Saturday, March 14, 2015

Why Smart Beta Strategies Work---It's Not What You Think

Smart Beta strategies are hot.  Go to any ETF conference and the majority of presentations will be about Smart Beta.  Smart Beta is any type of indexing methodology that is not market cap weighted, in an index like the S&P 500 companies are weighted by how large they are.  A Smart Beta approach to the S&P 500 would weight companies differently---equal weight, fundamental weight, low beta, high beta, etc.

It makes sense that Smart Beta is hot right now, ETFs are the future of asset management but we don't need another market cap weighted index.  So if you are going to issue a new ETF it has to be something different.  Add in the fact that the research shows that just about any other way to weight an index beats market cap weighting over time.

The million dollar question is why does Smart Beta outperform market cap weighting and will it persist?  One idea why is simple, curve fitting.  With computer technology you can find a bunch of anomalies that beat the market.  To guard against this you need to start with a valid premise that should work, and then test it to see if it does.  For example, weighting companies by fundamental factors, meaning the companies in the best shape get a larger weight in the index and the companies in the worst shape get a lower weight, should beat simple market cap weighting.  You can then test it and see what the results are.

The link below it a video from Rob Arnott talking about Smart Beta, it is a couple of years old but makes an interesting point about why Smart Beta strategies likely outperform.

He took a look at a bunch of different Smart Beta approaches and found that they beat market cap weighting.   He then took the opposite of those approaches and found that they also beat market cap weighting.  On the surface this makes no sense.  For example, if overweighting low beta stocks beats market cap weighting because their is something uniquely powerful about low beta stocks, the the inverse, overweighting high beta stocks, shouldn't beat market cap weighting, but it does.

The answer most likely is that Smart Beta strategies, regardless of how they are constructed, encompass two tilts---value and size---that do better than market cap weighting.  In a market cap weighted index you are always going to overweight the most expensive and therefore the largest stocks.  In most Smart Beta approaches you are going to tilt more towards value and smaller stocks, this is most likely where the outperformance comes from.

For example, we use the Guggenheim Equally Weighted S&P 500 (RSP) ETF a lot in our strategies. Like the name suggests it equally weights all the companies in the S&P 500.   According to Morningstar over the past 10 years (as of 3/13/15) RSP has an average annual return of 9.13% vs. a return of 7.76% for the S&P 500.  Also according to Morningstar, the average market cap of the portfolio is $21.9 billion vs. $73.4 billion for the market cap weighted S&P 500.  So where did the outperformance come from?  It could be a random anomaly that was curve fit, but most likely it is the size tilt towards smaller cap stocks.

Will Smart Beta outperformance persist?  Who knows.  However, it makes sense that value---buying something for a low price, and small stock outperformance should continue over time so any strategy that tilts towards those factors should be fine.

No comments:

Post a Comment