Relative breadth was again very narrow
Perhaps the most influential variable in determining the success or failure of a portfolio manager is the relative breadth of the market. This measure is not the absolute breadth used by technicians, the number of stocks up versus the number of stocks down; it is the percent of the S&P 500 companies that outperform the index. For example, if 300 companies in the S&P 500 outperformed the index in a given year, then the relative breadth would be 60%.
Normally, about 43% of the stocks in the S&P 500 outperform the index during a calendar year, based on data from 1986 to 1999. If we exclude 1998 and 1999, however, that figure rises to 45%. Note that the odds are still skewed against managers. That means that managers had only slightly better than a one-in-four chance of picking an outperforming stock. Multiply those odds by a 25- or 30-stock portfolio, and it becomes quite clear why most managers underperformed in 1998.
In 1999, the relative breadth increased to only 31%. Obviously 1999’s relative breadth was not as narrow as 1998’s, but it was still extraordinarily low. In fact, 1998 and 1999 were the two narrowest years in the last 14 when using our measure of relative breadth.
Little skill needed in 1999 to pick outperforming technology stocks
Although the relative breadth of the overall market was extraordinarily narrow again in 1999, the relative breadth of the Technology sector was extraordinarily high compared to that of various other economic sectors. This means it took very little stock-picking skill to invest in a technology stock that outperformed the S&P 500 during 1999.
Nearly 70% of the stocks in the Technology sector outperformed the S&P 500 during the year. That implies one could have figuratively “thrown a dart” at a universe of Technology stocks to pick a winner for the year, and that the skill needed to pick a winner was relatively small when compared to the skill necessary to pick outperformers in a number of other sectors.
This may explain why there has been a trend among individual investors to shift away from managed funds to self-trading. Individuals have focused primarily on the Technology sector. Our guess is they attribute their success to their superior skill relative to the diversified fund manager rather than to chance. Our guess is also that few compare their performance to Technology indices or Technology funds.
The percent of the companies that outperformed the market in the Basic Industrial sector was the closest to the Technology sector’s, and it was only 51%. Thus, while the probability of picking a stock that outperformed the S&P 500 from the Technology sector was roughly 70%, the best that one could have hoped for in a sector outside of Technology was 50/50.
The probability of picking an outperforming stock was amazingly low in certain sectors. For example, investors looking at Financial or Consumer Cyclical stocks had roughly a four-out-of-five probability of picking up a stock that would underperform the overall market. But investors in Consumer Staples stocks had a 92% probability of picking a loser.
We wrote a report several years ago highlighting that managers were going to have a difficult time outperforming during the second half of the 1990s if the profits cycle continued to decelerate. Market leadership tends to narrow when the profits cycle decelerates because the market becomes extremely “Darwinistic.” Only the fewer and fewer companies that can continue to grow their earnings tend to outperform during such periods, while the performance of the companies whose earnings suffer lags.
That description certainly fits 1998 because the profits cycle decelerated for the fourth straight year. However, that does not help one understand why there was narrow leadership in 1999 because the profits cycle actually re-accelerated. Typically, when that happens, both relative and absolute breadth widen.
We continue to believe that 1999’s narrow leadership was the result of a speculative bubble in the “Nifty 50” and Internet stocks. Both interest rate and earnings fundamentals during the year did not support their outperformance. See “Ignoring the Catalyst,” Quantitative Viewpoint, December 15, 1999. We expect relative breadth to widen in 2000, and expect an increasing number of active managers to outperform their bogeys.
C&Ds and relative strength were the best performers for 1999
If we look at the annual performance of the roughly 45 strategies that we regularly monitor, the best performer for 1999 was our C&D Index, which was up more than 71% for the year. It was the strongest year for lower-quality stocks since 1991, when the C&D Index was up more than 90%.
The second-place finisher for 1999 was our Relative Strength Model, and it was up 64.3%. It should not be surprising to most investors that this particular model performed so well, given the “Nifty 50” environment that was so pervasive during the year.
Third- and fourth-place finishers should also be no surprise: High Five-year Projected Growth (+60.6%) and High Beta (+60.3%). Technology stocks are often characterized as high-growth, high-beta stocks, so it seems natural that these strategies should perform well during a year in which Technology dominated the market’s performance.
