Hossein Kakavand
Finance Program, Stanford University

Home | Research | Publications | Links | Photos

Research Projects

  • PROJECTION: Credit Derivatives Pricing
    Abstract: From Bottom-Up (03) to Top-Down (04) portfolio credit derivative pricing via Markovian intensity projection. To value portfolio credit derivatives such as options on portfolios, portfolio indecies and tranches one needs to compute the distribution of the portfolio default count over time. For the factor model, proposed by Duffie (01), this can be done by conditioning and integrating against the distribution of the common risk factor. This is executed in Eckner (07). However, this procedure is already computationally quite demanding for only one common factor and bocomes computationally inhibitive for more than one common factor and for intensities with feedback. We use a Markovian projection technique to compute the distribution of the portfolio default count over time. Specifically, we construct a time-inhomogeneous Markov process that shares the count distribution of the default count over time. The computation of the state (count) distribution of this Markov chian are considerablly less demanding and as such result in large computational savings.

  • SECTORS: Effects of Sectors on Risk Premia in Structured Credit Derivatives
    Abstract: The work by Eckner (07) extensively studies the spread components of structured credit products under the factor model introduced by Duffie (01). It shows that such securities compensate investors for expected losses due to defaults, pure jump-to-default risk, correlation risk, as well as the risk of firm-specific and market-wide adverse changes in credit conditions. It provides a framework that allows a decomposition of ”structured” credit spreads and applies this decomposition to CDX index tranche. Further it presents computational techniques that make a certain class of fully dynamic intensity-based models for portfolio credit risk, just as computationally tractable as the static copula model. However that work does not consider sector factors due to the resulting computational burden. In this work work we propose a new model to study the effects of effects of sector factors on explaining risk premia in structured credit derivatives, while increasing the computational requirements by a small factor.

  • MADNESS OF PEOPLE: Effects of Herding in Financial Markets on Asset Price
    Abstract: We study the boundedly rational behavior of agents in financial markets in which agents tend to imitate the investment preferences of other agents. This phenomenon is known as herding, which contradictes the assumptions of classical asset pricing theory and affects the asset prices in financial markets. Specifically, we study a model of a market populated by two groups of boundedly rational agents. One group observes some private information, but fails to extract the information of other agents from prices. Since information diffuses gradually across the population, prices underreact in the short run. The underreaction means that agents trading on momentum can profit by following the trend. However, if they can only implement simple strategies, their attempts at arbitrage must inevitably lead to overreaction at long horizons. We study a model of underreaction and overreactions in financial markets as a result of agents boundedly rational behavior.

  • INFORMED TRADING: Optimal Trading Stratgy of an Informed Trader in the Presence of an Arbitrageur
    Abstract: In the context of a financial market, informed traders are those that possess intelligent information that is not generally available to the public. This information can result in a trading strategy that would intuitively provide higher returns than that available to the public. Informed trading under Various conditions has been studied previously. These studies include the case of a single risk-neutral insider that aims to maximize profit in the presence of noise traders. Also, the optimal trading strategy with more than one informed trader with correlated information (imperfect competition) among them has been studied. In this work, we consider a trader who posseses information about the price of an asset at a specific horizon and aims to maximize the profit from this information. This is done in the presence of an arbitrageur who hopes to profit from the trader’s activity. The arbitrageur is aware of the horizon but uncertain about the trader’s information and learns from observed market activity. This is a dynamic game with asymmetric information. We present an algorithm for computing perfect Bayesian equilibrium behavior and conduct numerical experiments. Our results demonstrate that the trader’s strategy differs in important ways from one that would be optimal in the absence of an arbitrageur. In particular, the trader’s actions depend on and influence the arbitrageur’s beliefs. Accounting for the presence of a strategic adversary can greatly reduce transaction costs.