James Saft, Reuters Columnist
That’s something to think about, given that international diversification has long been sold to investors as the next best thing to a free lunch.
The finding, in a paper published in June by Thijs Markwat, a Dutch quantitative researcher at Robeco Asset Management, should send many scrambling to re-think their asset allocations.
Markwat defines a global crash as a week in which the four major equity markets – Europe, the United States, Asia and Latin America – all turn in losses that would be in the bottom 5% in terms of performance. At minimum, that would be the equivalent of a 5.9% fall in global stocks over a week.
While the chances of that happening in 1992 were only 0.1% per week, by February 2010 it had risen to 1.5%, theoretically implying a global crash on average about every two years.
“The main conclusion for risk-managers reads that geographical diversification opportunities are almost monotonically decreasing, as the global crash probabilities keep rising. To keep the risk of portfolios at acceptable levels, risk-managers should think of other ways, besides geographical diversification, to diversify their exposures,” Markwat wrote.
While it is obvious to anyone paying attention that equities are riskier than many assumed 20 years ago, this is a stark illustration of exactly how volatile – and highly correlated – international equities have become.
My guess is that as the reality of this volatility and correlation becomes widely understood, investors will demand higher risk premia for shares. In other words, global stocks as an asset class may well have further to fall.
The idea that global diversification could act as a hedge gained in popularity in the 1970s and ’80s as investors observed that major markets often tracked one another only loosely, offering in essence cheap insurance for a portfolio.
While correlations between the U.S. and European markets were only 0.51 in 1992, by 2010 that had risen to 0.83. The link between U.S. and Asian shares rose during the period from just 0.41 to 0.64. That implies there was still significant diversification value, though that value is shrinking over time.
Globalisation
The Asian crisis of 1997-98 seems to have driven a permanent increase in correlations. Having reported on financial markets at the time, that makes sense to me.
When Thailand started to fall apart, the prevailing wisdom was that this was totally irrelevant to U.S. investors, a misconception many continued to hold right up until the point of the Long-Term Capital Management failure.
That event ushered in the boom, bust and bailout pattern we are still following today.
Investors began to understand, slowly, that global equity markets are reasonably tightly linked. This drove appetite for financial news, and also led, inevitably, to more attempts by investors to get out in front of a rout in one place by selling up in another, even if they themselves had no exposure in the original market to begin selling off.
So, what came first, the chicken or the egg? While surely this is a fundamental-driven phenomenon, a reflection of the globalisation of the economy, it is also in part driven by the behaviour of investors in reaction to that globalisation.
The bigger question for investors is: What should they do now?
Clearly, there is still value in diversification, even taking into account currency risk, as lots of diversification can be gained across asset classes. Just take a look at returns on U.S. 10-year Treasuries against the S&P 500 for proof.
There is also, plainly, only very imperfect correlation within fixed income classes and regions, implying real value to an internationally diversified fixed-income portfolio.
The same forces that are driving interest rates up in Spain –- and fixed income returns down -– are helping to drive bond prices in Denmark and Germany strongly higher, leading to negative interest rates in many cases.
While neither end of that trade may seem appetising, imagine for a minute you are an Italian saver: you will be heartily glad to be getting a tiny coupon on your German or U.S. bonds when compared with the losses, real and potential, on your domestic bond holdings.
Correlations in equity markets are also, very probably, going to continue rising. The major shocks, like the Asian crisis, tend to have a big and lasting impact on correlations. A euro crisis that turns truly horrible would surely be big enough to prompt a global sell-off in shares, and to drive correlations significantly higher.
A pessimist will see this as driving equities lower. An optimist will realise that markets usually overshoot and, eventually, equities will be a fantastic buy.
Keep your powder dry.
(At the time of publication James Saft did not own any direct investments in securities mentioned in this article. He may be an owner indirectly as an investor in a fund. You can email him at jamessaft@jamessaft.com )