The two worst-performing strategies were our A Index and High Dividend Yield. This makes sense given that the A Index typically has the highest dividend yield of any of our proprietary “Quality” Indices. The A Index was down 6.5%, and High Dividend Yield was down 7.7%.
If one negates the “Nifty 50” effect, as our “Quality” Indices do by nature of their equal-weighted construction, then equity market performance in 1999 did indeed reflect that the profits cycle was re-accelerating. Lower-quality stocks handily outperformed higher-quality ones. The constituents of our “Quality” Indices are equal-weighted, and thus, 1999’s “Nifty 50” effect is somewhat negated by their construction, i.e., large-capitalization stocks and small-capitalization stocks have the same influence on the index’s performance.
Thus, our B+ and A- Indices performed relatively poorly despite the fact that many of the “Nifty 50” Technology stocks are contained in those indices.
GARP: growth at a reasonable ridiculous price
The Alpha Surprise Model is a GARP-based, growth at a reasonable price, strategy, and has historically been one of our most popular stock selection strategies. The model has outperformed the S&P 500 ten out of the last 13 years. As a blend of growth and value, it has been able to outperform during “growth” markets and “value” markets.
Given the narrow leadership of the market the past two years, it has become increasingly difficult for our strategy to outperform. It appears that growth at a reasonable price has been replaced by growth at a ridiculous price.
In 1999, our Merrill Lynch vs. Consensus Positive Earnings Surprise Model was up 20.6% compared to the S&P 500, which was up 19.5%. This performance was particularly impressive because we equal-weight the model’s portfolios, and hence the model only had a 31% chance of outperforming the S&P. This model comprises 75% of the Alpha Surprise Model.
The Dividend Discount Model, which comprises the other 25% of the Alpha Surprise Model**,** was up a mere 4.2%. This is not surprising, given that most value-oriented strategies performed poorly during the year.
However, adding just 25% value to an outperforming strategy in a market that does not care about valuation killed the overall Alpha Surprise Model’s performance. The Alpha Surprise Model was up 8.4%, underperforming the S&P 500 by 1,110 basis points.
Further demonstrating the efficacy of our EPS Surprise Model, our Merrill Lynch vs. Consensus Negative Earnings Surprise Model was among our ten worst-performing strategies in 1999. The model was up 3% for the year. However, adding 25% overvaluation or using the bottom S&P 500 companies by Alpha Surprise Model, one would have outperformed our traditional Alpha Surprise Model by 62 basis points.
Searching for overvalued stocks in 1999 appeared to be a good idea. As mentioned, our top-performing strategy last year was our C&D Index. It was up 71.4% and its current portfolio trades at 70 times current-year earnings. Our second-best-performing strategy was our Relative Strength Model. This model was up 64.3% and its portfolio trades at 87 times current-year earnings.
The “Nifty 50” by market capitalization, which we have been underweighting since the end of April, was up 20.1% in 1999 and currently trades at 78 times earnings. In contrast, the “Not-So-Nifty 450” was up only 7.7% in 1999 and trades at a relatively mere 26 times current-year earnings. Based on the performance of our Dividend Discount Model and the “Not-So-Nifty 450,” it actually hurt managers to search for undervalued stocks in 1999. This is an important point for GARP managers who may have underperformed this year.
We expect 2000 to be a better year for active management
As we have repeatedly stated, the more optimistic one is about the rejuvenation and expansion of the global economy, the more one should emphasize active management over passive management of one’s bogey. Active managers have had difficulty outperforming their benchmarks not because they suddenly became idiots, but rather largely because of the global deflation and recession, and the resulting decelerating profits cycles. As stated, market leadership tends to narrow when profits cycles decelerate, and markets become very “Darwinistic.” The narrower the leadership, the lower the probability that an active manager will outperform.
We expect global profits cycles to continue their present upturns. Accordingly, we believe that 2000 will be a better year for active managers. The time to index and constrain tracking error was in 1995. Today, investors should be encouraging managers to deviate from their benchmarks.
