We may well be going through a 50-year storm in the financial markets.  Lots of folks are talking about unprecedented events in the housing market and about the risks of financial innovation such as Credit Default Swaps.  With so much change it may well seem like all the old rules are out, but at the heart of Lehman’s demise is a much much older killer - leverage.  Leverage kills by vastly amplifying risk. It does so in a non-linear fashion which makes it really hard for folks to appreciate just how much more risk they have added.

Imagine you have 100 dollars invested and the asset takes a 20 percent hit.  You lose 20 bucks.  With 1x leverage you would borrow an extra 100 dollars and the same 20 percent decline would cost you 40 bucks of equity.  With 2x leverage you would borrow 200 dollars and the 20 percent decline would cost you 60 bucks.  Looks linear so far.  But that is only the loss itself.  Let’s look at what is happening to your leverage post loss.  If you had no leverage, you still have no leverage.  At 1x you wind up with 80 of equity but still have 100 of debt so your leverage went up to 100/80 =  1.25x.  And at 2x starting leverage you wind up with 200 of debt and only 40 of equity so your resulting leverage is 200/40 = 5x.

In fact at only 4x leverage, a 20 percent loss leaves you with no equity and *infinite* leverage!  Your lenders who provided the leverage are all going to want their money back instantly. That means your business cannot absorb any other shock. You can literally be knocked over with a feather. Yet on the upside 4x leverage only gives you 5x returns, i.e. if the asset goes up 20% you make 100 dollars (instead of 20 dollars without leverage).  So as you push up leverage you increase your catastrophic downside risk much more rapidly than your upside.  Now just to put the above example in perspective, Lehman was operating not with 4x but with 30x leverage!

At some point, I will write a post as to why Wall Street was not able to learn this lesson after the failure of Long Term Capital.