Andrew Ross Sorkin’s column today is entirely based on what he learned talking to Jim Millstein, the chief restructuring officer at Treasury, who seems to be very happy to talk now that he’s officially announced Treasury’s plan to exit its investment in AIG. I spoke to him for 70 minutes this afternoon, and now have a much clearer idea of how Treasury is thinking, how its math works, and why there’s a disconnect between Treasury and critics like Kid Dynamite.
Millstein made a number of interrelated points.
First, the really big picture here is being missed. There’s now an end in sight to a huge and enormously complex corporate restructuring, of an entity — AIG — which was too big to fail, too big to manage, and which had an enormous black hole at its heart known as AIG Financial Products. Today, AIG is set to emerge as a viable entity roughly half its former size, small enough to fail, with the black hole gone. That’s not only a substantial achievement; it’s also a good proof of concept when it comes to the FDIC’s new resolution authority.
This involved a big strategic change of direction at AIG and Treasury. When Treasury installed Ed Liddy as AIG CEO in the immediate aftermath of the bailout, says Millstein, the idea was very much to sell off everything — essentially, to liquidate AIG entirely. But that’s no longer the vision: instead, the idea is now to keep AIG going as a good-sized US insurance company, with a very strong property and casualty franchise and a solid life insurance franchise to boot. That company looks as though it’s going to be worth something north of $60 billion, given its inherent profitability and general stock-market valuations of insurers.
But there’s an enormous difference between an insurance company you’re trying to liquidate, on the one hand, and an insurance company which you want to survive as a going concern, on the other: it’s not just a difference of taking various assets off the auction block. Rather, it all comes down to credit ratings: in order to be viable as a going concern, any insurance company needs a solid investment-grade credit rating.
If AIG was just selling off its assets or putting its insurance operations into run-off mode, then its credit rating wouldn’t matter so much — although the higher AIG’s credit rating, the easier it becomes to unwind AIGFP’s derivatives positions without facing enormous margin calls. But Treasury looked at the bids that AIG was receiving for its assets, and determined that they were being lowballed by the likes of MetLife, since potential buyers smelled a fire sale. As a result, Treasury needed to credibly be able to say that it didn’t have to sell off all AIG’s assets.
In order to do that, Treasury needed to take a large chunk of AIG’s debt and convert it into some kind of equity. That’s why Treasury ended up owning tens of billions of dollars in preferred stock: the ratings agencies don’t consider preferred stock to be debt, and so they disregard it when assigning their ratings.
Now a lot of the arithmetic being done by the likes of Sorkin and KD is based on that preferred stock essentially being debt. After all, that’s how AIG itself shows it on their website. But it’s a very peculiar kind of debt: in fact, to a first approximation, it really is that nerdy joke, the zero-coupon perpetual bond. There’s a dividend associated with the preferred stock, but AIG is under no obligation to pay it, and it’s non-cumulative: if AIG doesn’t pay the dividend then it doesn’t remain on AIG’s books as any kind of obligation. And there’s no maturity date, either. So the obligation that AIG has to Treasury is essentially zero: it has to pay back $0 per year, in perpetuity.
The only real value to the preferred stock is that unless and until AIG starts paying the coupon, it can’t make any dividend payments on its common stock. So the preferred stock is not entirely without value. But no one in their right mind would actually pay money for it.
So when Treasury swaps its preferred stock for common stock, it’s swapping something with essentially zero secondary-market value for something much more liquid and marketable.
Of course, Treasury brought this on itself, back in February 2009, when it swapped cumulative preferred stock paying a 10 percent coupon for new non-cumulative preferred stock. Without that move, there would never have been any equity value in AIG at all — AIG would have been a loss-making entity in perpetuity. But of course Treasury owns most of the equity in AIG, so it essentially made the decision to swap debt in an insolvent AIG for equity in a solvent AIG. And the reasoning was that the liquidation value of an insolvent AIG was much lower than the market value of a solvent AIG which could operate as a going concern.
At some point, Treasury was always going to insist on converting its new zero-coupon perpetual bonds into something a bit more useful, like secured debt or unsecured debt or cumulative preferred stock or common stock. They were always a halfway house, a way of getting here from there. And in the end, Treasury decided that the easiest and most profitable thing to do would be to just convert them all into common stock.
I’m not sure I would have made the same decision. AIG is making about $8 billion a year at this point, which is more than enough to support a bit more in the way of debt without making too much of a dent in its credit rating. If Treasury had converted say $20 billion of its current preferred stock into new preferred stock paying a 5% coupon, that would pay Treasury $1 billion a year in perpetuity, and could probably be sold at or near par. Instead, that $1 billion a year is being valued on a p/e basis in the stock market, at between $8 billion and $12 billion. That’s less than the $20 billion (ish) it would be worth if it looked more like debt.
But Treasury wants to exit its investment, and selling $20 billion of perpetual AIG preferred stock would be decidedly non-trivial. Selling AIG stock is a lot easier. So Treasury decided to simply convert everything to common stock, in an attempt to get out of the insurance business as quickly as possible.
Looked at this way, it’s silly to assign hard dollar values to the Series E and Series F preferred stock and then complain when they’re being swapped for equity worth less than that sum. Instead, the only number which matters is the total amount of money which Treasury ends up getting from selling off bits of AIG and, ultimately, AIG itself. And there’s a secondary consideration, too: Treasury wants to do that sell-off as quickly as possible.
Treasury’s exit strategy certainly maximizes the speed of the sell-off. And Millstein makes a credible case that at the end of the day, Treasury is going to get out of AIG more money than it put in — some $13 billion or so in profit. That sum is not nearly commensurate with the risk that Treasury took when it bailed out the insurer. But really, Treasury had no choice: when it was bailed out, AIG had a whopping $2.4 trillion in derivatives contracts, which would have caused major systemic consequences if they had been unwound in a Lehman-style forced liquidation. We would all be much poorer, today, if AIG had not been bailed out. Any profit on the bailout is just gravy.
So it’s easy to get caught up in the weeds here. But rather than getting caught up with the relative valuations of Series C and Series F, the big picture is relatively simple: Treasury put about $47.5 billion into AIG, and the Fed added a bunch more. The Fed is soon going to get paid off in full, with interest. And Treasury is going to end up with an equity stake in AIG worth something north of $60 billion; it’s optimistic that it’ll be able to sell that stake in the market, much like it’s selling off its Citigroup stake right now. That equity stake is a matter of choice; Treasury could have structured things many other ways, and probably could have ended up with something less liquid but more valuable if it had wanted to do so.
Millstein is a fan of common equity, and is looking forward to the day when he can start selling off the government’s AIG stake in the secondary market. Then we’ll be done with AIG, we won’t have big losses to show for it, and we will have dealt with the AIGFP black hole in the interim. It’s a pretty impressive achievement, all told. And the technical dynamics of exactly what the government is doing with its current slightly peculiar preferred stock are ultimately something of a distraction.
(A couple of footnotes, which don’t fit into the broader narrative: right now, AIG has the right to borrow $22 billion more from Treasury, in the form of that Series F perpetual zero-coupon preferred stock, at any time. Under this exit plan, AIG has to use that whole credit line to pay off the Fed, and then needs to repay it with various asset sales, including the sale of the assets it’s getting from the Fed. So the plan puts Treasury at less risk that suddenly it will have no choice but to send lots of money to a hungry AIG. And, AIG won’t only be an insurance company: for the time being, it still owns an aircraft leasing company called ILFC. But it has said that ILFC is non-core, and it will be happy to sell it at the right price.)
Update: It seems that Kid Dynamite had a similar conversation.