Two bond market memes are currently permeating financial-market discussions: First, that U.S. interest rates are at historically low levels; and second, that bear markets in bonds are always fast and severe. The first is only partially true; the second is pure fantasy, in our view. The U.S. experienced two major bear bond markets in the 20th century, both of which lasted for multiple decades and were characterized by interest rates gradually rising from very low levels.
These eras are instructive for investors who, like us, expect to face a similar period in the coming years. The lesson for building a portfolio in such an environment comes down to one word: balance. First, though, a historical perspective: The first major bear market in U.S. bonds lasted from 1899 to 1920; the second from 1951 to 1981.
Although there are important differences between the 1951-1981 period and the current state of the bond market, some of the similarities are uncanny. Rates rose gradually during the first 20 years of this era before spiking higher in the 1970s as inflation accelerated. Currently, we anticipate gradually rising rates and modest inflation for the foreseeable future but recognize there are possible risks to this forecast. The Federal Reserve also expanded its securities holdings by 1,100 percent in the decade leading up to 1951, a number that significantly trumps the 315 percent increase over the last decade on a percentage basis.
Analyzing the differences between those periods provides a number of significant fact-based (as opposed to fantasy-based) conclusions about historical major bear bond markets. First, optimal portfolios remained well diversified between equities and high quality bonds. Second, presuming an ability to adjust bond portfolio duration, overall portfolio efficiency remained relatively stable across various levels of equity exposure. Third, conservative investors were better off targeting slightly shorter bond portfolio durations than aggressive investors. Fourth, portfolios lost efficiency as equity allocations rose above 75 percent. And finally, investors should have strategically over-weighted small-cap equity relative to large-cap equity.
One significant conclusion from this is that investors were well-served by maintaining reasonable bond allocations, even if they were limited to long-duration bonds. However, as expected, portfolios with very large long-term government and corporate bond allocations exhibit lower risk-adjusted returns (Sharpe ratios) than portfolios with 35 percent or more in equity. Perhaps surprisingly, portfolios with greater than 75 percent allocated
to equities also exhibited a significant drop in efficiency.
Once the constraint of only holding long-maturity bonds is relaxed, and bond portfolio duration becomes dynamic, the efficient frontier flattens out and portfolio efficiency remains relatively stable across all risk profiles. This analysis suggests that more-conservative investors should target a lower duration (about 3.7) than more-aggressive investors (about 6.3). Why? Conservative investors generally hold relatively small equity allocations and relatively large bond allocations, so reducing duration in the bond allocation helps maintain overall volatility at reasonable levels without concentrating portfolio risk.
Next, equity allocations above 75 percent were self-defeating in the period under analysis. Larger allocations resulted in additional risk, but lower return, on average. Although we do not want to over-interpret this historical data in regards to setting guidelines, it is important to note that an all-equity portfolio was not an optimal portfolio for even very aggressive investors during the last major bear bond market.
Finally, their higher risk-adjusted returns and slightly lower correlation to government bonds resulted in an overweight for small-cap equity relative to large-cap equity. We believe this serves as a reminder to construct well balanced equity portfolios. In general, these facts of bond bear markets argue for maintaining well-diversified multi asset class portfolios and appropriately reducing bond portfolio duration instead of dramatically reducing the size of bond portfolios.
Although the popular misconceptions around bond bear markets might imply much more dramatic action, they are not supported at all by historical experience. In the context of a more expanded set of assets classes than is available for this historical analysis, we have incorporated these conclusions into our strategic advice as well as our current tactical positioning by generally reducing government bond exposure and overall bond portfolio duration in favor of other less interest rate-sensitive asset classes. These include meaningful allocations to high yield credit, emerging market debt, developed market equity, emerging market equity, hedge fund strategies, private equity, and private real estate.