Alternative Weighting in Indices

[JanSvenda]

Hey,

 

just want to see what others think about equal weight indices and their performance vs. market cap weight.

 

I was intrigued by fundamental indexation roughly two years ago, because Research Affiliates were boasting with the performance of RAFI (the fundamental index built by Arnott et al. - core paper can be found here - https://www.researchaffiliates.com/documents/FAJ_Mar_Apr_2005_Fundamental_Indexation.pdf). If you actually look at the performance of the index, the sole reason for the long-term outperformance (roughly 1.5%) is its performance in 2009. RAFI outperformed market cap SP500 by 16%. If RAFI earned SP500 return in 2009, it would clock in roughly the same long-term return since inception. This is obviously not mentioned by the marketing staff of RAFI. In fact even Arnott in 2005 did not hint the possibility of this, but that is slightly beside the point.

 

I ended up co-writing a short paper on the topic - https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2911929 where we said that the outperformance is likely due to the rebalancing of the index (timing might play a role as well). In that research, we also looked at another index that had a 'long-enough' track record and that is Equal Weight SP500, RSP by Guggenheim which was started three years before RAFI in 2003. This index performed even better than RAFI in 2009 (which points out that RAFI is not unique) likely due to the same effect.

 

What the rebalancing actually does is that it 'automates' the 'buy low' part of the old adage. By selling stocks that are relatively up (i.e. are a larger part of the index) and buying stocks that are down the index can be an interesting alternative to your usual SPY. The obvious downside is that in a bull market the indices are prone to underperformance, however, the outperformance in the crisis years could be large enough to create a meaningful difference.

 

I wonder if the indices are going to behave similarly when the next crisis (and potential rebound) comes or whether this was a one-off scenario. A small-cap equal weight might be an even better way to take advantage of this. I also wonder if anyone is a holder of RSP or RAFI here. Don't worry I am not going to hit you with a copy of Security Analysis.