Correlation measures are meant to provide an idea of the relationship between two variables. It answers the question as to how one variable relates to another. Do they tend to move in the same direction? Do they move in the opposite direction or are they actually unrelated to one another? Correlation figures can also be misleading because they tend to change over time. A negative correlation between say two variables could suddenly turn highly positive or vice versa. So correlations can be useful in devising a portfolio but one needs to be aware of this major caveat. There are also other, more subtle factors that should be considered when studying correlations. Take "Cat Bonds" as an example, which are insurance linked bonds that basically pay out in the event of a specific natural disaster such as the earthquake/tsunami that recently ravaged Japan. Looking at the correlation between this particular sub asset class and the stock and general bond markets, even over long periods, one gets the impression that they are not related (see graph). In fact, their very nature should ensure that for most events there should be no positive correlation, but like for any investment, this rule can break down when "extraordinary" events occur. The Japanese quake was extraordinary in the sense that it was not only the largest experienced in Japan but also amongst the top five largest in the world for more than one hundred years. This should therefore be considered as an "outlier" event, something of extremely rare occurrence but which can provoke the sort of correlations that one is trying to avoid in the first place.
Here is an outline of the sequence of events that would help explain the sudden sharp rise in correlations between Cat bonds and the equity markets: The Japanese earthquake/tsunami devastated a large part of the north of the country, causing an estimated $25 billion in damages to be covered by insurers. This is on the back of a series of other natural disasters (albeit at a smaller scale) since the beginning of the year, which have depleting the reserves that insurance companies had set aside for such events. On the stock market front, the Japanese disaster comes on the back a number of other events that have the potential of derailing the global economic recovery. These include the ongoing European sovereign debt crisis that seem to be worsening, turmoil in the Middle East that show no signs of ebbing and commodity prices that continue to rise. In the context of such a backdrop it is no surprise that markets would respond with a correction following the damage caused to the third largest economy in the world. Still, these are extraordinary events, never before observed in such a configuration and are unlikely to repeat themselves. Nevertheless, it is important to note that no matter how rare an event may be, it still has the potential of occuring and by not accounting for it is understimating the risks involved.
