Small-Cap Contrarian Rules

RULE 33

Small-cap investing: Buy companies that are strong financially (normally no more than 60% debt in the capital structure for a manufacturing firm).

RULE 34

Small-cap investing: Buy companies with increasing and well-protected dividends that also provide an above-market yield.

RULE 35

Small-cap investing: Pick companies with above-average earnings growth rates.

RULE 36

Small-cap investing: Diversify widely, particularly in small companies, because these issues have far less liquidity. A good portfolio should contain about twice as many stocks as an equivalent large-cap one.

RULE 37

Small-cap investing: Be patient. Nothing works every year, but when smaller caps click, returns are often tremendous.

What we see clearly is that the small-cap and low P/E effects are separate. Buying small companies does give you higher returns, but only if you select low P/E stocks." The low P/E effect continues to be strong regardless of the size of the company. Even buying the largest group, the return on the lowest P/E quintile is significandy above the highest P/E set. Sanjoy Basu, in a 1983 paper, obtained similar results, showing that the low P/E effect occurs for companies of any size, and further, also demonstrated problems with Reinganums methodology.

Do these results hold for other contrarian strategies? Yes. We found very similar performance by using small-cap price-to-cash flow, small-cap price-to-book value, and, to a lesser extent, small-cap price-to-dividends. The price-to-book results are shown in Table 15-2. Buying

When the Compustat tapes were divided into live groups by market size, and also sorted into five P/BV groups, the annual retums (in the five columns below) of the lowest P/BV were significantly better, regardless of the size of the company.

♦Market capitalization on January 1, 1995.

medium or small-cap stocks using one of the contrarian strategies not only provides higher retums over the highest P/E group, but magnifies them, because they outperform their larger co-peers. Once again, though, remember the caveat about the lack of liquidity and the transaction costs in the peanut-sized stocks.

Why do contrarian small-cap stocks do better? Although I cannot give you a definitive answer at this time, I think there are at least two reasons. First, these stocks are not glamorous, but are dam good businesses, such as the local bank or a small company in an unexciting but expanding industry. The stocks, like their big-cap siblings, get knocked down too far when they are unpopular and then recover just like their larger kin. Too, they are often overlooked by all but local or regional investors, and can thus plod along at cheap prices for years. Often a larger company sees they make a good fit, and they can also be bought on the cheap. A merger is consummated well above the previous trading range.

Naturally there have to be other caveats. I talked about survivorship bias in small cap stocks. Doesnt this also hold for small contrarian stocks? Not as much, but lets explore the question a little further. Usually the sizzlers are more thinly financed. Large numbers of IPOs swarm into the marketplace to meet the heavy demand for the concept of the day

Analyzing Ritters results, for example, indicates many IPOs were simply startups, usually high on expectation and low or nonexistent on actual revenues and eamings.* His 1991 study also showed that value industries had been in business for much longer than the concept industries. The average age of the value IPO was 12 years, compared to

Table 15-2

Small Cap or Small Low Price-to-Book? 1970-1996

Beware of Nasdaq and Small Stock Trading Costs

Before you msh out to buy small-cap contrarian stocks, remember diey have the same narrow markets endemic to all small stocks. It will cost big to make a ti-ade. Transaction costs are not only die costs of commissions for switching from one stock to another, which, while they can be high, would still leave you with a big part of the pie.

The more important spoiler is the spread: the difference between the price you can buy a stock for and what you can sell it for, which we examined in detail earlier. Even today, while having a relatively minor impact on larger companies, the spreads for pint-sized companies can be enormous-30% to 40%, sometimes more. It has been estimated that the average buy-sell or bid-ask spread of die Russell 2000 (the 2,000 smallest stocks among the top 3,000 traded) is 2.75%, against less than one half of 1% for the S&P 500. Remember, there are thousands of companies smaller dian diose indexed in die Russell 2000. Normally, the smaller die company the higher the spread. As weve also seen, die

several for the glamour segment. (The study excluded penny stock offerings, and stocks not traded in major markets, as well as several other classifications which, if included, would probably have made the comparisons for the glamour sector even worse.).

What this boils down to is, yes, there is very likely a survivorship bias, but one that understates the returns of small-cap conti-arian stocks and overstates those of high-priced small companies. Why? Because, as noted, contrarian companies are normally in unexciting but profitable industries that have the financial strength to endure. Failures are more hkely to be in exciting new sectors where growth is rapid, but competition is intense and financing is flimsy. Innovation and technology advance rapidly, wiping out the temporary advantage of thousands of such companies.

The latter are die ones diat likely dropped out of the less sophisticated databases a couple of decades back. While the jury is out until we get further proof, my assumption is that the retums of high flyers in the small-cap sample are overstated because of diis problem. If a more comprehensive index could be cons04icted that included the thousands of IPOs brought forward in die last few decades, and notjust die small-cap survivors, I believe the statistics favoring the small-cap contrarian strategies would be even stronger. I have stated for years that the small-cap effect is almost entirely a contrarian effect for this reason. Do I have any takers on this bet? Professors Banz, Reinganum-are you there?