Review of Tweedy Browne's "What Has Worked In Investing"
In last week's review of Morningstar's MFI article, the article's author referenced a pamphlet from investment company Tweedy Browne titled "What Has Worked In Investing" (download it here). MagicDiligence took a look at the pamphlet and there were some interesting results referenced in there that may be of interest to Magic Formula investors, as the MFI strategy adheres very closely to some of the mechanical methods referenced within. In this article, I'll attempt to summarize the basic conclusions provided in the pamphlet and add some of my own thoughts as well.
At the very beginning, Tweedy mentions 5 mechanical (statistical) criteria that they believe identify stocks that may present investment potential beyond the returns of the general market. These 5 criteria are:
1) Low Price in Relation to Asset Value, or Low Price-to-Book ratio.
2) Low Price in Relation to Earnings, or Low Price-to-Earnings ratio. This can also be manifested as a high earnings yield, or Earnings-to-Price as a percentage.
3) Significant Purchases by One of More Insiders. Large share repurchases can be a component of this.
4) Significant Decline in a Stock's Price.
5) Small Market Capitalization.
The pamphlet then goes on to cite numerous studies examining the performance of stocks meeting one or more of these criteria. These studies are spread out in different time intervals, and several of them examine foreign markets.
The first 2 are closely related, and a third criteria, low price-to-cash flow, exhibits similar results. When the universe of stocks in any exchange (U.S. or foreign) are lumped into groups of the lowest vs. highest P/B, P/E, or P/CF, the groups of lowest multiples consistently outperform the groups of higher multiples, usually in a linear pattern (that is, the lowest outperform the preceding higher group). A couple of interesting studies looked at the performance over 1, 3, and 5 year periods. Speaking generally, the lower multiple groups outperform about 70% of the time over a 1 year period, about 85-90% of the time over a 3 year period, and astoundingly, 100% of the time over a 5 year period. Over the long term, the lowest multiple groups outperform the highest multiple groups by about 4-6% annually for any of the three multiples.
Let's jump ahead to the 5th criteria, small market capitalization. There are many reasons to own small cap stocks, but the only one that matters is that they outperform larger cap stocks in aggregate. The pamphlet shows several studies, across several stock exchanges, that shows that an exchange's smallest stocks by market cap outperform the larger ones over periods of time. Similar linear performance amongst deciles (10 groups) or quintiles (5 groups) are exhibited, where the smallest caps outperform the 2nd smallest and so forth.
A study is also cited that shows that combining market cap with one of the low P/B, P/E, or P/CF leads to the best performing class of stocks, statistically: small-cap value stocks.
I cite these three criteria first because they are relevant to Magic Formula Investing. The strategy's "bible", Joel Greenblatt's The Little Book that Beats the Market, has a similar study where stocks are broken into deciles and exhibit the same linear improvement based on the MFI statistics of adjusted earnings yield and return on capital. Also, Greenblatt shows that including small market cap stocks improved the performance of MFI by about 10% a year over just large cap stocks. What Tweedy's pamphlet does is lend additional credence to the notion that small-cap, value-based stocks are the best way to go.
The other two criteria may be something we can apply to "magic diligence" to further drill down into the MFI stocks to find the most attractive ones.
Let's start with #3 - insider purchasing. I found the data here to be less compelling than for the 3 aforementioned attributes, but still interesting. Tweedy references a study that showed stocks with significant insider buying outperforming the market by anywhere from 7-20% over short periods of time (1-3 years). This makes some sense, as any edge the insiders have (new products, new marketing campaigns, etc.) is likely to be of a more immediate nature. Considering the Magic Formula holding period is just one year, looking for insider purchases could be a good way to identify exceptional MFI stocks.
Finally, there is #4, a significant decline in a stock's price. I was a bit skeptical on this one, as there are also several studies that show that "winners keep on winning"... that is, stocks that have outperformed the market over the past year continue to outperform. Tweedy only references 2 studies, both of which show that stocks that have underperformed over a previous 5 year period tend to outperform over the succeeding year-long period, by 18% no less.
While this is interesting, I'm not so sure it can be used to help us find better Magic Formula stocks. For one, many of the small-caps in the MFI screen barely have 5-year histories trading on one of the large exchanges. Second, there are other studies (specifically in O'Shaughnessy's What Works on Wall Street) that show the opposite effect over shorter time periods. At the same time, that book also confirms most of the other studies from Tweedy's pamphlet.
In all, the publication is an interesting read, and confirms many other studies that say the same thing: value-oriented, statistically-based investing methods greatly outperform market averages, particularly when they involve small-capitalization stocks. If nothing else, it serves as further evidence that Magic Formula investors are on the right track to vastly outperforming the market during their investing career.