Wind Shift or Sea Change?

inds can change quickly and dramatically. Sailors in a race react accordingly by rapidly adjusting the sails. It's tiring, but over a short distance it's feasible. Passage making, the art of making a long voyage, is a very different proposition. The sails are initially set and the ship moves with the wind. Tactical adjustments are made, though they are much more subtle. During periods of contrary winds the hands at the helm will tack. Progress is slowed. The experienced captain continues to watch and wait, knowing the winds will return. No need to re-design the ship or the voyage plans. Should the captain encounter a sea change, more dramatic responses might be required. The key is discerning a wind shift from a sea change.

During the latter part of the 1990's equity investing (to some) appeared to be a no-brainer. Growth winds were strong and apparently steady. If an investor wasn't heavily invested in high tech growth stocks and funds, they were thought by many to be missing the boat. Now, a few years later, the equity market indices — which were flirting with 10 year average returns near 20% at year end 1999 — have normalized to levels that reflect long-term reality. The gale force equity winds of the 1990's have subsided (see Figure 1).

As shown in Figure 2, the S&P 500 had produced a 10 year annualized return of 18.2% as of 12/31/99. Just three years later, at year-end 2002, the 10 year average was a much saner 9.3%. The S&P Midcap 400 had a 10 year average return of 12.0 at year-end 2002, down from 17.3% in 1999. The Russell 2000 Small Cap index had a 10 average return of 7.2 percent on 12/31/02. At year 1999 it had averaged 13.4% over the prior ten years.

From the vantage point of "3 years later", we can clearly observe that equity investors have experienced rough seas over the past several years. The critical issue at this point is rationally interpreting — for ourselves and for clients — what we've actually experienced. Very simply, have we experienced wind shifts or a fundamental sea change? Our conclusion is very important inasmuch as it will govern our response.

First, some perspective on the past. The moving targets we call "annualized return" are indeed susceptible to gyrations in the markets, despite the fact that measuring performance over multi-year periods attempts to dampen out the bumps and dips of short term performance. The data in Figure 2 clearly illustrate this. In the short space of three years the moving 10 year average returns for three prominent indicies declined by as much as 49%.

Figure 3 presents the 1, 3, 5, and 10 year annualized returns as of year-end 2001 and year-end 2002. The impact of the equity market decline of 2002 (and in some cases including 2000 and 2001 as shown in Figure 1) produced some staggering changes in the 3 year and 5 year annualized average returns. For instance, the 3 year average annualized total return for the S&P 500 Index was -1.0% for the three year period ending in 2001. Advance one year. As of year-end 2002 the 3 year average was -14.6%. Over a 5 year period the same index went from 10.7% (year-end 2001) to -0.6% (year-end 2002). Similar eviscerations were observed in the S&P Midcap 400 index and Russell 2000 small cap index over the 3 and 5 year periods. The ten year averages were not as dramatically affected. This, of course, serves as a valuable reminder to utilize longer periods to generate expected returns going forward. However, as shown in Figure 2, even a 10 year period may not be a long enough time frame to obtain accurate long-term market bearings.

Now, let's move beyond the indexes and examine the aggregate performance of the U.S. equity market as of year-end 2002. One method for comparing, and revealing, the performance of individual stocks and equity mutual funds beyond what benchmark indices provide is presented in Figures 4 and 5. Only companies (i.e. stocks) which are U.S.-based and had at least 12 months of performance history were included in the analysis. The selection criteria for mutual funds eliminated those with less than 12 months of performance and those with more than 15% of their portfolios in cash, bonds, or non-U.S. stock. Redundant funds were also omitted (i.e. only "Distinct Portfolios" were selected). Data used in this analysis were extracted from Morningstar Principia Pro.

As shown in Figure 4, there were 6,004 stocks in existence for the full year between January 1, 2002 and December 31, 2002. The arithmetic 1 year mean return for this group of stocks was -11.5% and the median return was -14.4%. (For comparison, at year-end 2001, the average return for 6,197 stocks in existence during 2001 was 18.1%, with a median return of 3.5 percent). The average return in 2002 of -11.5% was better than any of the indexes reported in Figure 3 during 2002. During 2002, 63% of domestic stocks (in the Morningstar database) had a negative one-year return. Of those companies whose stock had a negative return in 2002, the average was -44.2%. The mean return of the companies whose stock had a positive return was 43.4%.

