If You Can't Explain It, You Lose
April 11, 2011
Complexity can be invigorating, as you try to tease out what's happening, get a grip on it and explain it.
But there are some areas that beggar understanding.
Such a product, for me, is the collateralized debt obligation.
This is, of course, a synthetic bond. A bond or series of bonds based on a lot of bonds.
One entity buys a pool of bonds, such as mortgages, slices it into tranches and sets rules on payouts to investors. The collateral? The bonds in the pool.
What is hard to follow is why, in the first place, you would want to create a synthetic bond like this. Very often, simple is better. Create one bond. Lend it out to one party. Make sure you collect. You both have skinny in the game. If someone doesn't pay out, you know who to collect from. And you know what's going to happen to your bottom line, if you don't.
But, if you are going to play the scale game and try to spin out returns to investors from large pools of bonds, you best know exactly what's going on.
Which is why BlackRock set out to model collateralized mortgage obligations in great detail, before it started investing in them on behalf of its clients.
Part of its model, more than a decade ago, was to pull in payment information from federal mortgage institutions such as Freddie Mac and Fannie Mae. That was so that the company could know what exactly was being paid off. Or not.
So when a lot of attention got paid to Lehman Brothers, Bear Stearns, American International Group, et al., at the height of the credit crisis precipitated by falling housing prices and failures to pay off mortgages, not much attention was paid to BlackRock.
Here is its operating income for the last five fiscal years: $471.8 million in 2006, before the crisis. Nearly $1.3 billion in 2007 and $1.6 billion as the crisis emerged. A drop to $1.3 billion (again) in 2009. And a surge to $3.0 billion in 2010.
That's a pretty good record, for a company that built its risk analysis and investment management platform, known as Aladdin, on the premise that it should understand-as completely as possible-the financial instruments in which the company would be investing.
As you'll see in reading "BlackRock Gets Data to Give Models Might" on page one, if you can't model it, you can't explain it. If you can't explain it, you won't understand it.
And if you don't understand it, you're going to lose your shirt.