ALPHA GENERATION REVIEW: QAS TOP 20 U.S. SMALL CAP INDEX
The QAS Top 20 U.S. Small Cap Index continued with strong performance this year. As of 7/13/17, the index is up +6.2% Year-to-Date (vs. Russell 2000 +5.7%), and up +36.2% 5yr average annualized return (vs. Russell 2000 +22.1%). This is a very sizable alpha reading, and we would like to look inside and find out what parameters contributed to this performance using Bloomberg <PORT> functionality.
“Small Cap” is considered by portfolio managers and researchers as one of the most volatile and difficult size segments to work with. Therefore, to reduce risk, portfolio managers tend to increase diversification of small cap portfolios and hold hundreds of small cap stocks with a minimum position size. However, when one increases portfolio size to 100 stocks and above, one tends to increase correlation with the benchmark, and the risk/return profile starts to look very much like the one of the benchmark as well.
An intelligent investor may raise a reasonable question – why would I pay additional fees for a product that performs in close range with a much cheaper ETF product based on a broad-based index benchmark?
As part of our on-going model portfolio research we created multiple “building block” strategies with a clear investment objective to outperform relative to their respective benchmarks. Our research revealed that if you can effectively control the risk parameters, then an optimal portfolio size can be as low as 20-30 stocks.
Our risk control process has nothing to do with traditional volatility analysis. We believe that the volatility parameters can be misleading in evaluating investment opportunities. Instead, we focus on our own risk framework that analyzes simultaneously multiple time-frame momentum inputs (both absolute and relative readings must be positive) that identify the best stocks in the index that have the biggest upside potential (long only structure), and therefore, the lowest risk level. The rebalancing frequency and allocation scheme are also important. Our “sweet spot” combination is a “monthly rebalancing + equal weight” structure.
There are several points of analysis that we track periodically in our Bloomberg <PORT> Attribution report.
First of all, we can clearly see what portion of our alpha (active return) was contributed by two main inputs, selection (stock picking process/algorithm) and allocation (position sizing technique and rebalancing frequency). As you can see, the “selection” input contributed 79.36 out of 93.07 points of the alpha (85%), and the “allocation” input contributed 13.47 out of 93.07 points (14%).
Secondly, we are interested in a sector-based attribution analysis. It is clear that our portfolio structure works well in identifying alpha generating stocks in almost all sectors except the health care sector, which we should look into with a more detailed analysis. However, we can make the conclusion that our process is relatively stable across all major industry sector groups. We believe that this is a very attractive element for a reliable portfolio construction process.
All model portfolio and index products are available for use by institutional investors (www.qas-service.com)
Quantitative Analysis Service, Inc.