by direct simulation of the underlying; construct
confidence intervals around the estimated portfolio weights and
hence around the barrier option price; evaluate the performance of
our methodology when employing realistic dynamics of implied
volatility surfaces.
| |
Nonlinear Term
Structure Dependence: Copula Functions, Empirics, and Risk
Implications Niklas Wagner, Markus Junker, Alex Szimayer
This paper documents nonlinear cross-sectional dependence in the
term structure of U.S. Treasury yields and points out risk
management implications. The analysis is based on a Kalman filter
estimation of a two-factor affine model which specifies the yield
curve dynamics. We then apply a broad class of copula functions for
modeling dependence in factors spanning the yield curve. Our sample
of monthly yields in the 1982 to 2001 period provides evidence of
upper tail dependence in yield innovations; i.e., large positive
interest rate shocks tend to occur under increased dependence. In
contrast, the best fitting copula model coincides with zero lower
tail dependence. This asymmetry has substantial risk management
implications. We give an example in estimating bond portfolio loss
quantiles and report the biases which result from an application of
the normal dependence model.
| |
Adjusting the measure
change function in Lévy markets
Jens Wannenwetsch
This article focuses on the problem of incompleteness in a
financial market where the underlying security is modeled by an
exponential Lévy process. Incompleteness entails non-uniqueness of
the martingale measure, which again means that contingent claims
cannot be priced uniquely unless further constraints are imposed on
the model. Starting from an estimation of the distribution under the
historical probability measure we choose a change of measure which
is able to reproduce the observed volatility, skewness, and kurtosis
of the risk-neutral distribution while at the same time remaining as
close as possible to the historical measure as measured by relative
entropy. Doing this amounts to reformulating the problem of finding
a suitable martingale measure into a finite-dimensional non-linear
minimisation problem with linear constraints. The results are
compared with Black-Scholes prices and Lévy prices obtained by using
the prevalent Esscher change of measure. By construction this method
explains very well different shapes of the volatility smile.
| |
Optimal portfolio
choice with discontinuous price processes and multiple regimes
Andrew Lim, Thaisiri Watewai
This paper concerns the problem of optimal investment and
consumption, with power utility, discontinuous price processes, and
regime switching. Regime switching is modelled by a finite state
Markov chain, and unlike traditional regime switching models,
changes in regime may be accompanied by jumps in the asset price at
the instant of transition, where the distribution of the jump sizes
are conditional on the regime before and after the transition. This
enables us to model a situation where a transition from a `good'
regime to a `bad' one (for instance) is likely to be accompanied by
a downward jump in the price, while transitions from a `bad' regime
to a `good' one is likely to be accompanied by an upward jump.
Expressions for the optimal investment portfolio and consumption
policy are obtained using stochastic control methods. It is shown
that regime switching models with jumps at the instant of transition
have optimal solutions that are significantly different from those
associated with traditional regime switching models where changes in
regime are typically not accompanied by jumps.
| |
An Asset Based Model
of Defaultable Convertible Bonds with Endogenised Recovery
Nick Webber and Ana Bermudez
We describe a two factor valuation model for convertible bonds
when the firm may default. The underlying state variables are the
asset value of the firm and the short riskless interest rate.
Default can occur exogenously, or endogenously at a time a cash
payment is made by the bond. We endogenize the recovery value of a
defaulted bond through assumptions concerning the character of the
reorganization period following default. We use a tailored
Lagrange-Galerkin discretization, coupled with a Lagrange multiplier
method for free boundaries, to value convertibles in the model. Our
framework enables us to specify numerically and financially
consistent boundary conditions and inequality constraints. We
investigate the affect of changing the default, recovery and loss
specification. The affect of introducing a stochastic interest rate
is quantified, and asset and interest rate delta and gammas are
found. We find that the value of the convertible bond is more
sensitive to the initial asset value when its conversion rate is
higher. Its sensitivity to interest rate changes is about one tenth
that of a corresponding defaultable straight bond, chiefly due to
the presence of the conversion feature.
