back start next


[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [ 107 ] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155]


107

An Attack on Contrarian Strategies

Marc Reinganums findings supported Banzs. Reinganum concluded from his worlc that the only low P/E effect is in small stocks. In his work, Reinganum divided his P/Es into deciles. His lowest decile has a P/E ratio of slightly over 1, compared to 6.5 for his next lowest decile. A P/E of I, of course, indicates serious trouble. Healthy companies- large or small-do not trade at one times eamings. (This in effect means the investor receives a 100% after tax payback on his investment each year on the stock.)

Most of the stocks in his low-P/E sample were very small. Twenty-two percent had market values of $13 million (which are inordinately small capitalizations in recent years), and fully 65% had market values under $68 milhon. By comparison, only 2.4% of companies in the largest capitalization decile traded at P/Es of 1.

What do Reinganums results really show? Like Banz, he was actually measuring something very different from what he thought. In his case Reinganum is demonstrating a small, troubled company effect, not a low P/E effect. What these findings indicate is that investors react sha ly to the financial difficulties of smaller companies. The fear of bankmptcy or severe operating problems drastically shrinks their prices. This reaction is even more severe than for larger companies. Small companies that recover bounce back sensationally; those that dont are delisted. This also seems supported by the fact that small companies normally trade at higher, not lower, P/Es than larger concems. Again, the then-novice professor seems to have had little understanding of what lay under the statistics.

But what about other studies reputed to show the superior performance of small-cap stocks? Roger Ibbotson, a distinguished academic researcher, who is best known for his studies in collaboration with Rex Sinquefield on long-term retums on stocks and bonds, computed the value-weighted retums* on what was termed the second tier: all stocks other than the large companies in the S&P 500 from 1958 through

* A value-weighted index assigns a weighting to each stock in the index proportional to the market value of its stock.

demic researchers. Ive written five additional articles on small-cap research to the present time. The criticisms discussed above have never been challenged.



1979.* Over this 22-year period the groups total retum was 967% versus 504% for the S&P 500. Other researchers came up with similar effects.

These findings too are problematic. By weighting companies by their market values, Ibbotson assigns small companies to a relatively minor part of the sample, although there may be several thousand of them. Second, the liquidity problems endemic to the Banz study are also imbedded in this one. Adjusting for transaction costs of small companies takes a big bite out of their retums. Compounding the difficulty is that there are large numbers of big companies that are not in the S&P 500. These stocks were put into the small company sample. Many of the large banks, insurance companies, and a good number of other bigger concems traded on Nasdaq or its predecessor during this time. Weighting the companies by market size means the performance of 50 companies with a market value of $40 million, or 20 companies with a market value of $100 million, count for no more in performance than one company with a market value of $2 billion.

So what the study almost certainly calculated was the performance of medium and some larger-sized companies. In a word, the "small company effect" the researchers believed they were measuring was drowned out!

Another issue that questions the credibility of these results can be termed "survivorship bias." Many of the small companies measured ran into problems and dropped off the face of the earth. In the less rigorous databases prior to the 1980s, hundreds disappeared. Their businesses faded, and their stocks went off the screens unnoticed.* Of the scores of PC companies underwritten in the hot new issues market of the early eighties, very few are around today. The same was the case with the computer leasing companies of the late sixties and undoubtedly will be tme ofthe Intemet group shortly. The over-the-counter market, which at times traded over 5,000 companies, did not keep statistics on nonsur-vivors. Also ignored is the high likelihood of price declines as failing companies near Chapter 11.

You should note two other points, if you own or are thinking of buying small-cap stocks. First, confirming evidence that small stocks do not do well over time comes from the Loughran and Ritter and Forbes studies of IPOs, which we will look more closely at in the next chapter. Loughran and Ritter examined 4,753 IPOs between 1970 and 1990.* They found the average retum of these companies over five year periods was 3%, versus 11.3% for the S&P 500. No big bucks from small cap-

* Most likely traded, if at all, in "the pink sheets," where there are normally only one or two market-makers and spreads are enormous.



We Have a Tie-in to Mecca

What advocates of small-cap stocks may have lacked in research, they made up in chutzpah. Take Rex Sinquefield, who coauthored with Ibbotson. He became one of the founding partners in Dimensional Fund Advisors, a firm formed in late 1980 to take advantage of the small-cap effect. On the board or serving as advisors were ten high-powered University of Chicago professors or alumni, including Roger Ibbotson, Eugene Fama, Merton Miller, Myron Scholes, and Rolf Banz. Sinquefield, when asked what he thought of active portfolio management, sneered, "crap."° Sinquefield was hyping his new small-cap product at die time, whose hype naturally included the spectacular performance I questioned above.

* The most commonly used index to measure small-cap performance.

stocks here. This sample is much closer to what Banz or Ibbotson believe is small cap than those that were examined in their own studies.

The last point to be aware of is that small companies are expensive. Sure, everybody likes the idea, held forth by Banz and his colleagues, that thousands of these bargain basement values lie around waiting to be plucked up by the percipient investor. But once again we have a case of an excidng theory, in the words of Thomas Huxley, "killed off by nasty, ugly litde facts." The average P/E of the Russell 2000* over the last 20 years has been above that of the S&P 500 consistendy with only one or two exceptions since the indexs inception in 1979. Dividend yield has also been significantly lower than that of large stocks, while price-to-book value has been higher than that of large caps about half the time. So once again small companies have not been in the bargain basement as a group-with one important exception that well look at shortly.

Right or wrong, the academic thinking swept through the investment world like a brush fire. Billions of dollars and probably millions of investors, either directiy or indirecdy through their retirement plans, have purchased small-cap stocks. What we have seen is that the Banz research, the major study demonstrating superior performance, is so seriously flawed that adjusting for only some of the errors wipes out most, if not all of its claims. There are also major questions about the Ibbotson study. But to me the most telling point is what happened when the professors took their pet theory out of the laboratory and put it into practice.



[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [ 107 ] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155]