Bonds have had a good run for 40 years, but it's unlikely they'll maintain their pace relative to stocks.
Given the poor performance of stocks over the past year and decade, there has been ample discussion about the relative performance of stocks versus bonds. Some even argue that investors should allocate entirely to bonds, not only because bonds are the safer investments, but because they believe bonds will outperform stocks over the long run. In other words, if bonds can deliver higher returns with less risk, why bother with stocks?
Stocks Versus Bonds, the Past
By looking at the returns over the past 40 years (shown in Exhibit 1), the argument that bonds might outperform stocks looks to be valid. Not only have the average annual stock returns been poor over the past 10 years, but relative to bonds, stock returns look mediocre over the past 20, 30, and even 40 years. (View the related graphic here.)
But advisors should view this appearance with skepticism. First, note that over the 20-, 30-, and 40-year periods, stocks actually performed quite well, even if some bond categories did better. For the very long term, it is no longer a contest (Exhibits 2 and 3). One dollar invested in stocks 83 years ago easily outgrows $1 invested in bonds. Over almost two centuries, the returns on the stock market have been consistently high and roughly in line with stock returns over the past 40 years. (View the related graphic here.)
Long-term history provides two major insights:
1) Stocks have outperformed bonds.
2) Stock returns are far more volatile than bond returns, thus more risky. Given the additional amount of risk, it is not surprising that stocks don't out- perform bonds every period--even over extended periods of time.
Stocks Versus Bonds, the Future
How likely are stocks to outperform bonds going forward? To try to figure out the future, let us look in more detail at what happened during the past 40 years.
Despite the substantial decline in yields over the past 40 years, the bulk of the bond returns come from the income return portion, or yield (Exhibit 4). On average, the bond income return from coupon payments was more than 7%. Capital gains caused by the yield decline made up the additional return. (View the related graphic here.)
Today, yields are much lower (Exhibit 5). As of the end of the second quarter of 2009, the yield of the long-term government bond was 4.3%, and the intermediate-term government bond yield was only 2.51%. For bonds to continue to enjoy the same amount of capital gains over the next 40 years, a rough estimation would put the yield into negative territory, especially for intermediate-term government bonds. This is simply impossible, because it implies that investors would be willing to lend their money to a borrower and pay the borrower an interest rate. Over the past 40 years, bond investors have enjoyed abundant returns because of a high-yield environment followed by a steady decline in yields. (View the related graphic here.)
To analyze which asset class is more likely to outperform going forward, let's take a deeper look at the historical data and the current market environment. We analyze each component of returns going forward for stocks and bonds as follows:
current yield + capital gain
current yield + earning growth + P/E changes
First, given the current low-yield environment, it would be almost impossible for bonds to generate the same amount of capital gains as they did in the past. In fact, a reasonable estimate might be that there will be no more capital gains going forward, since yields may be at least as likely to rise as to fall. If there were no future fall in yields, all of the return would have to come from the coupon return.
That means the total returns for bond investments would likely be between 3% to 4%.
For stocks, the annualized dividend yield from January to July 2009 for the S&P 500 was 2.59%. If stocks produce more than 2% in capital gains per year on average, they will likely beat bonds. Stocks' capital gains can be decomposed into nominal earnings growth and changes in the P/E ratio. Historically, U.S. long-term nominal earnings growth has been roughly 5%, which is comparable to the nominal GDP growth. We will assume the market valuation level (operating P/E of S&P 500) will stay the same over the next 40 years.
We would then have an equity return of around 7%.
Even if we forecast a decline in the valuation level, the 10-year average P/E level would need to fall from just about 20 to below five to get equity returns around 3%.
Bonds not only have outperformed stocks by a large margin over the past year because of the financial crisis, but they also roughly matched stocks over the past 40 years. Will bonds continue to outperform?
Upon closer examination, we show that stock returns over the past 40 years were virtually in line with the long-term historical average. On the other hand, bond returns were not only much higher than their historical averages, but also higher than their current yields. This high bond return is due to higher interest rates in the 1970s and a subsequent declining interest-rate environment. This scenario for bonds is very unlikely to repeat in the future, given today's low-interest rate environment. Investors hoping that bonds will outperform in the coming years will likely be disappointed.
Stocks tend to outperform bonds over time but are much more risky, even over longer periods. Bonds can outperform stocks over a long period, but investors need almost perfect timing to get in and out of the market to realize such returns. We believe the right strategy is to follow a disciplined asset-allocation policy that considers the return and risk trade-offs by taking advantage of the diversification benefits between stocks and bonds over time.
As Warren Buffett wrote in his 2009 annual shareholder letter: "When the financial history of this decade is written, it will surely speak of the Internet bubble of the late 1990s and the housing bubble of the early 2000s. But the U.S. Treasury bond bubble of late 2008 may be regarded as almost equally extraordinary."