Draft Version: January, 2008
This paper tests the hypothesis that many investors overlook persistent firm differences in the amount of investment required to grow sales and profits. I focus on asset-intensity (operating assets/sales) as a proxy for the net investment a firm needs to attain a given growth rate and show that firms with heavy (light) asset-intensity have lower (higher) subsequent stock returns. A long/short portfolio based on this effect yields abnormal returns around 0.4% per month. Asset-heavy stocks miss consensus analyst earnings forecasts 16% more often than asset-light stocks, suggesting investor forecast error is the source of the return difference. Consistent with the hypothesis that investors make a mistake about the investment required to grow, asset-intensity is a stronger predictor when expected or ex-post investment is high.
The dataset covers all stocks traded on the NYSE, AMEX, and NASDAQ from July 1963 to June 2006. ADRs, REITS, financials, and stock indexes are excluded. All stocks must have at least a $25mm market value so the results are not screwed by very small, illiquid firms.
All the accounting variables are from Compustat.
Asset intensity, defined as operating assets/sales, effectively measures how productive a firm is. A high operating assets/sales ratio would imply a 'heavy' asset intensity measure and a low operating assets/sales would imply a 'light' asset intensity measure. For non-geeks, Warren Buffett would be a big fan of light asset intensity firms, or firms that can generate a lot of revenue (and hopefully profit) off little capital investment. These sort of firms usually have the proverbial 'moat' that value investors mentions always mention: something these firms own allows them to produce a lot of value on small investments.
In an efficient market, it should be the case (controlling for risk) that firms with light asset intensity should have lower expected returns (less risky and more profitable) than high asset intensity firms (more risky and less profitable); however, the issue is quite complicated. Raife tackles the problem and finds that light asset intensity firms actually have higher returns.
Not surprisingly, light asset intensity firms persistently have higher returns on existing assets and higher returns on new investments (for heavy asset intensity firms the effects are opposite). Moreover, using the asset intensity measure to create portfolios works in nearly every investment situation Raife tested: small versus big, within industries, after controlling for investment, accruals, and return on equity, etc. This strategy is a great tool for separating winners from losers (coupled with Piotroski's F-Score, you could probably create a super portfolio).
Specifically, after sorting the stock universe based on their asset intensity ratio every July 1 (presumably after all the accounting data is available), long the light intensity quintile and short the heavy intensity quintile--you will earn .49% a month after controlling for the Fama-French risk factors.
For institutional investors, this strategy is simple: have your data monkey pull all the required data off of FACTSET or CAPITAL IQ or Bloomberg or WHATEVER OTHER EXPENSIVE DATA SERVICE THAT IS GETTING EXPENSED TO THE FUND YOU ARE MANAGING. Next, with the data, sort all the stocks on their asset intensity. Long the top decile and short the bottom decile.
For individual investors, this strategy is even simpler: go to Yahoo Finance, or any number of free screening services and rank stocks on Return on Assets (essentially captures the asset intensity). Then, invest in the high ROA stocks and short the low ROA stocks.
Now, if you really want to get some market beating returns, use the asset intensity measure within the growth portfolio. First, sort all the stocks on the book to market ratios, then within that portfolio sort on the asset intensity. If you do this sort of strategy, you would have earned .63% per month above and beyond size, value, market, and momentum risk factors (See Table IV).
Implementation Issues and Remarks:
Raife does a great job explaining the economics of why and how alpha is generated using this strategy--I buy pretty much everything he says.
Implementation of this strategy, if you use a more simple version and just focus on ROA to sort your stock universe, is VERY easy. That said, I'm not sure anyone would want to set up a long short portfolio that is solely based on asset intensity; however, when building your stock picking model, or when choosing individual stocks, it is important to note that on average companies that earn a lot of sales (which presumably lead to more profits) on a small amount of assets are a good bet.
In summary, from an implementation/cost perspective this strategy deserves a 9/10 or a 10/10 Investment Potential Rating. But from a return potential perspective I would give it a 6 or a 7--the returns are rock solid, but they are not pure arbitrage or really breathtaking. Overall, I give the asset intensity investment strategy a 7.5/10 Investment Potential Rating; good, but not truly outstanding.
This is a great paper and a great explanation of how fundamental stock characteristics can have real affects on asset prices--you need to read this paper if you are a serious value investor.