Seminar: Winter 2006

Yacine Ait-Sahalia (Bendheim Center for Finance, Princeton University)

We analyze the portfolio selection problem of an investor facing both Brownian and jump risks. By decomposing the two types of risks on a well-chosen basis, we provide a new methodology for determining the optimal solution in closed form, up to a constant. We show that the optimal solution is for the investor to focus on controlling his exposure to the jump risk, while exploiting differences in the asset returns diffusive characteristics in the orthogonal space. We then examine the solution to the portfolio problem as the number of assets available to the investor increases, and study the asymptotic distribution of the investor's wealth and optimal portfolio.

(joint with Julio Cacho-Diaz and Tom Hurd)
top of page

Maher Aoun and Alyce Campbell (Calypso Technology, Inc.)

The talk gives an overview of capital markets technology and sheds light on the mathematical models' lifecycle from the quant's computer to the trader's desk.
top of page

Terry Tse (Deutsche Bank, New York)

In this presentation we will review some of the most important events that drove credit markets in 2005. In particular, we will investigate the crisis in the US auto companies and their subsequent impact on the credit default swap market. We will discuss the mechanics and peculiar dynamics of the credit default swaps contracts in the context of these events. Live trading examples will be presented. Finally, we will also touch on the impact of macro-economic issues such as the yield curve and oil prices.
top of page

Jayaram Muthuswamy (Dept of Finance, Griffith University, Australia)

Weak-form tests of market efficiency using portfolio returns have often invoked an "expected returns are constant" equilibrium model because insignificant sample autocorrelatinos then legitimately imply weak-form market efficiency. In this paper, the problem of price non-synchronicity is re-examined in the context of its strong potential to come directly as a wedge between theoretical efficiency, and its real
world testability. The basic argument here is that full information reflectively of prices is paradoxically impossible to achieve if the asset has not had the opportunity to trade at the instant of measurement of its price - and to therefore fully reflect available information in the first instance. Using a simple model to represent price non-synchronicity in a portfolio, it is shown that the conventionally perceived upper bound for spurious first order autocorrelation under the null hypothesis of a random walk for individual asset prices, is far too conservative. In addition, time variation in expected returns further exacerbates the successful
execution of tests of the Efficient Markets Hypothesis: if expected returns change through time, sample autocorrelations can deviate significantly from zero and the markets still be pronounced efficient.
top of page

Alessia Falsarone (Citigroup Investment Research)

We introduce a traditional top-down vs. bottoms-up approach to modern relative value investing with respect to global equity instruments (both straight and equity-linked). In particular we overlay fundamental, multi-factor driven calls to pure volatility pair trades under a synchronized pricing scenario. Next, we discuss the impact of phase-locking events, modern prospect theory, and “half-life” of a strategy in portfolio and individual equity selection models. Most importantly, we analyze the key determinants of effective modeling for equity strategists vis-à-vis externalities introduced by a changeover in financial metrics, comparability parameters and other disruptions of near-term factor stability.
top of page

David Li (Credit Derivatives Group, Barclays Capital, New York)

This seminar will be part of the Stanford-Tsukuba workshop on Thursday and
Friday March 2-3.
top of page

David Lando (Copenhagen Business School)

top of page

Takaki Hayashi (Columbia University)

We consider the problem of estimating the (integrated) covariance/correlation of two diffusion price processes when they are observed only at discrete times in a nonsynchronous manner. The popular approach in the literature, the realized covariance/correlation estimator, which is based on (regularly spaced) synchronous data, is problematic because the choice of regular interval size and data interpolation scheme may lead to unreliable estimation. We first introduce our approach (proposed in Hayashi and Yoshida 2005) -- an alternative estimation procedure which is free of any `synchronization' processing of the original data. We show why and how the procedure works well. We then present advances of the theory, accompanied by some examples.

(Joint work with N. Yoshida.)
top of page

Yakov Kanter (Morgan Stanley)

After a brief introduction to MBS, I will talk about the problems facing MBS modelers on the Street today.

top of page