| |
Distribution-Invariant
Dynamic Risk Measures Stefan Weber
The paper provides an axiomatic characterization of dynamic risk
measures for multi-period financial positions. For the special case
of a terminal cash flow, we require that risk depends on its
conditional distribution only. We prove a representation theorem for
dynamic risk measures and investigate their relation to static risk
measures. Two notions of dynamic consistency are proposed. A key
insight of the paper is that dynamic consistency and the notion of
"measure convex sets of probability measures" are intimately
related. Measure convexity can be interpreted using the concept of
compound lotteries. We characterize the class of static risk
measures that represent consistent dynamic risk measures. It turns
out that these are closely connected to shortfall risk. Under weak
additional assumptions, static convex risk measures coincide with
shortfall risk, if compound lotteries of acceptable respectively
rejected positions are again acceptable respectively rejected. This
result implies a characterization of dynamically consistent convex
risk measures.
| |
Liquidation Triggers
and the Valuation of Equity and Debt
Zvi Wiener, Dan Galai, Alon Raviv
Net-worth covenants provide the firm's bondholders with the
right to force reorganization or liquidation if the value of the
firm falls below a certain threshold. In the event of default,
however, many bankruptcy codes stipulate an automatic stay of assets
that prevent bondholders from triggering liquidation. To consider
this impact on the valuation of corporate securities we develop a
model where liquidation is driven by a state variable that
accumulates with time and severity of distress. In addition, current
distress periods may have greater weight than old ones. The
liquidation trigger can be adjustable to a wide array of bankruptcy
codes and jurisdictions.
| |
On the Martingale
Property of Stochastic Exponentials
Bernard Wong and C. C. Heyde
We present a necessary and sufficient condition for a stochastic
exponential to be a true martingale. It is proved that the criteria
for the true martingale property is related to whether an auxiliary
process explodes. An alternative and interesting interpretation of
this result is that the stochastic exponential is a true martingale
if and only if under a `candidate measure' the integrand process
does not explode. Applications of our theorem to problems arising in
mathematical finance are also given
| |
LIBOR Market Model:
from Deterministic to Stochastic Volatility
Lixin Wu and Fan Zhang
LIBOR market model is the benchmark model for interest-rate
derivatives. It has been a challenge to extend the standard LIBOR
market model so as to cope with the volatility smiles and/or skews
that are pronounced in the swaption markets. In this talk we extend
the standard LIBOR market model, which takes forward rate or swap
rate as state variables, by adopting stochastic volatility.
Specifically, we adopt a multiplicative stochastic factor for the
volatility functions of all relevant forward rates. The stochastic
factor follows a squared-root diffusion process, and it can be
correlated with the forward-rate processes. We derive approximate
processes for swap rates after the change to forward swap measures,
and develop a closed-form formula for swaption prices in terms of
Fourier integrals. We then develop a fast Fourier transform
algorithm for the implementation of the formula. The approximations
are well supported by pricing accuracy. By adjusting the correlation
between the forward rates and the volatility in a way consistent
with intuition, we can generate volatility smiles or skews of the
swaption prices similar to those observed in the markets.
Calibration of the model will also be discussed.
| |
FX Instalment Options:
Pricing, Applications, Risk Management
Uwe Wystup and Susanne Griebsch
We compare pricing techniques, present a new closed form
solution and analyze the limiting case. Joint work with Susanne
Griebsch and Christoph Kühn, Goethe University
| |
Minimizing Shortfall
Risk Using Duality Approach Mingxin Xu and Steven Shreve
Option pricing and hedging in a complete market are well-studied
with nice results using martingale theories. However, they remain as
open questions in incomplete markets. In particular, when the
underlying processes involve jumps, there could be infinitely many
martingale measures which give an interval of no-arbitrage prices
instead of a unique one. Consequently, there is no martingale
representation theorem to produce a perfect hedge. The question of
picking a particular price and executing a hedging strategy
according to some reasonable criteria becomes a non-trivial issue
and an interesting question. In this paper, we study the duality
approach in minimizing the shortfall risk proposed by Föllmer and
Leukert (2000). First we extend the duality results in Kramkov and
Schachermayer (1999) to utility functions which are state dependent
and not necessarily strictly concave, as our model requires, and in
the generality of a semimartingale setting. Then we specialize the
duality results to the problem of minimizing shortfall. We next
focus on the mixed diffusion case where we explicitly characterize
the primal and dual sets in terms of the characteristics. We provide
upper bounds for the value function using duality results. Each
upper bound produced in this way corresponds to a dual element. For
lower bounds, we pick a particular strategy which we call the 'bold
strategy' and compute the corresponding value function. In the cases
of bonds and call options and constant parameters, closed form
solutions for the upper and lower bounds are computed and numerical
examples given. This research provides for the first time a method
of checking the quality of a hedging strategy according to the
principle of minimizing shortfall in an incomplete market model.
| |
Multifractal Spectral
Analysis of the 1987 Stock Market Crash
Rossitsa Yalamova, Cornelis A.
