Events 2006-2007

Professor David G. Luenberger

Stanford University and LSE

Dynamic Pricing of Soft Derivatives

A soft derivative is an asset whose payoff is a function of a dynamic variable that is not a marketed security. An example is an option on the profit a company will make during the year. Profit is not a traded security. In such cases, standard derivative theory does not apply because it is impossible to replicate the asset with marketed securities. Most "real options" associated with business operations are soft derivatives. This talk will present a method for pricing such derivatives. This price will not render the derivative redundant, as does the Black-Scholes price, but it renders the derivative irrelevant in the sense that no risk-averse individual will want to include it in his or her portfolio. The price is found as the solution of a partial differential equation that generalizes the Black-Scholes equation.

Dr Robert Barnes

Managing Director, Equities, UBS Investment Bank

Evolving Execution Landscapes: Liquidity, Fragmentation, Dark Pools

Today's economic environment reflects increasingly dynamic evolutions of market micro-structure which ultimately shape order execution strategies. Come and learn about recent commercial trends and regulatory imperatives that impact markets, including MiFID which goes 'live' on 1 November. Expect more competition, complexity, and change. MiFID sets the regulatory framework encouraging competitive new entry and entrepreneurial opportunity. The industry already is responding.

Dr Michael Simmonds

Director, Equity Quantitative Analytics, Lehman Brothers

The use of quantitative models in execution analytics and algorithmic trading

With the dramatic rise of electronic trading, the traditional cash markets have increasingly been looking towards quantitative analytical tools to both improve trading quality and measure performance. As the regulatory climate enforces a clearer demonstration of best execution these two areas are becoming even more important. I will discuss how we develop algorithmic trading tools for the trading floor and our clients. I will also discuss the modelling of market microstructure and how we integrate these modelling tools into the whole pre- intra- and post-trade process. In particular I will show how we model transaction costs and how this can be integrated into automated trading strategies.

Dr Jerry T Parwada

School of Banking and Finance, University of New South Wales, Australia

Do foreign investors pay more for stocks in the United States? An analysis by country of origin

We examine whether and why implicit equity transaction costs incurred by institutional investors across 51 countries differ from those of domestic institutions in the United States. Using a proprietary institutional trading dataset that discloses the home country of the initiator of a trade, we analyze round-trip implicit transaction costs conditional on the trader's country of origin. The average daily equally (trade) weighted disadvantage of foreign institutional investors for purchases is 3.3 (3.3) basis points (bps) and 4.1 (3.0) bps for sales of common stocks and American Depository Receipts that traded on the NYSE, AMEX and NASDAQ between 1 July 1999 and 30 September 2004. Therefore a roundtrip daily equally weighted disadvantage to foreign institutional investors is approximately 7.4 (6.3) bps. Several institutional background factors related to foreign investors’ countries of origin are relevant in explaining their disadvantage. (Joint work with Terry S. Walter & Donald W. Winchester).

Professor Sheri Markose

Economics Department, University of Essex
Director, Center for Computational Finance and Economic Agents, University of Essex

Dynamic Learning, Herding and Guru Effects in Networks

It has been widely accepted that herding is the consequence of mimetic responses by agents interacting locally on a communication network. In extant models, this communication network linking agents, by and large, has been assumed to be fixed. In this paper agents make binary decisions, to buy or sell a unit of an asset in a radically decoupled market environment where own payoffs are observed but there can be no computable winning strategies. Typically, agents learn to win by having 'right' connections or advisors. Agents adaptively modify the weights of their links to their neighbours by reinforcing 'good' advisors and breaking away from 'bad' advisors with the latter being replaced randomly from the remaining agents. The resulting network not only allows for herding of agents, but crucially exhibits realistic properties of socio-economic networks that are otherwise difficult to replicate: high clustering, short average path length and a small number of highly connected agents, called "gurus". These properties are now well understood to characterize 'small world networks' of Watts and Strogatz (1998). Results will be demonstrated on the CCFEA 'herding simulator'.

Professor Jagjit Chadha

Head of Quantitative Economics, BNP Paribas, London

Herding in Financial Markets: Some Evidence from Economists' Surveys

We examine intra-month forecasts made by market economists for the main US indicators and find a number of stylised facts consistent with herding in beliefs. For all indicators we examine, we find that the empirical density function is significantly narrower than the actual data. We also find that there is some evidence of inefficiency, insofar as not all information seems incorporated in the market median. We undertake a specific examination of the economists' forecast of non-farm payrolls and find that as we approach the announcement date, there is no tendency for the forecasts to improve and we also find some evidence of leadership. At face value, this means that the forecasts from the set of market economists are more tightly dispersed than the actual market data and do not reflect the true uncertainty about outcomes. In effect, asa groups, economists' subjective priors are less dispersed than the data they expect - which is a form of risk loving behaviour. One possible explanation is that they herd around an initial forecast, which is public information and this swamps or outweighs any incentive to collect/use price information as the month goes on. An interesting corollary is that individual herding leads to aggregate risk loving. The role this kind of herding plays in generating financial market volatility is also considered.

Dr Alexander Lipton

Managing Director, Merrill Lynch

Dynamic factor models for credit correlation

We present a dynamic factor model for large inhomogeneous baskets of credits and show how to calibrate it to the market and use for pricing and risk management of bespoke tranches. We also discuss how a static version of the model can be used as a very attractive alternative to the base correlation framework for standard credit indices.

Professor Michael Dempster

Director, Centre for Financial Research, University of Cambridge
Managing Director, Cambridge Systems Associates Limited

Empirical copulas for CDO pricing using minimum cross entropy

The principle of minimum cross entropy is defined and used to choose an empirical copula calibrated to market data which is closest to industry standard Gaussian or Student t copulas. Computationally intensive numerical procedures are described and remarkably good fits to iTraxx and CDX index tranche prices demonstrated. This describes work in progress and outstanding research questions posed concern techniques for understanding the temporal evolution of copulas for multiple maturity CDO pricing.