John Lombardo submits:
Tune into any financial news network and you are likely to hear that the first decade of the millennium was a lost decade for US equity investors. Most investors and investment professionals likely sit back, nodding their heads in agreement.
Of course they are right, aren't they? Didn't the NASDAQ Composite and the S&P 500 produce negative annualized returns over the decade while the return on the Dow Jones Industrial Average trailed inflation? Hadn't the S&P 500 lost nearly 10% cumulatively over the period while NASDAQ suffered a cumulative loss of nearly 45% (charts 1 & 2)?
click for enlarged images
Yet the performance of these indices fails to provide an accurate picture of US equity market performance over the last decade. First, the tech-laden NASDAQ Composite Index was disproportionately impacted by the technology bubble. This largely accounts for its horrendous underperformance relative to the S&P 500 and the Dow Jones Industrial Average. Additionally, NASDAQ includes ADRs (American Depository Receipts), which represent units of foreign equities traded in the US, and therefore does not exclusively reflect the performance of US equities.
The Dow Jones Industrial Average has significantly outperformed the other two benchmarks and may well be the best known of the three indices. Yet it is comprised of merely 30 stocks and experienced eight constituent changes between April 2004 and June 2009. Such high turnover in a modest number of holdings makes it suspect as a gauge of overall market performance.
Additionally, the index is price, rather than capitalization, weighted. Both the S&P 500 and the NASDAQ Composite are capitalization-weighted, meaning that the performance of stocks with larger market capitalizations has a greater impact on the performance of the overall index than those with smaller capitalizations.
In a price-weighted index, higher priced stocks have a greater impact on performance than lower priced stocks. While it is hard to make a case for price-weighting, the case for cap weighting is relatively straightforward: companies with larger market capitalizations comprise a larger portion of the investable universe.
On the other hand, the S&P 500 is broadly diversified across economic sectors, consists of a reasonably large number of stocks, and encompasses roughly 75% of US equity market capitalization. As such, one might consider its performance a reasonable approximation of US equity performance in its entirety. This has not consistently been the case and over the last 12 years the S&P 500 has in fact been an extremely poor approximation of domestic equity performance.
When companies with the largest capitalizations perform significantly better or worse than the average stock, the index provides a distorted picture of US equity performance. To shed light on this issue, we compared the performance of the capitalization-weighted S&P 500 Index with an equally-weighted version of the benchmark. In the latter methodology, each stock in the index has an equal bearing on performance, providing an indication of the average performance of the 500 stocks in the index. We compared data for the two weighting methodologies from 1990 through 2009.
From 1990 through 1997, the performance of the cap-weighted and equally-weighted S&P 500 indices differed by at least 394bps in each calendar year except 1994, when the cap-weighted benchmark outperformed by 37bps. Despite the significant calendar-year differences, the annualized returns over the period were very similar, with the cap-weighted index outperforming by less than 100bps.
However, over the last two years of the decade, as the technology bubble built, the largest stocks in the benchmark dramatically outperformed. The cap-weighted index gained more than 24% annually, more than double the annual return on the equally-weighted index. As a result, the cap-weighted index significantly outperformed over the course of the decade (charts 3 & 4).
Clearly the cap-weighted index was not a good indicator of broad equity performance over the last two years of the 1990s bull market, but it has been an even worse indicator over the last decade. The equally-weighted index outperformed the cap-weighted index in seven of 10 years excluding its 1bp outperformance in 2006. In four of those years it outperformed the cap-weighted index by at least 11.5 percentage points.
Over the first seven years of the decade, the equally-weighted benchmark outperformed the cap-weighted benchmark by eight percentage points annually. After lagging moderately in 2007 and 2008, the equally-weighted index outperformed the cap-weighted index by nearly 20 percentage points in 2009. Over the first decade of the millennium the equally-weighted index outperformed the cap-weighted index by more than six percentage points annually (charts 5 & 6).
Of course, the S&P 500 is a proxy for the US large cap equity market, but equally weighting it makes it less so. Because the smaller capitalization stocks within the index have just as much impact on performance as the largest capitalization stocks, the equally-weighted performance is more representative of the performance of mid-cap stocks.
Indeed, the performance of the equally-weighted S&P 500 from 2000 through 2009 is very similar to that of the S&P 400 Mid Cap Index. While the Russell 1000 Index has a cap range very similar to that of the S&P 500, its median market cap is roughly half of the S&P 500. The equally-weighted Russell 1000 Index actually provided a reasonably attractive annualized return of 7.95% over the last decade (chart 7). Inflation averaged approximately 2.16%, making the real return on the equally weighted Russell 5.79%; below historical averages, but far from a wash out.
The last decade has been enormously challenging for the US equity markets. To suggest otherwise would be foolish. Real returns have been subpar and increased volatility has resulted in a more significant reduction in risk-adjusted returns. Yet the decade has not been the unmitigated disaster that broadcast journalists would have you think nor as bad as cap-weighted indices would suggest.
Additionally, we have not addressed the performance of the US small cap market in this analysis. While equally-weighting small cap portfolios poses greater challenges due to liquidity issues, we note that the Russell 2000 has produced positive results on both a cap- and market-weighted basis.
Should a stock carry a larger weight in a portfolio simply because it has a larger market capitalization? For that matter, should a stock be held at all simply because it carries one of the largest capitalizations in the market. The answer to both is a resounding no! At the end of 1999, many of the largest companies by market capitalization were vastly overvalued.
As a result, they have languished over the last decade. Investors purchasing good companies at inflated prices do so at their own peril. Those who bought Intel ( INTC ) and Cisco ( CSCO ) at their peaks in early 2000 may need to live to be as old as Methuselah to achieve a reasonable return on their investment. On the other hand, after a decade of underperformance, many of the largest companies by market-cap may now be undervalued. If so, being equally-weighted will not be as beneficial over the next decade. As such, the choice is not as simple as whether to equal or cap-weight a portfolio.
Proponents of passive management have long argued that active managers cannot consistently outperform cap-weighted market indices. However, as the evidence has continued to mount against the precept, we have seen the inception of equally-weighted and even fundamentally-weighted passive strategies.
Investors now have multiple passive approaches, with significantly varying performance, from which to choose. Yet such a choice requires an active decision. While continuing to espouse passive strategies, proponents of efficient markets theory are themselves changing their approach to meet the reality that cap-weighted indices are not particularly efficient nor remotely unbeatable.
Disclosure: No positions
See also 2011: The Year of Trading Dangerously on seekingalpha.com