That nearly two-thirds of all U.S. stocks produced a negative return in 2002 is certainly dismal news. However, it is encouraging that 37% of all U.S. stocks had a positive return during 2002 — a year which according to major equity benchmarks was a total disaster. If taking our bearings from prominent equity indicies — which were uniformly hammered during 2002 - it would seem unlikely that so many firms would have generated a positive stock return, yet they did. And, as noted above, the average return for those companies that did have a positive return was an astonishing 43.4%.

All of this leads to a critically important realization: equity indicies tend to hide as much as they reveal. As a classic example of this, consider that during the three year period ending in 1999 (the roaring conclusion of the bull market) the S&P 500 Index had a three average annualized return of over 27%, yet during that same three year period (1997-1999) 46% of all U.S. stocks had a negative 3 year annualized return. Moreover, for the 5 year period ending in 1999 nearly one-third of all U.S. stocks had a negative annualized return, despite the S&P 500 reporting a 5 year annualized return of 28.5%. These findings suggest that using prominent indicies as our primary measure of the equity market can distort perception. Indeed, by doing so, a wind shift might be mistaken for a sea change, and vice versa.

Back to the data in Figure 4. The difference between the mean and median stock return over the one year period suggests an upward bias produced by stocks which had a one year return in excess of 100%. Inasmuch as a stock cannot lose more than 100% there is a limit on the size of a negative return. There is, however, no such limit on positive returns, hence there will generally be an upward bias in mean return compared to median return. This upward bias is particularly noticeable in shorter time frames (e.g. one year or less). In fact, there were 174 stocks which had a one year return in excess of 100% during 2002. The highest reported return among individual stocks during 2002 was over 4,200% and the lowest return was -99.96%.

A revealing statistic is the "Share-Weighted Annualized Return" reported in Figure 4. This figure is not generated by Morningstar. The share-weighted return (as calculated by the author) assigns a greater weight to stocks that have more shares outstanding and less weight to thinly held stocks. For instance, General Electric had the largest amount of shares outstanding within the Morningstar database as of December 31, 2002. In fact, of the 465 billion total shares outstanding, GE had 9.95 billion, or 2.14% of the total. Therefore, the 2002 return for GE (-37.7%) was multiplied by .0214. This same calculation procedure was applied to all 6,004 stocks, resulting in a share-weighted average return of -26.0%. Recall that the raw average one year return for all 6,004 stocks was -11.5%. The share-weighed return obviously suggests that the return of the widely held, large cap stocks was far worse than the average stock. Indeed, the average return of the largest 100 stocks (as measured by number of shares outstanding) in 2002 was a negative 27%. Those 100 largest companies (representing less than 2% of 6,004 companies) accounted for 43% of the total shares of stock outstanding.

The share-weighted return in 2002 is also close to the S&P 500's one year return of -22.1%. The S&P 500 Index is a market-cap weighted index, a technique that has the same basic effect as share-weighting the returns of individual stocks. Therefore, it is the return of the largest firms that most dramatically affects the overall return of the S&P 500.

Obviously, when the stock of widely held firms (e.g. General Electric, Cisco Systems, ExxonMobil, Intel, BICO, Pfizer, Oracle, Microsoft, Citigroup, Wal-Mart, AOL Time Warner, etc.) performs poorly, proportionally more investors are affected. The U.S. equity market is one in which a small percentage of companies account for a disproportionately large portion of total outstanding shares of stock. For this reason, the share-weighted return is likely to be a more accurate reflection of what is being experienced by most investors compared to the simple arithmetic mean return for all stocks. For this same reason, virtually all equity market indicies are market-cap weighted. However, as previously noted, indicies which are market-cap weighted (or, in this case, share-weighted) can create blind-spots with regard to how the broader market is actually performing.

The share-weighted return was only calculated for a one year period only inasmuch as the number of a companies shares outstanding at one point in time (e.g. 12/31/2002) cannot reliably be applied to performance data over a period beyond one year.

Over the five years ending December 31, 2002, the picture provided by the indices (Figure 3) and the aggregate return of all stocks (Figure 4) diverges. The three U.S. equity indices had 5 year annualized returns (as of 12/31/02) ranging from -1.4% to 6.4%, whereas the raw average return for the 4,767 stocks in existence over the entire five year period was -8.8%. The median return for all stocks over the same period was a dismal -4.0 percent. Moreover, 59% of all stocks had a negative 5 year average annualized return. Over 10 years, the indices paint very respectable annualized returns ranging from 7.2% to 12.0%, yet fully one-third of the 2,579 stocks which survived for the entire ten year period generated a negative 10 year annualized return as of December 31, 2002.

