A LOOK AHEAD AT OPTIONS PRICING AND VOLATILITY

Academics on sabbatical leave are usually ‘taking stock’ of their previous research, thinking about writing a book or starting new projects. I have chosen a rather indirect path to enlightment: to spend my sabbatical in a hedge fund trading options and forecasting volatility. Hopefully, this request of the Editor-in-Chief of Quantitative Finance for writing a ‘general article on volatility’ will be an opportunity to reflect a bit on the interrelation of option pricing theory and quantitative trading strategies as I see them today. First, the usual disclaimer: a state-of-the-art survey paper on volatility trading is impossible to write, at least for me. There is simply too much material. For this reason, I will limit this paper to a few topics that I am familiar with, namely where pricing theory and trading interact. We are at a turning point in pricing theory. On the theoretical side of things, the input of mathematicians and physicists on the field has been tremendous. Confluent hypergeometric functions, integro-differential equations, multi-factor models — just to mention a few examples — have been applied to pricing options and many papers have been published in quantitative finance journals such as this one. If a researcher even thinks about a probability distribution as an alternative to the Black–Scholes model, chances are that this distribution has already been tested (usually by a graduate student or a Wall Street quant), a paper has been written and it has been put to work somewhere. So, is option-pricing theory dead? I believe that the answer is ‘no’, but that we need to broaden our horizons. First point: until now, option-pricing theory has worked almost exclusively on improving the ‘pricing kernel’. It has not considered other very important elements that go into the price of an option. I am referring to the effects of news about the underlying asset that will be released during the lifetime of the option at known dates. To generate profitable trades, option traders have to reconcile stylized facts about the movement of the underlying asset and the history of implied volatilities with the current price of an option and determine whether it is fairly valued. This judgment will necessarily take into account the fact that there are earnings announcements, releases of macro-economic statistics, ex-dividend dates, takeover proceedings and court judgments, the results of clinical trials (in the case of biotechnology and pharmaceutical stocks) or other idiosyncratic news that will be released over the life of the option. Option pricing theory has proposed very little in terms of handling these very real issues. For theorists, this is a controversial point: the efficient markets dogma assures us that all news is ‘priced in’ so we need not worry. But is this really the case? I don’t think so. The impact of a theory improved in this direction would be great. Second point: very recent structural developments in options markets may lead to exciting new trading techniques and improved the way in which we view and price listed options in aggregate. This major development is the advent of electronic trading of equity options as a primary way of trading in the U:S: markets (‘Direct Market Access’). To some extent, parallel developments are taking place in Europe. Electronic access allows traders to view the market as composed of multiple exchanges — each with its bid-ask spread and depth — and to trade options on several underlying assets across different markets systematically. This leads naturally to portfolio trading of options. An option pricing theory that can handle in a practical way multiple underlying stocks has yet to emerge.