I was in Amsterdam last week, inspecting volatility of canal prices (more later) but volatility was all over the place last week. On Monday, the PRA strenuously defended matching adjustment in Court with the idea that price movements are simply ‘short term’.
It may also assist the Court to know that the PRA is supportive of the principles underlying the MA, not only because (properly implemented) it more appropriately reflects the risks to which annuity providers are exposed but also because it enables firms to “look through” short term volatility in the market price of credit risk to which they (as buy-to-hold investors) are not exposed.
At the same time, our old friend golden.labrador@dogs.k9.gov mailed us to point out that if a firm had sold its bond portfolio at the height of the crisis (see chart above) when spreads had exploded, it would have been considerably worse off than if it had stuck to a buy and hold strategy, as matching adjustment allows you to do.
But where does this take us?
It takes us to another old friend: Harry Hindsight. Clearly, if you had kept your position when spreads had exploded to 700 basis points, or if you had bought around then, you would have done very well. Buying low and selling high is always profitable. But how do we translate that into some sort of algorithm for guaranteed profit? It’s a job for hedge funds, not central banks.
Suppose there is a surefire way to tell when spreads are artificially high, a trading signal to tell us when to buy assets (or to sell them when spreads are artificially low). The Bank of England could set up a hedge fund using this trading signal, and attract investors. Even better than a hedge fund, it could capitalise the guaranteed profits in advance and pay them out. Who needs to wait?
Even if that is possible, how does that justify the false accounting involved in matching adjustment? It gives all the wrong price signals to investors. Suppose bond prices collapse from 100 to 70, with liabilities marked down by the same amount. Then the book value of capital remains unchanged, so investors who buy the lifeco at book value are cheated of the gains they might have made if they had been able to buy at 30 lower. Contrariwise, speculators could short the lifeco stock and hedge by buying bonds at 70, to make a surefire profit. If the market recovers, they make 30 on the bonds, and lose nothing on the short position. If the lifeco collapses, they make 30 on the short position. It makes no sense to falsify a company’s accounts in order to protect the capital position set by the market.
You may object that the regulators will close the company to new business if the capital position is set by the fickle market. But why should the regulator have to do that, if volatility is truly short term? Capital is there to absorb losses caused by market movements. If all market movements are short term, and have no impact on the long term position at maturity, there is no point in a regulatory capital requirement at all, for there are no true losses to absorb.
Thus the problems of Solvency II are largely self-created. We create a capital requirement based on a regime designed for banks. Finding that almost no insurance firm can meet the requirement, we invent capital to meet the requirement. Did anyone ask whether that made sense?