Similar data for U.S. equity mutual funds as of December 31, 2002 are presented in Figure 5. It is startling to note that nearly all funds (97%) had a negative one year return. Why did this happen? Mutual fund managers tend to own many of the same stocks (e.g. hot companies). When the winds shift and the popular stocks start taking on water, the impact is rapid and widespread. Such was clearly the case during the 2002. Moreover, the majority of U.S. equity mutual fund assets reside in large cap funds - and many large cap stocks suffered in 2002. As of 12/31/2002, 55% of U.S. equity mutual funds had a negative 5 year annualized return. By comparison, 59% of individual stocks had a negative 5 year return as of the same date. Thus, the meltdown of 2000-2002 was sweeping in its effects. The potential protection offered by a diversified portfolio (the mantra of mutual funds) was largely ineffective against the broad-based equity sell-off. (On a technical note, these results are obviously affected by survivorship bias. However, such a bias does not present a comparative advantage to either individual stocks or mutual funds when comparing funds vs stocks.)

Over the one year period ending in 2002, U.S.-based equity mutual funds had an average return of -22.8%. When weighting each fund's one-year return by it's percentage of total net assets (of the entire group of 2,391 distinct domestic equity funds) the asset-weighted return improved slightly to -21.4%. In other words, the biggest funds performed slightly better (i.e. less worse) in 2002 than average-sized funds (where the average net assets per fund was $560 million). This technique is comparable to the share-weighted return reported for individual stocks in Figure 4.

As a side note, there continues to be a vastly uneven distribution of assets among U.S. equity mutual funds. The largest 100 funds (just over 4% of the total sample of 2,391 U.S. equity funds) held $788 billion in assets, or nearly 59% of the $1.339 trillion total. (Recall that this analysis has removed redundant funds, hence the total asset figures do not represent the entire domestic fund universe). So, as is the case with individual stocks, when widely held funds falter, a disproportionately large number of investors are affected. Calculating the net asset weighted return is one way of demonstrating this.

As shown by maximum and minimum return data in Figure 4, extremes in performance (both up and down) are characteristic of individual stocks. Mutual funds (Figure 5) tend to perform within narrower performance parameters. Compared to individual stock, mutual funds will always have maximum returns which are lower and minimum returns which are higher over virtually any time period. The 5 year period is a clear example of this. The parameters of 5 year stock performance were from -87.3% to 174.6 percent, whereas the worst fund lost 30.1% on an annualized basis and the best fund had an annual return of 22.9 percent.

Similarly, as shown in italics in Figures 4 and 5, the average positive return of mutual funds is lower than the average positive return of individual securities. But the good news is that funds also tend to have average negative returns which are better (i.e., less negative) relative to individual stocks.

Over the ten year period ending December 31, 2002, the performance of individual stocks and funds began to coalesce. The median return for 2,579 individual stocks which survived the entire period was 6.1% and for the 740 equity fund survivors the median annualized return was 8.1%. But, as previously noted, 33% of individual stocks had a negative 10 year annualized return while only 2% of the equity funds dipped below zero. In fact, 85% of the 740 equity funds had a 10 year average return of greater than 5%. Not thrilling results, but by comparison, only 54% of the 2,579 individual stocks generated a 10 year annualized return in excess of 5%.

It will be helpful to remind clients that mutual funds are not the most exciting approach to investing. But, perhaps excitement is overrated! In terms of performance, funds will never compete with the most successful single issues of stock in the short or long run. However, over the longer run, the vast majority of surviving equity mutual funds produce a positive return — not as often the case among surviving individual stocks.

In summary, wind shifts are to be expected. Sea changes are not. What we have experienced lately in the equity markets amount to wind shifts. Oddly, far too many investors are behaving as if the sea has changed as they flee equity investments. Part of this behavior may be due to the incomplete bearings provided by prominent equity indicies. Because of the dependence upon market cap weighting, the U.S. equity market as a whole is generally not performing as well, or as badly, as the major benchmarks indicate. Keeping this simple idea in mind allows the helmsman (planner, client) to avoid over-steering the ship. "A little left…a little right" may be all that is needed. The focus should be on passage making — which necessarily involves some course corrections — rather than over-reacting to the constantly changing winds.

Figure 1. Contrary Equity Winds

Figure 2. The Impact of the "Down Years"

Figure 3. The Impact of 2002.

Figure 4. Individual U.S. Stocks

Figure 5. U.S. Equity Mutual Funds