Style Rotation: This is the notion that there exist predictability in the relative performance of "box style" portfolios, i.e. those portfolios characterized by the factors of size (small cap, mid cap, large cap) and growth vs. value (value, blend, growth).
Style boxes were made popular by the Morningstar mutual fund ranking service, and now they are used universally. There are many ETFs that mimic these portfolios, so strategies that "time" relative performance of style are now easy (and relatively cheap) to implement.
To be sure, predictability of relative returns for style portfolios is at odds with the EMH, but perhaps not terribly so. In any event, many authors have found positive evidence for style rotation.
We'll look at this evidence and see what we believe and what we do not.
A good starting place is the article
The model developed in this article is less parsimonious than those we feel best about. Nevertheless it puts interesting issues into play. More amusing to me is that it also finds non-trivial "cheese."
The article has a nice literature review and bibliography. Essentially all of the papers you see listed there are accessible on-line from the UPenn library.
Mellon Bank View of Style Rotation
First, a quote from a Mellon Bank press release: "In deciding whether to favor growth or value stocks, the Mellon Equity model analyzes interest rate spreads, the term structure of interest rates, and several valuation and profitability indicators for baskets of stocks of both investment styles. We are now able to identify the combinations of variables that are favorable or unfavorable to growth and value stocks."
If this is true --- and if "favorable" means what everyone is expected to believe it means, we can replicate the Mellon analysis, and --- incidentally --- pick up a nice piece of cheese, if they are right. Telling whether growth or value will out-perform is a very nice thing indeed.
Sadly, the Mellon quote is from February 2004 and it called for Growth to out-perform. Well, we know that Value prevailed in 2004, 2005, and 2006. Oops!
Still, the Mellon research group does have some interesting reports. In particular, we'll consider their graphical methodology for judging tactical allocation strategies, especially for size rotation.
Arshanapalli et. al. use Grainger Causality as their main tool for variable selection, and this may seem strange to people who are more familiar with the model selection tools of regression. This is a useful reminder that time series variable selection is difference in some ways from regression variable selection. Along the way we will try to identify best practice for time series variable selection (for realistic financial contexts).
Note: The pain vanilla Wiki is not too good on this but the Wiki-Scholar article on Granger causality is fine. In fact, it has an elegant explanation by Sir Clive himself. Still, however well acknowledged, the idea of G-causality is pretty natural; it just happened to be a home run.
Related Web Resources.
There is a web site that gives up to the hour information on the style ETF from Russell. Arshanapalli et. al. use use four of these funds:
The site