Sunday, 23 February 2014

Why Economic Capital?

This post arose out of a conversation with a good friend and fellow consultant a while back. We were discussing the case of an insurer not subject to a market-consistent regulatory regime, and considering whether and why it would be necessary for such a firm to use an economic capital framework in their risk management function.

Our conclusion was that yes, it is desirable and necessary to consider your risk on an economic basis, no matter the regulatory or other rules your firm is subject to. Our reasoning relies on the proposition that the risk of running out of cash is the most fundamental of risks faced by a financial institution, something that I do not think is emphasised enough in the Solvency 2 regulations. I set out our argument below.

Let's start by defining an economic capital model. I'll use this phrase to describe any sort of risk model which uses “realistic” or market values of assets and liabilities. I'll also assume that the purpose of this model is to evaluate the probability of insolvency.

Now consider the converse argument: suppose that an institution operates in a regime where (for example) assets are recorded at book value and liabilities at the value of projected cashflows discounted at a single off-market discount rate. This was exactly the situation in all of Europe not too long ago, and is the current situation in some less developed economies. In this case the institution is insolvent if and only if the book value of assets is below the discounted value of liability cashflows. Economic values seem irrelevant.

But this statement isn't quite correct. Although the firm would then certainly be insolvent on a balance sheet basis, there is another situation in which it could be out of business: if it runs out of cash to pay its creditors, including policyholders or depositors. Cash is the ultimate constraint on any business, and liquidity is the most basic risk. It doesn't matter if the regulator claims that you are solvent, or the accountants swear that you are a going concern, if there is no money in your bank account to make payments when they are due, you are bust. Let's call these two crisis scenarios “balance sheet insolvency” and “liquidity insolvency”

So clearly this is an important risk, and equally clearly there is no ambiguity about cash, there is only one way to account for it. “Economic value”, “realistic value”, “market value” and similar phrases are all tautologies for cash. On the other hand for a life insurance company, with its long-dated liabilities and easily realisable investments, the risk of running out of cash might seem very remote. But here is the crux of the argument: an economic capital model is nothing more than a quantification of the future risk of liquidity insolvency. Consider: ignore the balance sheet values of assets and liabilities. To establish your current and future solvency you would want to project forward the cashflows from the assets and liabilities in all possible future states of the world, and track the value of the bank account. Provided this projected value is always positive, you are solvent with certainty. If the bank account value goes to zero in more scenarios than is acceptable (maybe 0.5% compounded over the number of years of your projection), then you have a problem. But now if you discount all the projected cashflows at some appropriate rate, and make an appropriate allowance for the risk associated with each future state of the world, what you end up with is nothing other than the economic value of the assets and liabilities.

So we can take a short cut in the calculation: rather than projecting forward the cashflows of assets and liabilities, discounting, weighting for risk and averaging, we can just look at the economic values. We now make the heroic assumption that market values are economic values, and that market-consistent values are likewise, and simply draw up an economic or realistic balance sheet to establish our current solvency.

And now we are almost done: having drawn up the economic balance sheet, we can subject that balance sheet to stresses in some way (instantaneous stress tests, one year of Monte Carlo scenarios generated by some model, take your pick) and check that the behaviour of the realistic balance sheet under stress is acceptable.

To summarise our line of thought:
  • Cash is the ultimate constraint on the business
  • Irrespective of the accounting and regulatory value of a firm's balance sheet, it needs cash available to make payments as they fall due
  • Economic values are discounted probability- and risk-weighted values of future cashflows
  • So using an economic balance sheet allows us to circumvent the need to project all cashflows
  • Hence even in an off-market regulatory or accounting regime, the economic balance sheet should be the basis for prudent risk assessment
Now there are a couple of assumptions and limitations I should point out here. First of all there is equivalence of what I've called economic values (being expected values of discounted, risk weighted future cashflows) and market or market-consistent values. Secondly note that we lose some information when using economic values rather than projecting cashflows and the bank account. Being solvent on an economic balance sheet basis doesn't mean that the firm is always going to be solvent on a cashflow basis, it just means that it will on the balance of probability. A certain liability and a risky asset could be of equal value, provided the asset is worth sufficiently more than the liability. But this is the reason we don't just look at the current value of the economic balance sheet, we look at stressed scenarios.

As a final note I'd state that I believe that liquidity issues don't get the attention they deserve in the life insurance sector. This is a big contrast with banking, where Basel III has put liquidity risk at the centre of risk management. Many corporate bankruptcies are down to liquidity issues, and in particular many high-profile financial failures have been liquidity related, often through collateral mechanisms. Think of IG Metallgesellschaft, Long Term Capital Management, Bear Stearns. The interaction of collateral and liquidity risk is quite interesting itself, I may blog about that topic in more detail at one point. With insurers increasingly using derivatives to manage their risks the issue of liquidity is becoming more and more important. All economic risks are simply proxies for, or transformations of this most fundamental of risks to the financial firm.

No comments:

Post a Comment