Down markets in equities have the interesting property of testing investors' appetite for risk.
In long-term bull markets, investors only have academic knowledge that equity markets can produce deep and occasionally prolonged periods of absolute underperformance. Reassured by their bands of advisers that such periods will be brief and equities always outperform in the long run, investors suffer from the illusion they will retain their long-term confidence in the asset class.
Once the downdraft hits -- such as the one that occurred last autumn -- the reality of the difference between volatility risk and uncertainty sinks in, and longer-term memories of huge equity market debacles come to the forefront of the limbic system. To illustrate, before the emerging-market crisis began, investors focused on the upper half of this decision tree. As the crisis unfolded the full range of possible outcomes was explored, and a bias toward the lower half of outcomes emerged.
Subsequently the pendulum has swung again to the more Utopian scenario.
However, it is in these very volatile environments that investors become more conscious of the very real merits of bonds as an asset class. These merits can be summarized in two categories: the positive case for global bonds; and a more realistic view of the true range of outcomes in global equities.
There are frequently neglected aspects of bonds, which make them a far more attractive asset class than conventional wisdom assumes:
* Global bond markets have been less correlated with "extreme" market conditions than global equity markets, and hence offer better risk diversification when investors really need it. Over the past 10 years, in "normal" markets represented by the mid-70% of monthly moves, global bond markets are some 30% correlated, against a global equity market correlation of 25%. In the top and bottom 15% of market moves, however, the bond market correlation rises to 40%, while the equity market correlation shoots up to nearly 70%.
* There is extensive research supporting the case that local bond markets are more predictable than equities. Hence, they offer more reliable returns to active management. If the determinants of bond market value, as opposed to equity market value, are considered, this is not too surprising. Bond value depends on credit and inflation expectations. Equity value also depends on credit and inflation, but in addition is strongly influenced by earnings growth and a much more unstable risk premium. The number of hard-to-predict variables is much larger.
* Credit considerations are usually ignored in comparing equity index returns with bond index returns. If we compared an index of global corporate bonds, which would yield somewhere between 50 and 100 basis points above the government indexes, the returns would be closer to those of the Morgan Stanley Capital International global equity index.
* Finally, if we take the category of bonds that has the most similar "duration" to equities -- namely, long-duration bonds -- they also offer more competitive returns to equities, especially on a risk-adjusted basis. Longer dated bonds, with their higher average yields, offer substantial returns, but are considerably less volatile than equities on average. This lower volatility also implies their returns are much less concentrated, which significantly reduces timing risk.
More realistic view
Apart from looking at the more positive side of bonds, a more realistic view of the full range of possible returns from equities should also change the degree of bias toward equities:
* There is huge "survivor bias" in the long run equity return studies used to support the increasingly heavy strategic overweighting in equities. They are based on the United Kingdom and the United States, and ignore that many other equity markets have been wiped out in the same time periods. As Laurence Seigal points out in his paper, "Are Stocks Risky? Two Lessons" (Journal of Portfolio Management, Spring 1997), the promising equity growth universe in 1900 included Germany, Russia, Japan and Argentina (to say nothing of Hungary and the then Czechoslovakia).
These were the markets devastated over the subsequent decades by war, sequestration and hyperinflation. Survival bias exists at the market level as well as the stock level. Of course, this bias is also present in bonds, but because bonds have a lower duration than that implied in equities, bonds are likely to be favored within this comparison, as are default recovery ratios.
* The corollary of this analysis is that we only have a very limited sample of long-term equity markets. There are multiple possible scenarios that could reasonably have played out during the past 50 to 100 years. We cannot infer that equities always will outperform just because they have done so in two samples from the surviving markets.
* Periods of major real underperformance in equity markets can be much longer term than most analysis suggests. When bonds are considered, the focus is on real returns rather than nominal. Equities also are capable of delivering significant long-term real negative returns.
* Equity market returns are highly concentrated. If you eliminate the top 5% of monthly returns in equities you will have lost 40% to 50% of your returns. Timing risk is therefore acute in equities, yet there might well be very real reasons injections into equity markets cannot always be smooth and consistent. Nor does this allow for the natural fear and uncertainty associated with down markets or the euphoria of up markets. It is the rare institutional investor who holds his or her nerve and buys into the most depressed markets after a major collapse.
* The final assessment of the case for bonds also includes behavior. There has been a growing degree of confidence in equities relative to bonds among investors and the pension fund industry in general in the past few decades. This implies that equities are again likely to be strategically overvalued relative to bonds, given the strong recovery from last autumn's events. Indeed, the historic excess returns on equities that have created this confidence cannot be explained satisfactorily within the framework of economic theory.
The equity risk premium is dependent on particular environments and conditions change. The past in economics and finance is not a stationary, independent sample period. Historic returns are time specific, and the laws governing them are insufficiently immutable to be sure they will be repeated. As Peter Bernstein noted, Benjamin Graham wrote in 1952 that the most powerful case for stocks was as "an essentially undervalued security form" (extract from Peter L. Bernstein Inc., Oct. 1, 1995), because investors were failing to understand a portfolio of stocks significantly diversified risk. Moreover, memories of the 1930s and the impact of war's turmoil on stocks were at the forefront of investors' minds in the early post-war period.
The post-war period, which has produced the exceptionally high-risk premium for equities, encompassed a time period when there existed a wide range of capital, foreign exchange and other regulatory controls. These permitted the buildup of inflationary forces that distorted the relative return on most assets: equities, bonds, property and commodities. In particular it created a strong anti-bond bias among investors. It is dangerous to assume these experiences will continue to dominate the investor psychology of the future.
We are in a different mode today. The diversification benefits of equities are well understood, and faith that the capitalist system will deliver ever-higher earnings is extremely strong. The implicit time horizon of equity investors is therefore stretching further into the future.
Bonds, on the other hand, while having been substantially revalued since the traumatic experience of the 1970s, are still not back to the long-term average long-term real yields. As 1994 demonstrated, and as we are again being reminded in the first half of 1999, the memory of the inflationary 1970s is still lurking in today's investor psychology. However, the disinflationary environment of the past two decades is gradually changing investor perceptions, and the risk of deflation, rather than inflation, has been rising. This could well produce a bond premium puzzle for future investors.
All this is not to deny equities as a hugely important element in portfolios. It is to remind us that bonds, and global bonds in particular, also have an important and frequently neglected role. A reassessment of that role -- both in terms of its scale and composition -- is appropriate.
Christine Downton is chief investment officer at Pareto Partners, London.