Monthly Archives: April 2016

Interesting Takeaways from “ Fact, Fiction and Momentum Investing” 2014, AQR

  • There is an important distinction between momentum and trending following. The former is a relative strategy (based on recent out or under performance) while the latter an absolute one (based on recent absolute performance).
  • Every factor strategy has its down time. So is momentum. Momentum’s worst time occurs after a 2-year plus bear market followed by abrupt market upswing (e.g. 1Q 2009). But like other factor strategy, it works across VERY long periods of time and in many places. Combining value and momentum historically gave you positive annual return in 4 out of 5 times. Momentum adds diversification benefit to a value strategy (i.e. worth having momentum strategy even if its expected return is zero)
  • Momentum is as strong in large stocks as in small stocks. There is indeed high turnover in momentum strategy (5-6 times of value). Hence it is much more appropriate to be exploited by a large institutional investor rather than by retail as the former enjoys 10 time less trading cost. However, momentum’s return does not seem to decline despite increased adaptation among hedge funds. Momentum’s premium should persist as long as behaviour bias (underreaction or delayed overreaction) or compensation for risk still exists.
  • There are many ways to construct momentum signals, e.g. price momentum or fundamentals (e.g. earnings) momentum. Taking an average of all reasonable measures tend to yield better portfolios.

Leave a comment

Filed under Uncategorized

Interesting takeaways from “Fact, Fiction and Value Investing”2015, AQR

  • Value investing can be expressed in systematic investing or discretionary concentrated investing. What is more important is how to properly construct the value factor.
  • Value factor exists in multiple asset classes. And the value strategies have high correlation across different asset classes. Reversal of long-term (5-year) past return can be used as a value measure for a certain asset class (e.g. commodities) where it is not obvious how to build a value measure.
  • Value factor seems to be weak in large-cap stocks. It indicates that value may be more relevant in small stocks or EM stocks or frontier market stocks.
  • Both Profitability/Quality and Momentum factor have positive return, even though they are negative correlated with value factor. That means a combination of profitability/quality, momentum and value seems a better approach – as they are thought like a system – than using one factor alone. The combination of factors should generate better Sharpe ratio.
  • Why Value works can be attributed to both risk premium (efficient market) theory and behaviour anomaly. The fact that Quality has positive return means behaviour anomaly is as likely as the efficient market explanation.
  • There appears some issue in the value measure construction in Fama French 5-factor model that ignores momentum. Value still exhibits positive risk premium when its construction is improved and momentum is included.

Leave a comment

Filed under Uncategorized