Tuesday, January 16, 2007

Debt, Derivatives and Interest Rate Conundrums
The rise of total debt in the US economy has been tremendous in the past few years, reaching 315% of GDP in 2006. Despite this high demand for borrowed funds and the sharp rise in Fed Fund rates, interest rates set by the free market (i.e. rates for borrowing longer term) have not gone up by much at all and the yield curve is inverted. Why? The answer, what Mr. Greenspan called a “conundrum”, may lie in derivatives - particularly credit default swaps. Let’s start from the beginning.

Any debt has two major risks: interest rate risk and default, or credit, risk. Interest rate risk is determined by the market – if rates rise and the debt carries a fixed interest rate then its price in the secondary market will drop until its yield to maturity becomes equivalent to current returns and vice versa. The rise and fall of prices are mathematically driven and therefore exact and objective. There are several ways to hedge this risk with futures, swaps, options, etc, but the process is also mathematically exact and not subject to interpretation. We'll ignore this risk for now.

Credit risk is a much more relative: it measures the ability of the borrower to repay principal and interest on a timely manner and has always been subjectively judged and rated (not always successfully or honestly, I might add). There are several credit rating and reporting agencies scoring business and individual borrowers based on cash flow, income, debt load, guarantees, collateral, etc. Credit risk is the main determinant of the interest rate a borrower has to pay. Enter financial engineering and Credit Default Swaps (CDS).

The massive power and ready availability of modern computers has allowed mathematical modeling to reach every nook and cranny of the finance business. A new “science” was born - financial engineering. Let’s stop for a moment to consider the oxymoron of that expression. Finance is certainly not an exact science, since it ultimately depends on people’s psychology and temperament - evanescent quantities, both. How, then, can you construct something solid on top of them, as the term “engineering” implies? No matter how much statistical analysis and back-testing you perform, people are not as predictable as steel beams. Yet investment banks, brokerages and trading firms are now full of “financial engineers” busy devising one model after another to fit what is, in the end, an impossibly chaotic process. Foolish? Yes and no.

Yes, because in the end reality always “outs”. Fundamentals govern the ultimate return of any investment, not expectations. You cannot make a silk purse out of a sow's ear and, no matter how much you may desire it, there is no philosopher’s stone.

No, because people are funny animals – if they believe something is true then they behave accordingly until proven wrong. Thus, belief is much more important than fact and nowhere is this more apparent than finance. Advertise there is gold in the bottom of Lake Erie and you will find people willing to invest millions to finance the construction of mining submarines. You can make a pile from taking advantage of “irrational exuberance”.

The most dangerous form of gullibility is that which is self-induced, i.e. “believing your own sh*t”. To wit, if your model shows that time and again (in the past 20 years, anyway) sub-prime borrowers were better credit risks on average than was originally calculated, well then “you have a market” and can make billions lending at lower rates than before. Presto, an industry is born. Never mind that the data pool was tiny or that low interest rates and plentiful cash lowered defaults only temporarily – the “model” showed an opportunity. Lenders rushed in – and how! The creation of CDS’s was extremely rapid. In the chart below from the Bank of International Settlements (BIS) you can see the notional amounts outstanding in CDS quadrupling to $20 trillion in just 18 months!

What does a CDS do? It supposedly uses the “market” to more accurately score credit risk, instead of the traditional credit agency methods. It also spreads out the risk amongst more participants who buy and sell such default protection. In theory it lowers overall portfolio risk and thus lowers average borrowing costs and/or allows for the lending of more money at a given cost.

Bingo! The next chart is also from BIS. It shows the phenomenal increase in issuance of Collateralized Debt Obligations and the performance of related indexes. (The recent price of the ABX.HE.06-02 BBB- index has collapsed to 93.41 and that of BBB to 95.31).

The conundrum is apparently solved: the explosion of credit derivatives has allowed interest rates to stay low even as debt burden expanded very rapidly. But keep in mind that the default risk has not gone away - it has only been spread amongst many more participants (including your very own pension fund, no doubt). Is that a benefit? Up to a point, yes. But beyond it lies this very simple fact: there is now much more risk shared among many more entities.

And that is not good engineering...