Don't Believe in Buy and Hold
A. Gary Shilling
Even during rip-roaring bull markets, investors miss a good chunk of the gains. Buy and hold RIP.
The recent 57% collapse in the S&P 500 index to its most recent low on March 9 followed hard upon the 38% decline in the 2000-2002 bear market. And the five-year recovery from that swoon didn't exceed the 2000 peak by much. The Nasdaq index, which nosedived 78% in 2000-2002, recovered only 44% of that decline before falling another 55%. There are only two other global bear markets since 1900 in which stocks fell over 40%.
No wonder that investors' faith in stocks has been shattered, and both institutional and individual investors have been withdrawing. The buy and hold strategy, which was validated by the earlier long, steadily rising market, doesn't work in severe bear markets. Only one of 1,700 diversified U.S. stock funds showed a gain in 2008, and that was a mere 0.4%. The average of these funds dropped 39%, precisely in line with the S&P 500's decline.
The buy and hold devotees say you can't time the market, and if you aren't in all the time, you risk missing much of the gain. A Spanish research firm found that if you removed the 10 best days for the Dow Jones industrial average in the 1900-2008 years, two-thirds of the cumulative gains were lost. But if you missed the 10 worst days, it found, the actual gain on the Dow tripled. These results are in line with our earlier research and reflect the fact that stocks fall a lot faster than they rise.
We eschew the buy and hold strategy because of what's known in classical statistics as the gambler's ruin paradox. The odds may be in your favor in the long run--in this case, your stocks may provide great returns over, say, 10 years. But if you hit a streak of bad luck, your capital may be exhausted before that long run arrives.
Or more likely, a severe bear market will scare you out at the bottom. Many investors bail out then and don't reenter until the next bull market is well advanced. This explains why the returns of mutual fund investors lag well behind the performance of the funds in which they invest. A widespread retreat is what makes a good bottom, as we've noted in many past Insights. All those who can be shaken out are. They've reached the puke point at which they regurgitate their last equities and swear to never ingest any more.
We've never understood the U.S. individual investors' fascination with stocks, almost to the exclusion of all other investment vehicles. Stock backers point to long-run annual gains of about 10% but neglect to note that about half of that came from dividends, which were much bigger parts of the total return in earlier years, although they may be again in the future.
Also, stock indexes are revised over time, dropping weak and fading companies and replacing them with robust and growing firms. So the performance of the Dow or S&P 500 over time is much stronger than the performance of the companies that were in those indexes, say, 30 years ago. This is known as survivor-bias.
Even with this upward bias, stocks way underperformed Treasury bonds in the 1980s and 1990s in what was the longest and strongest stock bull market on record. The superiority of Treasuries has been even more so since then. One reason that few realize this is because they don't know much about bonds, despite the simplicity of Treasury obligations and, so, they ignore them. Furthermore, commissions on stocks are usually much bigger than on Treasuries, so brokers favor them.
Our all-time favorite graph shows the results from investing $100 in a 25-year zero-coupon Treasury bond at its yield high (and price low) in October 1981, and rolling it into another 25-year Treasury annually to maintain that 25-year maturity. On March 31, 2009, that $100 was worth $16,656 with a compound annual return of 20.4%. In contrast, $100 invested in the S&P 500 at its low in July 1982 was worth $1,502 last month for a 10.7% annual return including dividend reinvestment. So Treasuries outperformed stocks by 11.1 times!
Long-time Insight readers know we have been recommending long Treasury bonds since 1981. Back then, we forecast secular and huge declines in inflation and interest rates. So we declared that "we're entering the bond rally of a lifetime." Unfortunately, that rally is over. Our target of 3% yield on 30-year Treasuries, down from 14.7% in 1981, was exceeded at the end of 2008 when the yield fell to 2.6%. Nevertheless, it was a grand finale to "the bond rally of a lifetime." The yield drop from 4.5% at the end of 2007 provided a 37.5% appreciation. Add in the 4.5% interest and the total return was 42% last year.
In the long run, the stock market rises with GDP, after accounting for intermediate trends in profits' share of GDP and P/Es. In the next decade, we foresee much slower growth in GDP than in the 1980s and 1990s and deflation, with profits' share of the pie falling along with declining P/Es. In this secular bear market, stock market average gains will probably be much lower with cyclical bull markets shorter and weaker, while bear markets are more frequent and deeper.
