What is the AIG trial about? I am tempted to say that it is about whether Delaware law required a separate class vote of common stockholders on a reverse stock split that AIG conducted in June 2009, but that seems sort of perversely contrarian. I mean, that is actually what it is about, you can read the complaint, I am exaggerating only a tiny bit.1 But why would anyone call Hank Paulson and Tim Geithner and Ben Bernanke to testify on the Delaware corporate governance standards applicable to reverse stock splits? They wouldn’t. It doesn’t make any sense. There’s something else going on here.
Ughhh right it’s re-litigating the bailouts! Here’s the thing. In 2008, AIG ran out of money. This was a thing that happened to a lot of companies in 2008. Those companies asked themselves where they could get more money. The same answer occurred to a lot of them: Maybe the government would give them the money. Mostly, it did. For instance, it gave money to AIG, and Fannie Mae, and Freddie Mac, and JPMorgan, and Goldman Sachs, and Citi, and Bank of America, and Intermountain Community Bancorp and hundreds of other banks. But not everyone! Lehman Brothers did not get the money it needed, and so it sank beneath the waves.
But the government exacted a higher price from AIG than it did from the big banks, demanding a 14 percent interest rate on its loan and a 79.9 percent stake in the company. This left AIG’s previous shareholders with less money than they would have had if the government had offered nicer terms, though with more money than they would have had if the government had offered harsher terms, and with much more money than they would have had if the government had just let AIG fail the way it did with Lehman that same weekend. (Then they would have no money at all.)
But Hank Greenberg, AIG’s pugnacious former chief executive officer, still owned a lot of AIG stock, and he sued, arguing basically that the government should have given AIG a nicer bailout, the way it did with JPMorgan and Goldman and Citi. This is not a real legal argument, at all, even a tiny bit, so his lawyers had to dress it up with the stuff about the reverse stock split, which to be fair is kind of a real argument. (Kind of.) But most of their rhetoric, and most of the celebrity witness examinations, are about the unfairness and malice of AIG’s bailout, not the boring legal stuff, because honestly nobody wants to hear about class votes on reverse stock splits.
But of course the AIG trial can’t really be about how AIG’s shareholders got treated unfairly, either, because that’s ridiculous. As far as I can tell there is one person on earth who thinks that, and it’s Hank Greenberg.2 So the AIG trial can’t be about what it’s about (class votes on reverse stock splits), or about what its participants think it’s about (the mistreatment of AIG shareholders), so it has to be about a weird third thing, which is the excessively generous treatment of the banks who were bailed out under less onerous terms than AIG was.
That’s a thing that gets re-litigated a lot so I guess one more time can’t hurt?
If you do want to think about that, it’s helpful to have a basic framework of what a banking bailout is, and what a bank is. Simplifying a lot, a bank is a thing that allows some people (“savers”) to put money somewhere, get paid interest on it and be confident that they’ll get it back, and that allows that money to be invested in the real economy. That is: Banks have risky claims on the real economy and hand out risk-free claims to their savers.3
This is a bit of magic that works most of the time, but not all of the time. Sometimes people realize that their risk-free claims are backed by risky assets, and might be riskier than they thought, and then they panic and pull their money out of the banks and that’s a big problem for the rest of the economy. This has been an extremely well-known problem for centuries, and the solution has been extremely well-known for about 140 years. It’s for the central bank to lend the banks money until the crisis passes.
This has nothing to do with subprime, or derivatives, or too-big-to-fail banks. It’s just a feature of banks, which are where the money is, but only in a probabilistic sense. Sometimes they are not where the money is, and that’s a crisis. But if the central bank (or, sort of equivalently, the government) lends them money, then the crisis will pass, and they’ll be able to pay it back with interest.
Many people dislike this, and it is sort of unseemly, but it really is a well-known set of facts. You can reduce the risk of banking crises happening, but not to zero, because of that core mismatch between banking’s risky assets and savers’ expectations of safety.4 And if crises do happen, central-bank support seems to be the only effective way to solve them.5
That framework helps explain why AIG’s bailout was so much harsher than everyone else’s. Like, for one dumb obvious thing, AIG wasn’t a bank. It ran into trouble in businesses like securities lending and insuring senior tranches of structured credit securities, which are not traditional banking businesses. Businesses — even financial businesses – run out of money all the time, and the government’s response to them is normally to ignore them as they go bankrupt. By that standard, leaving shareholders with 20.1 percent of the company, instead of zero, looks positively generous.
This is not a wholly satisfactory explanation insofar as lots of other businesses were rescued that were not traditional banks. Goldman Sachs and Morgan Stanley, for instance, became banks, but only barely, and have never exactly worked like traditional banks, taking lots of deposits to make a lot of small-business loans. The government guaranteed money market funds, which are also not banks.
