A day after the Treasury properly declined to bail out Lehman Brothers and 9 days after the government had to make explicit its previously implicit guarantee of Fannie Mae and Freddie Mac at still-unknown cost to taxpayers, Treasury Secretary Paulson has determined that AIG, the giant insurance company, was on the verge of a failure that would have roiled markets and the economy beyond what the government could accept.

Therefore, on Tuesday evening, the government announced that, much like the FNM and FRE deals, it would inject a massive amount of cash, $85 billion into AIG, in return for 79.9% of the company.

AIG's stock, which was trading $2.60 after hours on Tuesday and should be lower than that on Wednesday, is down from $56 on January 1 of this year and $65 a year ago today. With 2.69 billion shares outstanding, the total wealth to AIG shareholders in the past year is about $170 billion dollars, approximately the entire GDP of the Czech Republic, and more than the GDPs of Chile, Israel, or Singapore.

People may wonder how an insurance company failed...it had nothing to do with losses from Hurricane Ike. Instead, it was AIG's insurance of mortgage-backed securities which brought them down, with the final nail in the coffin being the downgrading of AIG which required them to raise capital as collateral against those insurance policies. Of course, raising capital is essentially impossible for them, and the cascading effect of a failure of AIG could have been a massive disaster for the markets and the economy.

While I understand why the government did what it did, just as I understood its actions with Bear, Stearns, at some point this is going to have to stop. As I write this on Tuesday evening, stock index futures are trading up. But I am not so sanguine because of the confusion the government is sowing by alternately bailing out some firms and letting others fail.

I have argued that the government's action in the case of Bear, Stearns was appropriate and that the likelihood of substantial loss to taxpayers was small. At some point, however, the moral hazard involved in bailing out investors (even if only bondholders) is very important.

It's a necessary thing for shareholders to be wiped out in events like these. And there is some justice in the fact that the CEOs and executives who made such terrible decisions probably own a lot of stock. But that doesn't make up for the damage done to ordinary investors and to lower-level employees, many of whom are required to take pay packages which include over-concentration in the company's stock, both in bonuses and in retirement account contributions. One thing which would go a step further toward imposing a modicum of discipline on these firms is if shareholder insist (better than government requiring) employment contracts for the highest levels of company officers which prohibit golden parachutes and giant severance packages if the executive has presided over a lack of success, whether that ranges from the modest to the disastrous.

My guess, without having done a lot of research, is that the government did the right thing (again), but they must go out of their way to make sure the market realizes, despite recent evidence to the contrary, that these uses of taxpayer money, even if only as "loans", is the exception and not the rule.

In the meantime, the British bank Barclay's has announced that it is buying Lehman's banking divisions for $250 million in cash and its NYC headquarters and New Jersey data centers for $1.5 billion. They got only the best parts of Lehman, for pennies on the dollar, without having to take any of Lehman's existing liability. This was possible because Barclay's didn't act until after Lehman had filed for bankruptcy.

When other institutions fail, I expect this same sort of action: Smart buyers will come in when they can get only the good stuff, without the "hair". Shareholders in the failed companies will be wiped out, but shareholders in the companies with management who dodged the bullets and have available cash for strategic purposes will be rewarded. (Of course the losses to shareholders in the failed companies will be stunningly rapid, while the gains to the smart company shareholders will probably be more slowly realized over months or years.)