This article begins a series dedicated to anticipating possible future financial crises, based on a special attention to financial correlation risk. The article has been published in the private sector for members of the EFPA website (European Financial Planning Association) and we’ve made it public in this blog.
What is financial correlation risk?
The correlation risk measures the possibility of future financial losses as a result of adverse changes in the correlation between assets (financial or otherwise). An example of financial correlation risks is the negative correlation between the interest rates and the price of commodities (gold, silver, oil, coffee, corn, wheat, etc.). If the interest rates go up, they produce losses on commodities. Another example of financial correlation risk is what happened during the Greek financial crisis of 2012. The positive correlation between the Mexican and Greek bonds caused very strong losses for Mexican bond holders. Non-financial assets are also exposed to correlation risk. For example, the levels of sovereign debt and currency value can cause economic losses for the exporters of a country. In 2012, North American exporters experienced losses due to the devaluation of the Euro. In the same way a low Gross Domestic Product (GDP) has an impact on the European or Asian exports due the high concentration of them in the United States.
The inadequate management of the variations in correlation between assets was, as we will see, the root cause of the real estate and financial crisis of 2007-2009. Before delving into what, in my humble opinion, will be a proposal of analysis to perhaps anticipate and be able to prepare ourselves for future crises or black swans of this type, I will describe an example of how the variations in the correlation between assets impacts the losses that are produced.
The ideal functionality
Let us suppose that an investor buys an Italian sovereign bond and is worried about the risk of Italian default. The investor could cover or offset this risk by buying a Credit Default Swap (CDS) insurance on the default risk of Italy. That way he transfers the risk of payment suspension from the bond issuer to a third-party insurance or CDS seller. How would that situation work in the event that the bond issuer actually did suspend payments (default)?
- The investor buys a 1.000.000 Euro bond from an Italian issuer.
- The Investor buys CDS insurance from an issuer (Deutsche Bank).
- The Investor receives the Italian bond voucher and pays the fixed spread of the CDS.
- The bond goes into default and the investor doesn’t receive the principle or outstanding vouchers.
- The insurance or CDS seller (Deutsche Bank) pays 1.000.000 Euros to the investor so that he recovers his initial investment.
- The value of the vouchers (fixed or floating) and the price of the insurance (fixed) should be managed in order to have a well calculated return on investment.
So far, we would be an example of a structured or complicated investment that happens at a highly sophisticated level of professional investors (not all). It’s not very common that personal investors come to achieve this type of coverage.
But what risk do we avoid?
We assume that if the Italian bond goes bankrupt, we won’t recuperate any of the initial investment. The investor is protected against the bankruptcy of the Italian bond through their purchase of the CDS insurance because they will receive 1.000.000 euros from Deutsche Bank if the bond goes bankrupt. The price that you pay for the insurance is of the fixed type (spreads) of the CDS which is calculated using the probability of default of the reference asset (Italy) and the correlation of default between Italy and Deutsche Bank, which is to say, what relation exists between what is happening in Italy and how that affects Deutsche Bank itself. In the event that the correlation between Italy and Deutsche Bank increases positively, that is, that the probability of the suspension of Italian payments and the possibility of Deutsche Bank going bankrupt are equal, positive and perfection correlation (correlation = 1) would cause the investor to lose everything, since he would not receive his initial investment on the Italian bond because of the Italian suspension of payments and wouldn’t receive the million because of the bankruptcy of Deutsche Bank.
What indicator tells me if the increases and is positive? One of the key indicators is the price of the insurance, that is, the spread of the CDS. The higher the correlation (greater risk of joint bankruptcy) the smaller the spread. In the case of positive correlation (Deutsche Bank and Italy), the investor finds themselves in the worst-case scenario, called WWR (Wrong Way Risk).
The Real Estate and Financial Crisis of 2007-2009
In the real estate market do mortgages and hybrid CDOs adequately control the correlational risk of default between quality and sub-prime mortgages? As we will see in the following post, the mistake of not controlling the variation of the default correlation matrixes between junk mortgages– or sub-prime mortgages– and the higher quality mortgages began the downward spiral that we’re all familiar with now. So, what should we fix our attention on for the immediate future? The evolution between the correlation of the stock market internally and externally with other debt markets, the correlation between volatilities and above all, the correlation of default of the issuers will be key to anticipating possible black swans in the system.