From the May 2006 issue of Wealth Manager Web • Subscribe!

May 1, 2006

Buy and H-o-o-l-d

Index funds and passive management are thought by some to be one and the same, but in practice, indexing has proven itself far less passive than is generally assumed. Index-fund owners may be poorer as a result.

The world's most popular benchmark of equities, the S&P 500, is not passively managed, courtesy of the evolving list of constituent firms as chosen by the Standard & Poor's index committee. That's par for the course in the business of indexing, of course, but it comes at a price.

Or so advises a new study penned by Professor Jeremy Siegel of the Wharton School and Jeremy Schwartz, senior analyst at Wisdom Tree Investments Inc. in New York. The buy-and-hold returns generated by the original 500 companies in the S&P 500 from March 1957 through December 2003 outperformed the continually updated S&P 500, they assert in "Long-Term Returns on the Original S&P 500 Companies," which appears in this year's January/February issue of the Financial Analysts Journal.

Siegel is no stranger to sifting over historical equity returns in search of enduring trends. His best-selling Stocks for the Long Run (McGraw-Hill, 2002), now in its third edition, reviews the track record for the U.S. equity market since 1802. The new FAJ paper's focus spans a mere half century, but the results may be more controversial. The Siegel and Schwartz study contradicts a previous study, thereby raising fresh doubts about the management of the S&P 500 and other benchmarks run with a broadminded approach to additions and deletions.

When it comes to the analysis, there are various assumptions that can be made about what's meant by "original." Spinoffs, distributions, privatizations and so on continually harass the otherwise simple concept of buy and hold. Indeed, an average of 20 new companies a year were added to the S&P 500 between 1957 and 2003, the FAJ paper reports. The S&P 500 is many things, but unchanging isn't one of them.

Considering a representative sample of the buy-and-hold possibilities, the paper forms three theoretical portfolios in 1957, each governed by different rules for treating the changes in the original list of 500 companies. In one respect, it doesn't really matter, since all three portfolios outperform the actual index by an annualized 46 to 55 basis points for the market-value- weighted portfolios (which is the weighting system for the S&P 500). What's more, the three portfolios post higher Sharpe ratios than the actual benchmark, indicating superior risk-adjusted results.

If nothing else, the Siegel and Schwartz study lends more ammunition to the argument that index providers should redouble their efforts to minimize turnover. What seems like the enlightened tweaking of an index incurs risk of picking the wrong stocks. Yes, in the real world of indices, some changes are warranted. But how much is too much? Whatever the answer, a good rule of thumb seems to be that less really is better.

In a recent interview with Wealth Manager, Siegel avers that his latest study highlights the fact that owning the hot growth stocks is no shortcut for winning the performance race. "One of the biggest misconceptions among investors is thinking of growth as return," he says. Nonetheless, cutting loose the fading industrial company in favor of the thriving technology firm has found favor in more than a few equity benchmarks in recent years. But as the following conversation with Siegel reminds, there's reason to wonder if second guessing the market is costing index investors money in the long run.

Your study finds that buying and holding the original S&P 500 companies from 1957 through 2003 would have outperformed the actual S&P 500 index. What's the lesson here?

There are several implications. One is a challenge to the notion that updating indices is absolutely essential for owning a well-performing portfolio. That line of thinking says that if you never update, you'll end up with old companies and poor returns. But that's not true.

The belief that the only way to make money is by holding stocks in sectors that are expanding isn't true, either. Some of the shrinking-sector companies in the S&P 500 have done well. A lot of it has to do with the fact that the shares in shrinking sectors were priced very reasonably. People knew that some of the companies in the index didn't have strong growth prospects. So, investors priced the shares down; if the companies still managed to turn out profits, shareholders got a good return.

Your study will surprise some people, in part because it contradicts a previous study.

That was the Foster and Kaplan study [Creative Destruction (Random House, 2001)]. But the authors weren't economists and they made a serious, erroneous set of calculations when measuring long-term returns. Their methodology was seriously flawed, and they now admit that. They just looked at the market value of the companies, then and now, without taking into account dividends, etc. One reason that no one took them to task was because everyone said they're probably right, and it sounds reasonable that the original companies wouldn't do as well, and that new companies are necessary. But we studied the results and found out that the original S&P 500 companies actually outperformed.

In fact, our results question the common notion among investors that you've got to own growth, that you've got to keep updating portfolios otherwise your returns will fall behind. It's often the case that the new firms aren't really providing the best returns.

3. What drove the returns of the original S&P 500 companies?

Many of the original companies remained low in price, but paid a good dividend, and they were value-oriented firms. When you calculate the returns with reinvested dividends, you find that the original companies tended to outperform.

Meanwhile, as time went on, and the S&P 500 became more popular, stocks added to the index tended to jump in price and were overvalued. But you don't want to continually buy stocks that are high in price and that everyone else wants to buy. Rather, you want stocks that are good values. It turns out that a lot of the stocks that people were enthusiastic about had fundamentals that didn't justify the enthusiasm. In nine of the 10 industries in the S&P 500, the new stocks added subsequent to 1957 underperformed the original stocks that were initially in the index.