The multifractal model of ssset returns captures the fat tails
and volatility persistence of many financial time series. The
multifractal spectrum computed from wavelet transform modulus maxima
lines provides information on the irregularity of the higher moments
of the distribution of market returns, in particular on the kinds of
singularities that occur in a market. We found that changes in the
multifractal spectrum display distinctive patterns around
substantial market "drawdowns" or "crashes." In other wordsm the
kind of singularites and the kinds of irregularity or "randomness"
changes in a distinct fashion in th periods immediately preceding
major market drawdowns. This paper focuses on this identifiable
multifractal spectral patterns preceding the stock market crash of
1987. Although we were not able to find a uniquely iodentifiable
irregularity pattern within the same market preceding different
crashes, we do find the same uniquely identifiable in various stock
markets experiencing the same crash at the same time. Morover, our
results suggest that crashes are preceded by a gradual increase in
the weighted average of the values of the irregularity exponents. At
a crash this weighted average irregularity value drops, while the
dispersion of the spectrum of irregularity coefficients jumps up. We
find that the Wavelet Transform Modulus Maxima (WTMM) methodology
provides a non-sampled complete measure of the changes in the risk
patterns preceding the stock market crash.
| |
Estimation of
Value-at-Risk and Conditional Value-at-Risk for Dynamic Hedging with
Jumps Yuji Yamada and James A. Primbs
In this paper, we present a Value-at-Risk (VaR) and conditional
Value-at-Risk (CVaR) estimation technique for dynamic heding with
jumps in the underlying asset model. At first, we approximate a
jump-diffusion process through its first four moments including
skewness and kurtosis using a general parameterization of
multinomial lattices, and solve the mean square optimal hedging
problem. Then our recently developed technique is applied to extract
the hedging loss distributions in option hedge positions. Finally,
we investigate how the hedging error distribution changes with
respect to non-zero kurtosis and skewness in the underlying through
numerical experiments, and examine the relation between VaR and CVaR
of the hedging loss distributions and kurtosis of the underlying.
| |
Asset Allocation with
Regime-Switching: Discrete-Time Case
Hailiang Yang and Ka Chun Cheung
In this paper, we study the optimal asset allocation problem
under a discrete regime switching model. Under the short-selling and
leveraging constraints, the existence and uniqueness of the optimal
trading strategy are obtained. We also obtain some natural
properties of the optimal strategy. In particular, we show that if
there exists a stochastic dominance order relationship between the
random returns at different regimes, then we can order the optimal
proportions we should invest in such regimes.
| |
Modeling Credit
Risk Yildiray Yildirim
Modeling Credit Risk This project provides an alternative
approach to the structural credit risk models. The
first-passage-time approach extends the original Merton model by
accounting the default may occur not only at the debt's maturity,
but also prior to this date. Default happens when the underlying
process hits a barrier. We also define default the first time the
underlying process hits a barrier as in the first passage time
models, and the liquidation time as the area of this process under
the barrier is greater then a constant barrier. We use this
technique to price risky debt, and separate default from insolvency.
| |
Completeness of
Security Markets and Backward Stochastic Differential Equations with
Unbounded Coefficients Jiongmin Yong
For a standard Black-Scholes type security market, completeness
is equivalent to the solvability of a linear backward stochastic
differential equation (BSDE, for short). If the interest rate is
bounded, there exists a bounded risk premium process, and the
volatility matrix has certain surjectivity, then the BSDE will be
solvable and the market will be complete. However, if the risk
premium process and/or the interest rate is not bounded, one gets a
BSDE with unbounded coefficients to solve. In this paper, we will
discuss such a situation and will present some solvability results
for the BSDE which will lead to the completeness of the market in a
broad sense.
| |
American Option
Pricing with Transaction Costs
Valeri Zakamouline
In this paper we examine the problem of finding investors'
reservation option prices and corresponding early exercise policies
of American-style options in the market with proportional
transaction costs using the utility based approach proposed by Davis
and Zariphopoulou (1995). We present a model, where investors have a
CARA utility, and derive some properties of reservation option
prices. We discuss the numerical algorithm and propose a new
formulation of the problem in terms of quasi-variational HJB
inequalities. Based on our formulation, we suggest original
discretization schemes for computing reservation prices of
American-style option. The discretization schemes are then
implemented for computing prices of American put and call options.