What I think this means is that, in our modern financial system, things that are not traditionally called “banks” serve banking functions.6 It used to be that the way that risky investments were transformed into risk-free claims was through bank lending and bank deposits. But lots of other things — repo, money market funds, asset-backed commercial paper — have now evolved to serve similar functions, not least because they could serve those functions without the cost of complying with bank regulation. One obvious lesson of the crisis is that these things are functionally like banking, not least because when they go awry they cause banking-like crises. And the policymakers drew the quite reasonable conclusion that those crises should be addressed with the bank-crisis toolkit, that is, by lending money until the crisis ended and the bank-like things could pay it back.
But this is a thing that they figured out over time, by trial and error. “Let Lehman fail, what’s the worst that can happen?,” they asked, and then found out.7 AIG, which failed the same weekend as Lehman, was big and creepy enough that it got liquidity support from the government, but not quite bank-like enough to get that support without a dose of punishment.
Because of course the big problem with all this government support is moral hazard: If you can always get a government bailout with no bad consequences to you, you will take excessive risks. So there should be bad consequences of a bailout. On the other hand, there need to be good consequences of a bailout, too. If taking a bailout meant that all shareholders were zeroed and all employees were fired, then banks would resist taking bailouts, and panic would spread. The ideal balance is that you want banks to
- not want to do anything that could make them need a bailout, but
- take a bailout as soon as they need one.
Striking this balance is hard, which is why the normal approach in the U.S. is to regulate banks to keep them from succumbing to moral hazard, rather than to punish them severely if they do. This doesn’t always work, because regulators can be wrong or lazy or deceived or captured or whatever. It also didn’t quite work in 2008 because the things that were regulated as banks were not the only things that ended up requiring bank-like bailouts, which led to inconsistencies like Lehman’s non-bailout, AIG’s harsh bailout and Goldman’s loving embrace by the Fed after it converted into a bank holding company. So now a big focus of financial-stability regulators is to figure out the things that might require bank-like support, and to regulate them in somewhat bank-like ways. Obviously both parts of this — the regulation now, and the foreshadowing of support in a crisis — are fiercely controversial.8
But in 2008 no one quite knew this, so they made it up as they went along, balancing moral hazard and the need to avert panic by trial and error. My own dumb model for why AIG’s bailout was worse than Citi’s goes like this:
- In March 2008, the government helped bail out Bear Stearns, guaranteeing a lot of its liabilities to enable JPMorgan to buy it for $2 (later $10) a share.
- This bailout was widely seen as too generous to Bear, and to JPMorgan, and as creating moral hazard, so that was bad.
- On the weekend of Sept. 13-14, 2008, the government was presented with Lehman Brothers and AIG. It took a much less generous approach, refusing to bail out Lehman and giving AIG the harsh terms to which Greenberg now objects.
- These decisions were widely seen as too mean to Lehman (at least, and probably to AIG, too), and created further panic, so that was bad.
- In October 2008, the government gave the big banks bailouts that were more generous than what AIG (and certainly Lehman) got.
- Those terms were, let’s just say for argument’s sake, just right.
That is a really dumb model, and I don’t think that Paulson or Geithner or Bernanke would exactly endorse it, but it has some plausibility, doesn’t it? It’s not like anyone knew what they were doing. You just zig-zag between alternatives until you get it more or less right. Did the bailouts that the big banks got do too little to prevent moral hazard? Quite plausibly, but, you know, you have until the next crisis to fix that. Did the bailout that Lehman didn’t get do too little to prevent panic? Almost certainly. Was the bailout that AIG got too harsh? No, sort of by definition, because AIG took it.
That doesn’t end the discussion over the bailouts — nothing will! — but it really ought to end AIG’s case: The board was offered a deal, the deal was unpleasant, the alternative was worse, it took the deal, it can’t now seriously complain. Except maybe about the reverse stock split thing.
(Corrects the date of the Lehman/AIG weekend in 17th paragraph.)
1 The claim is that the government took AIG shareholders’ property without just compensation. But what actually happened is that the government offered AIG a loan, on harsh terms that included a 14 percent interest rate and a 79.9 percent equity stake in the company, and AIG’s board of directors said yes. If I offer you a deal, and you agree to it, then I haven’t “taken” anything from you, even if your alternative to the deal was bad. (Actually, the AIG shareholders gamely argue that, but I have trouble believing that they could possibly mean it.)
But that 79.9 percent equity stake, it turns out, was more shares than AIG had authorized. So it needed a shareholder vote to authorize more shares. And shareholders did not want to give the government 79.9 percent of the company, whatever the board thought. So … the company just sort of issued the extra shares without the required shareholder vote. As a maneuver, it had a distinct fishiness to it. And that is actually the main interesting legal claim in the lawsuit.