Even during rip-roaring bull markets, investors miss a good chunk of the gains. Buy and hold RIP.
The recent 57% collapse in the S&P 500 index to its most recent low on March 9 followed hard upon the 38% decline in the 2000-2002 bear market. And the five-year recovery from that swoon didn't exceed the 2000 peak by much. The Nasdaq index, which nosedived 78% in 2000-2002, recovered only 44% of that decline before falling another 55%. There are only two other global bear markets since 1900 in which stocks fell over 40%.
No wonder that investors' faith in stocks has been shattered, and both institutional and individual investors have been withdrawing. The buy and hold strategy, which was validated by the earlier long, steadily rising market, doesn't work in severe bear markets. Only one of 1,700 diversified U.S. stock funds showed a gain in 2008, and that was a mere 0.4%. The average of these funds dropped 39%, precisely in line with the S&P 500's decline.
The buy and hold devotees say you can't time the market, and if you aren't in all the time, you risk missing much of the gain. A Spanish research firm found that if you removed the 10 best days for the Dow Jones industrial average in the 1900-2008 years, two-thirds of the cumulative gains were lost. But if you missed the 10 worst days, it found, the actual gain on the Dow tripled. These results are in line with our earlier research and reflect the fact that stocks fall a lot faster than they rise.
We eschew the buy and hold strategy because of what's known in classical statistics as the gambler's ruin paradox. The odds may be in your favor in the long run--in this case, your stocks may provide great returns over, say, 10 years. But if you hit a streak of bad luck, your capital may be exhausted before that long run arrives.
Or more likely, a severe bear market will scare you out at the bottom. Many investors bail out then and don't reenter until the next bull market is well advanced. This explains why the returns of mutual fund investors lag well behind the performance of the funds in which they invest. A widespread retreat is what makes a good bottom, as we've noted in many past Insights. All those who can be shaken out are. They've reached the puke point at which they regurgitate their last equities and swear to never ingest any more.
We've never understood the U.S. individual investors' fascination with stocks, almost to the exclusion of all other investment vehicles. Stock backers point to long-run annual gains of about 10% but neglect to note that about half of that came from dividends, which were much bigger parts of the total return in earlier years, although they may be again in the future.
Also, stock indexes are revised over time, dropping weak and fading companies and replacing them with robust and growing firms. So the performance of the Dow or S&P 500 over time is much stronger than the performance of the companies that were in those indexes, say, 30 years ago. This is known as survivor-bias.
Even with this upward bias, stocks way underperformed Treasury bonds in the 1980s and 1990s in what was the longest and strongest stock bull market on record. The superiority of Treasuries has been even more so since then. One reason that few realize this is because they don't know much about bonds, despite the simplicity of Treasury obligations and, so, they ignore them. Furthermore, commissions on stocks are usually much bigger than on Treasuries, so brokers favor them.
Our all-time favorite graph shows the results from investing $100 in a 25-year zero-coupon Treasury bond at its yield high (and price low) in October 1981, and rolling it into another 25-year Treasury annually to maintain that 25-year maturity. On March 31, 2009, that $100 was worth $16,656 with a compound annual return of 20.4%. In contrast, $100 invested in the S&P 500 at its low in July 1982 was worth $1,502 last month for a 10.7% annual return including dividend reinvestment. So Treasuries outperformed stocks by 11.1 times!
Long-time Insight readers know we have been recommending long Treasury bonds since 1981. Back then, we forecast secular and huge declines in inflation and interest rates. So we declared that "we're entering the bond rally of a lifetime." Unfortunately, that rally is over. Our target of 3% yield on 30-year Treasuries, down from 14.7% in 1981, was exceeded at the end of 2008 when the yield fell to 2.6%. Nevertheless, it was a grand finale to "the bond rally of a lifetime." The yield drop from 4.5% at the end of 2007 provided a 37.5% appreciation. Add in the 4.5% interest and the total return was 42% last year.
In the long run, the stock market rises with GDP, after accounting for intermediate trends in profits' share of GDP and P/Es. In the next decade, we foresee much slower growth in GDP than in the 1980s and 1990s and deflation, with profits' share of the pie falling along with declining P/Es. In this secular bear market, stock market average gains will probably be much lower with cyclical bull markets shorter and weaker, while bear markets are more frequent and deeper.
Comments