Let's Help Treasury Get Out Of GM

Publish date:

I'm pretty sure that there's one or two or thirty investment bankers currently handholding at the U.S. Treasury and General Motors in their debate over when and at what price Treasury should get rid of its remaining GM shares. I'm also pretty sure that those bankers are fed up with their principals' childishness. Thus, I guess, this Wall Street Journal article. On the one hand, you've got Treasury and its unfamiliarity with the concept of sunk costs:1

Earlier this summer, GM floated a plan with Treasury officials to repurchase 200 million of the roughly 500 million shares the U.S. holds in the auto maker, according to people familiar with the discussions. Under the plan, Treasury would sell the remaining shares through a public stock offering.

But Treasury officials aren't interested in GM's offer at the current price and aren't in a rush to offload shares, according to people familiar with the matter. The biggest reason: A sale now would leave the government with a hefty loss on its investment.

At GM's Friday share price of $24.14, the U.S. would lose about $15 billion on the GM bailout if it sold its entire stake. While GM stock would need to reach $53 a share for the U.S. to break even, Treasury officials would consider selling at a price in the $30s, people familiar with the government's thinking have said.

On the other hand, you've got, um, this:

GM executives have grown increasingly frustrated with [the government's 32%] ownership, and the stigma of being known as "Government Motors." Executives have said the U.S.'s shadow is a drag on its reputation and hurts the company's ability to recruit talent because of pay restrictions. Privately, executives are also irked at the continued curbs on corporate jet use.

What would you propose if you were one of those bankers and you wanted - as you do - to earn a fee by creating a transaction? There's a range of imaginable value-gap-bridging strategies. I might - pretty arbitrarily! - start from the premises that (1) management thinks it could unlock more shareholder value2 without the stigma and restrictions of government ownership and (2) management has some idea of what it's doing. From there I might propose some sort of contingent value right that allows the government to sell its stake today but share in some upside: let GM buy shares back today in the $20s, work to improve profitability on a firm foundation of a name immune from sarcasm and extensive private-jet travel, and if it works, give the government say 75% of any value over $30 that GM manages to build in the next few years.3

Perhaps that's overengineered though: if all you need to do to improve value is lift pay restrictions, maybe you just ask Treasury to lift the pay restrictions? And convince them that you're right? You're left with "Government Motors" but maybe you make that a selling point? "Buy GM cars, or your taxes will go up," something along those lines, I'm not an ad guy. Alternatively, it is just conceivable that executives could be found who could run a profitable automaker with only limited use of corporate jets, though that seems to be an untested theory.

But I suspect any of these proposals would founder on the fact that Treasury is not an economic actor. Mainly what it wants is the political benefit of being able to say that it "made money," or at least didn't "lose too much money," on GM's bailout, which is not at all the same thing as wanting to maximize its economics now. For instance my AIG math from last week looks very different from the official math of most of these bailouts, which usually turns on just zero-discount-rate accounting profit: if you put in $100 in 2008, and after a series of terrifying gyrations and restructurings over four years, you get back $100.25, then YOU WIN and you don't worry too much about discount rates or IRRs or the heart attacks and legal bills along the way. And unfortunately, accounting profit (loss) on the auto programs is really terrible.

But couldn't that be fixed, or at least mitigated? One basic lesson of financial engineering is that if you have a client who doesn't care about something that the market does care about, then you should buy as much of that something from him as you can. If that something is as fundamental as "time value of money" then your opportunities are limitless.

My stupid sketch goes something like this: let's say, roughly accurately, that GM's cost of five-year debt is about T+350.4 That means that a dollar from GM in five years is worth about 80 cents today. Meaning that if GM "bought" Treasury's shares for $33 (the IPO price, apparently acceptable to Treasury, though still a loss) "today," but "settled" the transaction in five years, it would only be delivering about $27 worth of present value. (That's still ~10-15% premium above today's price but hey you're getting your jets back.) Call it a "postpaid forward"5 and GM can book the shares today.6 The government can't exactly say that it's gotten its money back - among other things, it hasn't gotten the money yet, and when it does it'll still be at a loss - but presumably you could write some blather about obtaining certainty about getting some of it back, I dunno, again, not an ad guy.

This seems unlikely to happen, no? For one thing, it's too obvious; you might need to throw in some obfuscations so that it's not as simple as "Treasury gets out of GM by giving GM an interest-free loan to buy back its shares." But you can always find some obfuscations; AIG's bailout was restructured something like four times, starting out as a 12% loan and ending up as a success with a 5.7% annual return. Why shouldn't Treasury take advantage of the same sort of math to pay itself off here?

U.S. Balks at GM Plan [WSJ]

1.That is a totally unfair way of characterizing Treasury; they've got uneconomic constraints of their own. Take the TARP bank preferreds, which Treasury had no burning desire to hold, and so sold them at market prices, perfectly sensible. This prompted some rather amazing wailing in the vein of "you shouldn't sell at a low price, you should just wait until the price goes up and sell then." Because the price always goes up!

2.Because ... management is worried about shareholder value, right? It needs to pay higher salaries to managers, and give them more private jet trips, to sell more cars and increase the value of the company and enrich shareholders and other stakeholders, no? It's not just because managers, like, like higher salaries and more jets, is it?

3.This is complicated if most of the share sale is to the public - does GM give a contingent value right only on the shares it buys?

4.I see GM Finance's 6.75% of 2018 at 111.55, a 4.43% yield, which Bloomberg shows as 373bps over the 2017 Treasury though some of that is curve. And 5y CDS closed Friday at ~301 after being ~325-340 previously last week. So T+350, whatever.

5.But not like this one, which was postpaid in name (and taxes) only.

6.I guess mechanically GM (1) buys 500mm shares forward and (2) sells 300mm shares spot for its own account to the public, so that Treasury gets its fake $33 on all the shares and GM only has to stump up fake cash for 200mm of them. (Then GM uses those public offering proceeds to retire some actual, money-costing debt, essentially replacing it with free government debt.) I assume that public offering doesn't come at a premium - though, I mean, shouldn't it, with the jets, etc.? - which leaves a $2-3 gap still to fill in here. Maybe you do that with just stupid engineering - make it a ten-year forward!