Portfolio diversification is a fundamental pillar of modern financial management theory, but that theory dates from another era and has lost its relevance for today’s investors. Diversification, as we still understand it today, is a “myth”, and we need to revisit it, asserts a recent T. Rowe Price study.
In a majority of instances, diversity works the least when you need it the most says the T. Rowe Price report titled When Diversity Fails,and written by Sébastien Page, head of Global Multi-Asset and Robert Panariello, portfolio manager and quantitative analyst.
In reality, the situation is even more confusing based on monthly data collected between 1970 and 2017. Correlation strengthens in times of market stress and weakens in times of market elation; that’s the exact opposite of what an investor wants. Referring to the 2008 financial crisis, the authors write that “not only did correlations increase on the downside, but they also significantly decreased on the upside. This asymmetry is the opposite of what investors want. Indeed, who wants diversification on the upside?”
Indeed. When markets go up, any investor would like everything to go up and see correlations increase, for example, between corporate bonds and small-cap stocks, or between emerging market bonds and large-cap stocks. On the other hand, when markets fall, that investor would like to see correlations weaken: large-cap stocks resist the fall of small-caps or even compensate for it.
Unfortunately, none of the above holds. Match stocks to corporate bonds or real estate, or North-American stocks to emerging market ones, small-cap to large cap, correlations are very tight when markets drop, settling anywhere between .55 and .92 (1.0 indicates a perfect correlation).
When markets go up, correlations dissolve. In the best cases, for example, between value stocks and growth stocks, they stand near .4; the same reading holds between U.S. stocks and emerging markets stocks. These are among the best correlations one can find in the study when times are good.
The same general observations hold for the most common hedge fund strategies and for correlations between real estate and private capital. They even hold for risk factors like “value”, “size” or “momentum”. For example, event-driven, macro and relative value hedge fund strategies exhibit very high correlations to U.S. stocks in down markets, ranging between .7 and .8. In up markets, correlations are almost null near .05 and .1. Macro and relative value strategies even exhibit negative correlations when times are good: nearly -.4 in both cases.
The picture is a bit more convincing in the area of risk factors, but correlation readings there are not overwhelmingly convincing. For example, the currency carry and the size (small versus big) factors show relatively high correlations (.5 to .75) in down markets, and weak ones in up markets. But the cross-asset momentum strategy has a very attractive correlation profile in down as well as in up markets (.2 versus .8). Currency momentum shows a very strong downside resistance of 0.85, but a weak upside throttle of -0.12.
“Risk factors are not inherently superior building blocks,” warn the authors. “They deliver better diversification than traditional asset classes simply because they allow short positions and often encompass a broader universe of assets.”
Finally, confirming traditional evidence, the authors find that only two types of assets present optimal diversification and correlation : stocks and government treasury bonds. When markets move up, correlations between these two asset categories tighten, then move apart when markets fall. However, note the authors, correlations between the two asset classes have tended to tighten in more recent years.
The study’s observations find an echo among market specialists with a less conventional bent. For example, they are well grounded, finds Richard Guay, finance professor at ESG UQAM, in Montreal, and an ex-president of the Caisse de depot et placement du Québec. If you look only at market extremes, he says, “the study says that not much diversification holds there, and it is totally right”.
Guy Mineault, an economist and retired professor of Université Laval, in Québec, makes similar claims. “I’ve been saying for years that diversification, as we have learned it, doesn’t hold any more,” he asserts. Thought up in the early 1980s, the notion of “diversification” proposed to cut up the allocation of a portfolio between many geographic regions. “It made sense because stock markets in those days were not correlated. Today, they are.”
But the study is too pessimistic, thinks Richard Guay, “though it catches up toward the end, he adds. We can still find a diversification effect, but it is much more limited than what many investors think.”
None of the specialists interviewed consider that diversification is dead. Today, investors must look for it in new areas and new assets mixes. Our understanding of diversification needs to be... diversified. But that’s a whole new ball game to be explored in another story.