We examine the effects on the reservation option prices and the
corresponding early exercise policies of varying the investor's ARA
and the level of transaction costs. We find that in the market with
transaction costs the holder of an American-style option exercises
this option earlier as compared to the case with no transaction
costs. This phenomenon concerns both put and call options written on
a non-dividend paying stock. The higher level the transaction costs
is, or the higher risk avers the option holder is, the earlier an
American option is exercised.
| |
Pricing a class of
exotic options via moments and SDP relaxations
Mihail Zervos, J.B.Lasserre, T.Prieto
We present a new methodology for the numerical pricing of a
class of exotic derivatives such as Asian or barrier options when
the underlying asset price dynamics are modelled by a geometric
Brownian motion or a number of mean-reverting processes of interest.
This methodology identifies derivative prices with
infinite-dimensional linear programming problems involving the
moments of appropriate measures, and then develops suitable
finite-dimensional relaxations that take the form of semi-definite
programs indexed by the number of moments involved. By maximising or
minimising appropriate criteria, monotone sequences of both upper
and lower bounds are obtained. Numerical investigation shows that
very good results are obtained with only a small number of moments.
Theoretical convergence results are also established.
| |
A Tale of Two Time
Scales: Determining Integrated Volatility with Noisy High-Frequency
Data Lan Zhang, Per A. Mykland, Yacine Ait-Sahalia
It is a common practice in finance to estimate volatility from
the sum of frequently-sampled squared returns. However market
microstructure poses challenges to this estimation approach, as
evidenced by recent empirical studies in finance. This work attempts
to lay out theoretical grounds that reconcile continuous-time
modeling and discrete-time samples. The present study incorporates
microstructure noise in the modeling of the returns. We propose a
subsampling-averaging approach which takes advantage of the rich
sources in tick-by-tick data while delivering a consistent,
asymptotically normal estimator for the integrated volatility. Under
our framework, it becomes clear why and where the "usual" volatility
estimator fails when the returns are sampled at the highest
frequency. Our procedure is implementable regardless of the
magnitude of the noise. In the event that the noise is "almost
negligible ", our work provides an approach to finding an optimal
sampling frequency if one wishes to use the classical "realized
volatility", and to optimizing the subsample frequencies if one
wishes to incorporate more data into the analysis.
| |
Mean--Risk Portfolio
Selection Models in Continuous Time Xun
Yu Zhou and George Yin
This paper is concerned with continuous-time portfolio selection
models where the objective is to minimize the risk subject to a
prescribed expected payoff at the terminal time. The risk is
measured by the expectation of a certain function of the deviation
of the terminal payoff from its mean. First of all, a model where
the risk has different weights on the upside and downside variance
is solved explicitly. The limit of this weighted mean--variance
problem, as the weight on the upside variance goes to zero, is the
mean--semivariance model which is shown to admit no optimal
solution. This negative result is further generalized to a
mean--downside-risk portfolio selection problem where the risk has
non-zero value only when the terminal payoff is lower than its mean.
Finally, a general model is investigated where the risk function is
convex. Sufficient and necessary conditions for the existence of
optimal portfolios are given. Moreover, optimal portfolios are
obtained when they do exist.
| |
Arbitrage Pricing
Simplified William Ziemba and M. Kallio
This paper derives fundamental arbitrage pricing results in
finite dimensions in a simple unified framework using Tucker's
theorem of the alternative. Frictionless results plus those with
dividends, periodic interest payments, transaction costs, different
interest rates for lending and borrowing, shorting costs and
constrained short selling are presented. While the results are
mostly known and appear in various places, our contribution is to
present them in a coherent and comprehensive fashion with very
simple proofs. The analysis yields a simple procedure to prove new
results and some are presented for cases with frictions.
| |
Utility Maximization
with a Stochastic Clock and an Unbounded
Gordan Zitkovic
We introduce a linear space of finitely additive measures to
treat the problem of optimal expected utility from consumption under
a stochastic clock and an unbounded random endowment process. In
this way we establish existence and uniqueness for a large class of
utility maximization problems including the classical ones of
terminal wealth or consumption, as well as the problems depending on
a random time-horizon or multiple consumption instances. As an
example we treat explicitly the problem of maximizing the
logarithmic utility of a consumption stream, where the local time of
an Ornstein-Uhlenbeck process acts as a stochastic clock.
| |
|