I don’t like to be wrong, but I believe that I have made an error in one of my regular Washington Post On Leadership blog posts with respect to the central thing that bothered me most about Goldman’s behavior. I believed the initial reports that suggested Goldman let John Paulson pick the mortgage-backed securities for the now-infamous Abacus 2007-AC1 deal that he thought were going to plummet in value, which Goldman then sold to clients who were blissfully ignorant of Paulson’s involvement.

Instead it now appears clear that the principal buyer of the deal – ACA – was not in the dark.  In fact, ACA’s formal role was as the final selector of which securities were included in the deal.  It now appears that ACA’s principal for the Abacus deal, Laura Schwartz, had conversations with Paolo Pellegrini, Paulson’s deal principal, in which, among other things, she was informed that Paulson’s interest was in shorting the deal.

 It is a complicated and somewhat surreal deal, but I think about it as follows.  Imagine two players, P and A, in the following game of Fantasy Baseball.  Both sides have to agree on the composition of the team.  Both sides know that P is going to bet that the team will go on to lose and that A is going to bet that the team will go on to win. Each is free to suggest players.  The other can accept or reject players as they see fit.  Eventually they know they have to assemble a team, so they trade off with one another, though P always tries to get poor players included and A tries to get good players included.  When they are done, the fantasy team plays the virtual game and if the team loses, P gets a big payoff and if the team wins, A does.

In real life, Paulson was P, ACA was A, and Goldman was the moral equivalent of league commissioner. In the end, ACA was convinced that it had outsmarted Paulson in picking securities (i.e. players) that would perform well.  It turned out that Paulson was correct: they performed badly.

 So I no longer believe that Goldman attempted to deceive ACA and the other buyers.  ACA and others deceived themselves by believing that they were smarter than Paulson – and it turns out that they were terribly wrong.

 And so was I: my mistake.

The question remains as to whether Goldman and others should be spending their time creating bets like this.  The proponents argue that betting games like this help financial institutions manage their risk while detractors argue that it encourages financial institutions to take on excessive risks.  There will always be some level of speculation in every market.  For a trade to be made, the players have to have different points of view on what will transpire and make a bet on their view.  While some of the positions will actually be natural hedges (we have a big mortgage portfolio so we will short mortgages to create a risk-cancelling hedge), in no market can all be. So to the extent that we want to have capital markets, we need to have some level of speculation.  

 But as we now know from the insolvency of so many financial institutions in 2008-9 – which required massive government bailouts – the existing regulatory structure appears to have produced a lot of more excessive risk-taking than prudent risk management.  The challenge remains to figure out how to create a regulatory structure in which prudent risk management wins out over excessive risk-taking by our financial intermediaries and pension funds.