Wednesday, March 25, 2009

1000 times!

Hey, so xkcd is with me about the whole AIG bonus order of magnitude thing. Good company to be in.

Monday, March 16, 2009


Okay, so apparently folks are upset about how AIG is using its bailout money. AIG of course sponsors Manchester United. And Manchester United just lost badly to rival side Liverpool, despite being at home and strongly favored to win. It seems however that the current cause for concern is elsewhere. Namely, AIG has 1) spent a lot of money on paying back big banks that no one likes and 2) spent a lot of money on paying back executives... that no one likes.

So, by the numbers:
  • 173 billion => total bailout to AIG
  • 93 billion => total amount of the AIG bailout spent on big banks
  • 165 million => total amount of AIG bailout allotted for executive payouts
  • ~400 => number of executives among whom that 165 million is spread
  • ~750 billion => rough amount of Obama's bailout bill, passed last month
  • ~500 billion => rough amount spent on the military, annually, in peace time
  • ~600 billion => current cost of the war in Iraq.
Let's ignore the 93 billion in debt payments that have gone to Goldman Sachs, Barklay's, etc. The issue that accounted for *two* front page headlines in the business section of my paper today was the 165 million in bonuses. Why might someone object to that?

Well, one might be really bad at math and fail to realize that while 165 mil is a pretty big number, it's still not even 1/1000th of the overall AIG bailout. In terms of arithmetic significance, we're literally in the realm of rounding error here.

One might also be of the mind that no one, or at least no one who hasn't cured cancer or perfected nuclear fusion, deserves that much money. Such a person should probably prefer that AIG and friends be allowed to go bankrupt, since corporations doing hundreds of billions in annual revenue are inevitably going to make someone rich at some point.

Perhaps one might be motivated by simple vengeance. These executives crashed the economy. If anything, they should be punished not rewarded.

I'm not too worried about any of that though. I'm much more concerned about how AIG is deploying the other 99.9% of the money we gave them, and about why we gave it to them in the first place.

Wealth is something that an economy generates without theoretical limit, not a fixed quantity that can only be distributed in various ways. Therefore, I don't worry if someone has more money than I've got, so long as my lot is improving too.

Vengeance for its own sake seems silly too. It's a clumsy way of dealing with what I actually care about, which is moral hazard and incentives.

Before we get into incentives though, a bit more on these bonuses. The bonuses being paid out are contractually obligated "retention bonuses," that were negotiated prior to the current unpleasantness. Retention bonuses, though often abused, are in theory intended to keep high talent executives from jumping jobs.

The theory here is that even "huge" bonuses are justified if they can attract and retain top talent. In a decent year (i.e., not 2008), a company like AIG is making about 10 billion a year on 100 billion or so in revenue. If the best candidate is going to be even 5% more successful than the second-best, spending a fraction of a percent of revenue on bonuses makes sense. The huge corporations out there all have boards that know this, so the competition for top-flight talent is fierce, similar to a free agent market in sports.

Now, AIG is spending a lot to get the best talent, but are they actually getting the best talent? There's really no good way to know. It's unlikely that all 400 or so executives under consideration were directly responsible for massive losses, but it's quite likely that some of them were. Are we to believe that all available alternative leadership would have lost even more money? How could we ever know? What we do know is that it's relative. An executive has about as much chance of making big profits in an economy like this one as an NBA player has of shutting down LeBron James. The coach is pretty much happy if his defender causes LeBron to score less than usual, and that's pretty much how executives are judged too.

Here's what we do know. Executives are paid a lot of money just to hang around, irrespective of how well they do. They're also paid a lot more money if the company does well, either directly in bonuses or indirectly in reputation which can lead to bigger contracts in the future. This incentive structure seems to encourage shooting for the moon. If you take some risks and get crazy numbers, you'll get crazy rich. If you play it safe and get mediocre numbers (like, I dunno, 5 billion a year in PROFIT), you'll get passed by in favor of the risk takers. Since a failed risk still leaves you pretty comfortable, it's worth the shot.

How do you fix this problem? Well, one thought is to remove the safety net. Two of the more famous CEOs out there, Steve Jobs and Warren Buffett, get paid almost entirely in equity. Jobs makes one dollar a year (as he itemizes it: "I get 50 cents a year for showing up, and the other 50 cents is based on my performance."). Buffett makes 100K, and has indicated to the board that "he would not expect or desire it to increase in the future." Apple and Berkshire seem to be doing just fine, but it's worth noting that Jobs and Buffett created those companies and have devoted their lives to them. They weren't just mercenaries brought in to manage an already existing fortune 500 firm.

Another thought is that the "too big to fail" problem is really a "too big to manage" problem. Corporations like AIG are attempting to safely direct hundreds of billions of dollars in a very complicated economy. Part of the complexity comes from the sheer scale. Part comes from regulatory arbitrage. Part comes from the desire of smart executives to out-smart their rivals (perhaps there is such a thing as too much cleverness). It might be that the objector above who thinks no one should earn that much money is right, just for a different reason. There might be a limit to how big things can get before the entire endeavor is too complicated to reliably predict and you end up with a crap shoot. Boards probably have really bad heuristics for finding the "best" CEOs, and really bad metrics for determining if the ones they've got are actually performing better than the alternatives. It's entirely possible that no one can really understand something so big, so they just cross their fingers, sign the checks, and hope everything works out.

How do we encourage corporations to be smaller and more straight-forward? Well, we can start by only investing in things we understand, as Buffett suggests. There are probably other things that we can do too, but that's probably something for another day.

Thursday, March 12, 2009

Go big or go home

So I was watching some skateboarding at a local bar when I received a visit from the incorporeal form of Captain Obvious himself. The slogan for the X-Games should be "we who are about to die salute you." I guess for the first part, you can replace "Hail Caesar" with "Drink Red Bull" or something like that.

I was looking for a visual to support my point, and Google managed to find me one in 0.15 seconds. What a wonderful world we live in. Here you go:

Monday, March 9, 2009

Just nod if you can hear me

Over the weekend, I learned of two additional readers of this here screed. I knew of four readers previously, which means that the number of confirmed readers has jumped by 50%, over a weekend when I didn't post. Not bad.
So, if the six confirmed readers each told six friends a week, and those friends told six other friends, how many weeks would it take to reach the world population of 6 or so billion? Well, some considerations:
  1. We've pretty much all got the same friends and there's no guarantee of uniqueness in the 6/week requirement. Our little sub-net is well connected, but connections to other sub-nets might be harder to reach. Basically, within a week or two, we'd all be recommending this blog to each other, but distant lands like Russia, Botswana and the Mid-West might not have been reached at all.
  2. Social capital is an issue: this blog is lame, so there's a valid concern that recommending it to a friend might suggest to that friend that you too are lame. This concern might hamper the 6/week endeavor.
  3. It's one thing with like 5 readers, but there might be a threshold audience size at which I'd be too embarrassed by my own content. I'd have to choose between starting to make sense, or quitting out of shame. I'm pretty shameless, so we've got a ways to go, but I'd probably freak out long before our little mathematical growth experiment was able to recruit the entire population of, e.g., China.
Anyhow, I guess this has been my way of saying welcome to all new readers, confirmed and otherwise. See you around.

p.s. BFS ♥