What's the best performer among the surviving stocks?

By far, it's Philip Morris Corp., now called Altria Co. It had a return of approximately 20 percent a year, or nearly double the S&P 500's return over the period studied, 1957 to 2003.

What are some of the outstanding performing sectors over those years?

There were two very good performing sectors. One is consumer staples. Also, healthcare was a good performer. Not so much among the healthcare providers, but the pharmaceuticals have done well. Despite their problems today, the pharmas have very good compounded long-term returns over the last 50 years.

The third outstanding sector, which was a contracting one in terms of relative value, is energy. Keep in mind that returns are measured through 2003, before the tremendous surge in energy of the last two years. The energy sector outperformed the S&P 500 from 1957 through 2003, despite the fact that energy was a declining fraction of the market over that span. That was a surprise. We didn't expect to see the oils doing well, but they did extraordinarily well. They were churning out cash, paying good dividends, and their total returns proved to be quite outstanding.

How have sector weightings changed over time in the S&P 500?

Technology, health care and financials collectively were less than 5 percent of the market in 1957, and today they're about 50 percent of the market. Meanwhile, materials and energy together were about 50 percent of the market in 1957, and today they're about 12 percent.

The message seems to be that big-picture economic trends don't necessarily dictate winners and losers for stocks.

Right. One of the chapters of my latest book--Future for Investors (Crown Business, 2005)--is called "Growth Is Not Return." One of the biggest misconceptions among investors is thinking of growth as return. In the long run, a company doesn't have to grow, but as long it turns out profits, and the return on the shares is reasonably priced, you'll get an excellent return that's often better relative to a growth company. The way you structure your portfolio in the long run is very different from what you'd do in the short run.

Your new FAJ paper suggests that investors are better off buying the current S&P 500 companies and simply holding them as opposed to owning the S&P 500 index. Do you agree?

I'm not sure. One of the problems of holding stocks over time without rebalancing is that you become less diversified. On the other hand, holding value stocks will deliver relatively stronger performance in the long run--the evidence shows that. That's also a lesson understood by the great investors, like Warren Buffett: look for gems in stagnant or shrinking sectors because they're overlooked. Take a look at the type of stocks Buffett buys today, or over the last 40 years. It's not in the hot areas. Rather, it's where he sees value, in areas that are often overlooked, where people are paying reasonable prices, and in firms where he likes the management. Overall, he prefers a commitment to shareholder return and he shuns the hot sectors.

But everyone chases the hot stuff in the expanding industries, which are overpriced. For building a good investment portfolio, you don't care if an industry is expanding or not. Rather, you want to see if a company's turning out profits. A stock's price, its returns and dividends, are the most important factors.

9. What does your study say about S&P 500 index funds, and other investable equity indices?

Indexing has done well relative to active money managers, but the study shows that the way indices are formed means they might have a bias that's not favorable toward investors. One question is whether market-value weighting is the best way to go. We don't address the question in the paper, although our work finds that you don't always have to be in the next hot sector and that you don't always have to update the portfolio. Rather, you can find some real gems in the neglected sectors, in shrinking areas that can ultimately outperform a dynamically updated index such as the S&P 500.

Criticism of market-cap weighted benchmarks like the S&P 500 seems to be on the rise these days.

When the bubble burst in 2000-2002, if you owned a market cap index you held all those technology stocks and so you had to ride the bubble up and down. One begins to wonder if there's a better way to structure a diversified portfolio that's not subject to all the volatility.

A recently launched exchange traded fund--(FTSE RAFI US 1000 (NYSE: PRF)--asks and answers the question, offering what its sponsor claims is a better way to index stocks than by market capitalization. Have you heard of the fund, which weights stocks by a number of fundamental criteria, such as book value and cash flow, rather than just market cap?

I'm very familiar with it. I like fundamentals indexing. I think there's a long way to go with the idea, with many products to come.

Does that ETF move closer to exploiting the effect you document?


But isn't the market-cap portfolio still the best default portfolio? Modern portfolio theory suggests it is.

The market-value portfolio has certain optimality features under certain assumptions of an efficient market. But if some of the assumptions of the efficient market break down, market-weighted portfolios are no longer optimal. That's one of the points of a fundamentally weighted portfolio.

Professor Bill Sharpe, who was crucial in the development of the intellectual foundation that supports the use of market-cap weights, says that market capitalization ultimately conveys important information--information that shouldn't be dismissed lightly by investors.

That's the position of the efficient-market supporters. But there's a lot of debate. Was the market giving the right signals in March 2000, when the market-cap-weighted indices were valuing companies without profits at hundreds of billions of dollars? That's the question. The S&P 500 has been extremely successful as a market-cap index. But the question is, will the new fundamental indices chip away at market-cap-weighted indices over time?

Should everyone holding an S&P 500 fund sell and look for something better?

Perhaps we should start thinking about fundamentally weighted or some non-market weighted indices. In the end, I call the FAJ paper suggestive. It's contrary to the expectations that people have that you always need to update to enjoy index returns.

In other words, it's possible to beat the market and not necessarily engage in trading to do so.


Reprints Discuss this story
This is where the comments go.