Lookback Options A Discrete Problem（411-426）.pdfVIP

17 Lookback Options: A Discrete Problem* Key words: PDE, Monte Carlo simulation, transform method, discrete monitoring, exotic option The number of exotic options traded on the market has dramatically increased in the last decade. Correspondingly, a large demand has come about for the development of new, efficient, and fast methods for pricing these securities. This case presents numerical and analytical methods for pricing discretely moni- tored lookback options in the Black–Scholes framework. Lookback options are path- dependent options. Their settlement is based on the minimum or the maximum value of an underlying index as registered during the lifetime of the option. At maturity, the holder can “lookback” and select the most favorable figure of the underlying as occurred during this period. Since this scheme guarantees the best possible result for the option holder, he will never regret the option payoff. As a consequence, a look- back option is more expensive than a vanilla option with similar payoff function. An important feature of this contract is the frequency of observation of the underly- ing assets for the purpose of identifying the best possible value for the holder. Dis- crete monitoring refers to updating the maximum/minimum price at fixed times (e.g., weekly or monthly). In general, a higher maximum/lower minimum occurs as long as the number n of monitoring dates increases. As noted by Heynen and Kat (1995), the discrepancy between option prices under continuous and discrete monitoring can largely be due to the slow convergence of the discrete scheme to the continuous one as the number n of monitoring dates increases. This figure is quantified in an order of proportionality of 1/√n. Closed-form solutions for continuous sampled lookback option prices have been obtained by Conze and Viswanathan (1991) and Goldman, Sosin and Gatto (1979). However, few papers have investigated the analytical pricing of discretely monitored