Market Timing with the VKOSPI Refined Composite Multiscale Entropy Indicator

For the KOSPI200 Index (“Index”), a well-known negative and statistically significant relationship exists between the returns of the Index and VKOSPI, the official model-free implied volatility index. Our primary research purpose is to test VKOSPI as an informative and meaningful trading indicator and to generate some timely trading signals. VKOSPI Refined Composite Multiscale Entropy (“VKOSPI RCMSE” or “RCMSE”) as an alternative complexity indicator was derived from the VKOSPI time series. Both long and short positions triggered by large changes in the RCMSE and the trading strategies are evaluated in terms of their ability to generate alpha through an algorithmic trading platform that will outperform the index benchmark in steady risk aversion regime. There is some empirical evidence for expecting opportunities for positive future excess and risk-adjusted returns for long and short positions triggered by large movements in the RCMSE. This finding is more prominent during the periods of steady risk aversions. Trading strategy is tested on the data period from Jan 2, 2003 to May 29, 2015. The time-varying return-volatility relation is used to implement a set of dynamic asset allocation strategies by analyzing the RCMSE with its corresponding Index. With asymmetric volatility phenomenon in the Korean stock market, selected benchmark-outperforming dynamic asset allocation strategies are developed by processing meaningful signals from RCMSE indicator on the Index Futures.