Revisiting the Profitability of Market Timing with Moving Averages

• DOI: http://doi.org/10.1111/irfi.12132
• Published date: 01 June 2018
• Date: 01 June 2018

Revisiting the Profitability of

Market Timing with Moving

Averages*

VALERIY ZAKAMULIN

School of Business and Law, University of Agder, Kristiansand, Norway

ABSTRACT

In a recent empirical study by Glabadanidis (“Market Timing with Moving

Averages”(2015), International Review of Finance 15(13):387–425), the author

reports striking evidence of extraordinarily good performance of the moving

average trading strategy. In this paper, we demonstrate that this “too good to

be true”reported performance of the moving average strategy is due to

simulating trading with look‐ahead bias. We perform simulations without

look‐ahead bias and report the true performance of the moving average

strategy.We find that, at best, the performance of the moving average strategy

is only marginally better than that of the corresponding buy‐and‐hold strategy.

In statistical terms, the performance of the moving average strategy is

indistinguishable from the performance of the buy‐and‐hold strategy.

JEL Codes: G11; G17

I. INTRODUCTION

To time the market, traders often employ moving averages (MAs) of prices (the

10‐month simple MA is particularly popular). A common belief is that one can

identify the market trend by comparing the MA of prices and the recent price.

Specifically, a price above (below) the MA signals a bullish (bearish) trend.

However, whereas market timing with MAs has been extensively employed

by practitioners for more than half a century, academics had long been

skeptical of its usefulness. The study by Brock et al. (1992) marked the onset

of a change in the academics’attitudes toward technical analysis because this

study tends to suggest that one should not dismiss the value of market timing.

However, before the 2000s, on the academic side, there was consensus that

market timing does not work in stock markets. Whereas, on the one hand,

researches did find evidence that MA signals display statistically significant

predictive ability (Gencay 1996; Gencay and Stengos 1997; Gencay 1998;

Gencay and Stengos 1998); on the other hand, Sullivan et al. (1999)

* The author is grateful to an anonymous referee and Steen Koekebakker for their comments and to

Geirmund Glendrange and Sondre Tveiten for double checking the author’s results. The usual

disclaimer applies.

© 2017 International Review of Finance Ltd. 2017

International Review of Finance, 2017

DOI: 10.1111/irfi.12132

International Review of Finance, 18:2, 2018: pp. 317–327

DOI:10.1111/irfi .12132

© 2017 International Review of Finance Ltd. 2017