That’s the question I woke up with this morning. Sad, isn’t it.
The Wall Street Journal reported this week that Treasury will soon announce that it will use TARP funds to invest in life insurers, or at least those who snuck under the federal regulatory umbrella by buying a bank of some sort. The argument for the bailout is a version of the “No more Lehmans” theory: the failure of a large financial institution could have ripple effects on other financial markets and institutions that could cause systemic damage. For a bank, the ripple effect is primarily caused by two things: (a) defaulting on liabilities hurts bank creditors, and (b) defaulting on trades (primarily derivatives) hurts bank counterparties, if they aren’t sufficiently collateralized (think AIG).
My thought this morning was that life insurance policies are long-term liabilities that are already guaranteed by state guarantee funds, so we don’t have to worry about (a), and hopefully most life insurers were not doing (b) – large, one-sided bets on credit risk like AIG. So why not just let them fail and let the states take over their subsidiaries? But then I checked the facts, and it turns out that the limits on state guarantee fund payouts are pretty low. So the scenario is this: you hear bad things about your life insurer, you decide to redeem your policy (usually at a significant loss to yourself), turning it into a short-term liability, and then the insurer has to start dumping assets into a lousy market, pushing the prices of everything further down and hurting everyone holding those assets. Would this really cause a systemic crisis worse than we’ve already got? I don’t know, but no one in Washington wants to take that risk.
Ultimately, though, this goes back to the question of whether this is a liquidity crisis or a solvency crisis. If it’s a liquidity crisis – in which case you would expect to see lots of people redeeming their policies already – then there are better ways to prevent a run on the life insurers. For one thing, if the insurers really do have good assets to cover their expected payouts, the government could just boost the limits on the state guarantees, charge the insurers a premium for the guarantee (insurers already pay a premium for the backstop they get from the states), and pocket the money. Alternatively, the government could act as a reinsurer, taking on some of the payout risk in exchange for a corresponding proportion of the assets and premiums. Using TARP money might work, but since it just adds a few billion dollars to the insurer’s capital (without guaranteeing anything), it’s not a surefire solution.
If it’s a solvency crisis, though, we have to ask whether a few billion dollars of TARP money is enough. The Hartford estimates it is eligible for $1-3 billion of money. (I picked them because they are discussed in the WSJ story, not because I know anything else about them.) It also has $288 billion of assets. How do their assets compare with the assets of, say, a bank? In principle, insurance companies are more closely regulated, and their investment mix (in terms of bond ratings) is constrained. But it’s also true that insurers – especially the large ones – were investing in more sophisticated products in an attempt to earn higher yields. (For details, see pp. 156-76 of the Hartford’s latest 10-K.) And we know that you could lose a lot of money investing in AAA-rated assets. If this does turn out to be a solvency crisis, then this could be the first page of a long story.