Last week I analyzed the Nikkei 225 monthly closes since 1970 to back-test several monthly moving average strategies versus buy-and-hold (see here). Now let's make a similar analysis of the S&P 500. The index was created in March 1957. But I'm also including the spliced S&P Composite monthly closes from January 1950, the starting point for my data source, Yahoo Finance. By including these earlier years, we get a more complete view of the post World War II bull market that began six months earlier in the summer of 1949. See this monthly update for more on secular bull and bear markets.
This log-scale chart highlights the four periods we'll focus on:
Limitations of this Approach
Like the Nikkei study, this is a purely hypothetical exercise since there was no mechanism for investing in the S&P 500 until the first Vanguard index fund was launched in 1976. In addition, the Yahoo Finance S&P 500 data excludes dividends. Similarly, I've excluded interest earnings when the SMA strategies signal a move to cash. Finally, this approach doesn't factor in trading costs, which will impact SMAs, especially the shorter ones with more buy/sell signals. Despite these limitations, the analysis should still provide a general idea of relative performance for passive buy-and-hold (B&H) versus active management with the various monthly SMAs.
Charting the Relative Performance
The bar charts that follow show the nominal value of one dollar invested using B&H and fourteen monthly SMA strategies. I've divided the 60-year timeframe into the four secular bull/bear markets illustrated above. The boundaries are determined by peaks and troughs based on monthly-close highs and lows and thus don't correspond to the specific intramonth dates. They are November 1968, July 1982, August 2000 and the most recent close. Thus we will examine two secular bull markets (1950-1968 and 1982-2000) and two secular bear markets (1968-1982 and 2000-?). The chart for the last cycle extends to the latest S&P data. But one day we conclude that the cycle really ended in March 2009.
Comparative Performance: 1950 to Present
The first chart shows us the big picture. Over the past 60 years, B&H outperforms two-thirds of the monthly SMAs, including the 10-month SMA, which is the system highlighted in the The Ivy Portfolio. However, the authors, Faber and Richardson, point out that the 10-month SMA, as illustrated by their data from 1973, significantly minimized the downside risk (psychological and financial) of bear-market declines.
January 1950 to November 1968
This chart covers most of the postwar boom that began in the summer of 1949, six months before the earliest data available from Yahoo Finance. However, since the index gained over 20% in the last half of 1949, a chart understates the returns for the entire post-war boom.
November 1968 to July 1982
The nearly 14-year span in the next chart saw the combined effect of consolidation from extreme overvaluation in 1968, when the cyclical P/E10 was in the top quintile of market valuations (see the P/E data here), and the relentless erosion caused by the decade of stagflation.
The B&H performance was just shy of break even, returning 99 cents on the dollar, excluding dividends. The monthly-SMAs outperformed B&H, especially the longer ones, and they dramatically reduced losses during the three cyclical bear markets during these 14 years (declines of 36.1%, 48.2%, amd 27.1% illustrated here).
Side note: That 2-month SMA is a curious outlier that worked well during the recessions of this period (see this chart).
July 1982 to August 2000
The most interesting chart in this study is the amazing 18-year Boomer bull market — probably the most favorable extended period in US history for passive investing.
I use the "Boomer" label because this period coincided with the financial coming-of-age of the Baby Boomer generation. The 401(k) plan was introduced in 1980 and IRA rules were liberalized the following year. The US saw the massive growth of a new investor class with a limited understanding of markets and risk. The bulge of Boomers became a windfall for Wall Street, the oldest having just turned 35 in 1981. Stagflation was replaced by the Great Moderation. We entered an era of optimism, investor confidence, and witnessed the emergence of the Boglehead philosophy of B&H passive investing with low-cost index funds. The crash of 1987 was a minor blip for the passive investor, but it gave a nasty whipsaw to many of the monthly SMAs — further insuring their underperformance and validating the strategy of passive investing.
August 2000 to the Present
The Tech Wreck of 2000 marked the beginning of a massive secular change in the market, one that was confirmed by the more recent Financial Crisis. In less than nine years, the S&P 500 suffered two staggering declines — 49.2% and 56.8% from peak to bottom. Compared to B&H, the monthly SMAs showed their advantage for risk management.
Like my study of the Nikkei SMAs this is another back-test of investment strategies in a hypothetical environment. It is based on authentic, but partial, data — monthly closes excluding dividends, interest on cash, and trading costs. Naturally it would be foolish to cherry-pick a single monthly SMA strategy based on these charts, intriguing as the results may be, especially since the results for the various timeframes demonstrate that no single strategy is a consistent winner.
What we can conclude is that in secular bull and bear markets, passive management (B&H) is a successful strategy on the way up, but it is a losing proposition on the way down. The reverse is true for active management with SMAs. It's unlikely to outperform B&H on the way up, but outperformance on the way down is a virtual guarantee.
Unfortunately it's impossible to pin-point those secular tops and bottoms and change strategy on a dime. More than a year after the March 2009 market lows, it's still impossible to say for certain that we're in a new secular bull market.