Estimating the Out‐of‐Sample Predictive Ability of Trading Rules: A Robust Bootstrap Approach

Abstract

In this paper, we provide a novel way to estimate the out-of-sample predictive ability of a trading rule. Usually, this ability is estimated using a sample splitting scheme, true out-of-sample data being rarely available. We argue that this method makes a poor use of the available data and creates data mining possibilities. Instead, we introduce an alternative .632 bootstrap approach. This method enables to build insample and out-of-sample bootstrap datasets that do not overlap but exhibit the same time dependencies. We show in a simulation study that this technique drastically reduces the mean squared error of the estimated predictive ability. We illustrate our methodology on IBM, MSFT and DJIA stock prices, where we compare 11 trading rules specifications. For the considered datasets, two different filter rule specifications have the highest out-of-sample mean excess returns. However, all tested rules cannot beat a simple buy-and-hold strategy when trading at a daily frequency. JEL: C52, C53, C15, C18

Automated investment management: Comparing the design and performance of international robo‐managers

“…The advantages of robo-managers in deviating market situations (what-if scenarios; see Kleijnen, 2012) cannot be depicted with such a methodology (Pfiffelmann et al, 2016). Hypothetical market scenarios based on historical price patterns are needed to overcome this shortcoming (Hambuckers & Heuchenne, 2016). The simulation and forecasting methods used for this purpose examine how hypothetical path scenarios can be derived from past return series (Mansini et al, 2015).…”

Interest Rate Differentials and the Dynamic Asymmetry of Exchange Rates

“…,Bajgrowicz and Scaillet [2012],Hambuckers and Heuchenne [2016] orHsu et al [2016].The significance of in-sample performance is assessed with the stepwise-superior predictive ability (SSPA) test ofHsu et al [2010] to control for data snooping issues, whereas out-of-sample performances are compared with the test ofGiacomini and White [2006] and the fluctuation test of…”