European Floating Strike Lookback Options: Alpha Prediction and Generation Using Unsupervised Learning

Abstract

This research utilized the intrinsic quality of European floating strike lookback call options, alongside selected return and volatility parameters, in a K-means clustering environment, to recommend an alpha generative trading strategy. The result is an elegant easy-to-use alpha strategy based on the option mechanisms which identifies investment assets with high degree of significance. In an upward trending market, the research had identified European floating strike lookback call option as an evaluative criterion and investable asset, which would both allow investors to predict and profit from alpha opportunities. The findings will be useful for (i) buy-side investors seeking alpha generation and/or hedging underlying assets, (ii) sell-side product manufacturers looking to structure the European floating strike lookback call options, and (iii) market trading platforms looking to introduce new products and enhance liquidity of the product.

TOPICS: Options, volatility measures, statistical methods, simulations, machine learning

Key Findings

• • Outperformance of mean Sharpe of forward European floating strike lookback call options selected using K-means clustering was statistically significant using ANOVA test at p value = 0.0002. Tukey-Kraner HSD test showed that Sharpe risk adjusted return improved by 4.94% to 7.04%.

• • Evaluating Sharpe based on European floating strike lookback call option provided a clear choice of trading cluster of options, whereas this was not clearly apparent if the standard call option Sharpe was used as the only evaluating criteria for the selection of tradable option clusters.

• • Consistency and stability of the predictive results implied that the European floating strike lookback call options acted as a useful evaluation criterion for option investment.