Whether the figures were correct or incorrect, they were highly marketable. And Sinquefield was certainly an academic who could market. Here, as quoted in Forbes in a 1981 article, is part of the spiel. "Take the S&P 500 for 1930 through 1980; the retum is 9J4% a year. But for the bottom 20% in terms of market value, its 14.6%" The article continued, "Put another way: $1 million in S&Ps 500 would have grown to $93 million, but to $924 million in the small companies. Better than that you can hardly do."

For further effect Sinquefield adds, " Its not a fluke. Our computer tells us it happens but none of us knows why." Sinquefield resoundingly concludes," Were plugged into the academic world, where all the real good research is taking place. We have a kind of tie-in to Mecca.

The Forbes article went on to report the reaction of some of the critics, including me. "Forbes columnist David Dremans rebuff of North-westem Professor Rolf Banzs favorable study of small company performance was like a slap in the face with a wet fish. Dreman says basically that the University of Chicago study was based on faulty data, which skewed the results. Dreman said of the Banz study, in effect, garbage in, garbage out."

Still, give credit where credit is due. The marketing of the product, and the founding of a major investment firm. Dimensional Fund Advisors (DFA), on it could make a case study at tiie Harvard Business School, even if the academic work was faulty and has proved costiy to investors who followed it. By the end of 1996, Dimensional Fund Advisors had $11 billion under management, a good part from the small-cap product.

The professors associated with DFA went along with their Wall Street marketing cronies, happily wallowing in dollars. Ironically, it is this group, some of whom were among the original cmsaders to save the public from the clutches of Wall Street ignorance, who refuse to answer reasonable criticisms of the performance of peanut stocks.

But, you might ask, were the small stock findings really as badly flawed as Ive indicated? After all, as Sinquefield stated, they had a tie-in to Mecca as his firm was plugged into the finest academic research available. How well, then, did the Dimensional Advisor small-cap fund perform? Poorly, by any yardstick.

For the 5 years ending October 1990, tiie S&P 500 increased 77.6% in value, while the DFA Small Company Fund (called the 9-10 fund, which closely approximates the real small-cap universe) rose only 1.15%. These figures come from an article by Marc Reinganum, which argues that over time there are predictable reversals. Large cap

Should You Give Up on Small Stocks?

Dr. Reinganum, as we saw, attempted to dismiss the low P/E findings as a small-cap effect. Nothing could be further from the truth, as Table 15-1 indicates. The study, from 1970 to 1996, is an update of one I published in Forbes on July 23, 1990. It shows two important investment findings. First it demonstrates that there is a low P/E effect regardless of the size of the company. Secondly it shows a definite small-cap effect, which results in superior returns to smaller companies-but this effect

TabU 15-1

Small Cap or Small Low P/E? 1970-1996

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

Market

*MaRKET CAPITALIZATION ON JANUARY 1, 1995.

may beat small stocks in one 5-year period, and predictably will typically get beaten the next time around. Good sleight-of-hand! Alas, the "predictable reversal" Reinganum promised never came. From its inception in 1981 to tile end of December 1997 the DFA 9-10 Fund trailed the S&P 500 by 54%. Some of the academics involved with the fund were disturbed by its performance. One said, "This has never happened before, were stunned."

DFA continues to market the small-cap funds on the basis of the same problematical returns, reporting that small stocks earned 45% annually at the bottom of the Depression. Any novice in the math of compounding rates of returns knows that if you start with these enormous results as a base, whether theyre correct or not, the S&P 500 is likely never to catch up. It appears, then, that Sinquefields direct line to Mecca is out of order, and has been for the last 72 years.

is contributed entirely by small, low P/E stocks. Lets look at these results more closely.

The study was again done in collaboration with Dr. Eric Lufkin of the Dreman Foundation. To build our database, we took all the companies on the Compustat tapes with market value greater than $100 million" and divided them by both P/E and market size. The lowest 20% of stocks-as ranked by price-to-eamings ratios-are in column one, progressing to the highest 20% in colunm five. Looking across the table, the lowest 20% of stocks by market value ($100 milHon to $500 million) are in the top row, the largest 20% (larger than $5 billion) in the bottom row.

The lowest P/E group outperformed the highest group for companies in all five market-size categories. The low P/E effect exists for stocks with market values of $1 to $2 billion, as surely as at $100 to $500 million. The difference in retums is formidable. Had you invested $10,( in the lowest P/Es of the lowest market-size group back in 1970, it would have been worth $1,003,( in 1996-more than 4 times as much as the $240,000 for the highest P/E group.

The results are even more lopsided for companies in the second-lowest market cap range (row two). The lowest 20% of P/Es here provided an 18.8% annual retum, versus 10.4% for the highest P/E group.

Moving to the largest stocks (stocks with a market value greater than $5 billion) in row five, the lowest P/E group provided a 14.2% retum versus 11.9% for the group average and 8.7% for the highest P/E stocks. Whats more, these low P/E behemoths with supposedly sluggish futures outperform every category of high P/E stocks including even the smallest-cap group ($100 to $500 million) handily.

The table also demonstrates that the low P/E strategy is even more effective as company size is reduced. Returns rise as market size decreases, but again we have to be very careful when we get down to "midget" companies. The superior retums, low P/E or not, are likely to be offset to some extent by higher transaction costs. In addition, there is the greater risk that the companies down here will not survive.

The study provides some backing for an investment strategy in medium-sized and smaller stocks almost identical to the one I recommend for large low P/E issues. With the smaller issues you must be prepared to buy them and hold them or the tumover costs will eat you up. This means you have to be very sure the stock is an extra-good value.

Here are five pointers that should be helpful in following the contrarian approach for smaller and medium-sized companies. As you can see they are littie different from the mles given for their big brothers, but the penalties for breaking them could be even more severe for your portfolio.