Hank Paulson’s $700bn “bailout” plan unleashed a flurry of alternative proposals, as most people recognize that time is running out. There is an urgent need for a significant intervention to break an accelerating downward spiral that is threatening the very survival of the financial core of the world economy.
Most proposals, including the one just agreed to by Congress, have in common a few general principles. First, they recognize the need to recapitalize the financial system and to improve the liquidity of several key asset and insurance markets. Second, there is agreement on the need to protect taxpayers by giving the government a share of the upside as well. Third, most see moral hazard as a reason to limit the extent of intervention and, in particular, to punish shareholders. Not doing so, the argument goes, would make future crises more likely as it would encourage the financial sector to repeat the excesses that caused the crisis in the first place.
We share the first two “principles” but are less persuaded by the third one. The main problem of the standard moral hazard view is its disregard for the incentive problems it generates within crises. In real life, unlike in many of our models, crises are not an instant but a time period. This time dimension creates ample opportunity for all sort of strategic decisions within a crisis. Distressed agents have to decide when and if to let go of their assets, knowing that a miscalculation on the right timing can be very costly. Speculators and strategic players have to decide when to reinforce a downward spiral, and when to stabilize it. Governments have to decide how long to wait before intervening, fully aware that delaying can be counterproductive, but that the political tempo may require that a full-blown crisis becomes observable for bickering to be put aside. Each of these agents is in the game of predicting what others are likely to do. In particular, the likelihood of a bailout and the form this is expected to take, change the incentives for both distressed firms and speculators within the crisis. These incentives are central, both to the resolution of the current crisis as well as for the severity of the next crises.
A standard advice stemming from the moral hazard camp is to subject shareholders to exemplary punishment (the words used by Secretary Paulson during the Bear Stearns intervention). This is sound advice in the absence of a time dimension within crises. With no time dimension, all shareholders were part of the boom that preceded the crisis and as soon as the bailout takes place the crisis is over; the next concern is not to repeat the excesses that led to the crisis. Punishing shareholders means punishing those that led to the current crisis, and it is better that they learn the lesson sooner rather than later.
However, this advice can backfire when we add back the time dimension. Now, the expectation that shareholders will be exemplarily punished if the crisis worsens delays the decision to inject much needed capital by stabilizing investors. As a concrete example, sovereign wealth funds are now much less eager to inject equity into the US financial system than they once were. Conversely, destabilizing speculators and shortsellers see the value of their strategy reinforced by the policy of exemplary punishment. For both reasons, crises become more acute, as the equity market becomes extremely one-sided when uncertainty and risk rise during bouts of panic and confusion. The anti-moral hazard strategy turns into a crisis enzyme.
This perspective leads to several observations regarding how the details of the bailout, many of which are yet to be determined, should be arranged. One objective must be to signal to strategic investors waiting in the sidelines that prices will stop falling and thus discourage speculative waiting. Speculators will not expect that prices of securities will be lower than those established at the Treasury’s auction (if indeed an auction is used), at least in the period that immediately follows the auction. Thus the date of the auction provides a clear deadline to any speculative waiting. Announcing a timetable for purchasing a given list of securities may therefore have a salutary effect on prices even before the actual purchases take place.
To the extent possible, the first securities to be purchased should be those where the evidence of mispricing is greatest. For instance, certain AAA-rated tranches of subprime mortgage backed securities have been trading at prices that are hard to justify except by the extreme illiquidity of the market. If these securities were first on the Treasury’s list this would signal to speculators that the possible gains from speculative waiting will soon disappear.
One risk is that if some of the holders of a particular security are especially distressed, this may lead to fire-sale prices when the Treasury purchases the securities. To some extent, this risk is mitigated by the profits the taxpayers would make on this purchase. Still, there is a concern that purchases at excessively low prices would harm other security-holders, if nothing else from having to mark-to-market their remaining holdings. One way to partially avoid this situation is to commit to purchasing a sufficiently large amount of each security to minimize the impact that any particular security-holder’s distress will have on auction prices.
Finally, the Treasury’s plan contains as-yet-unclear provisions for giving the government an equity stake in the companies it assists. Presumably this will involve diluting the holdings of current shareholders. One way to take into account the within-crisis incentives this policy generates would be to give special consideration (for instance, lower dilution) to firms that raised fresh capital since the start of the crisis.
To be clear, our position is not that the standard moral hazard concerns should be disregarded. Instead, our argument is that it is important that when designing policies to address it, we are more mindful of the perverse incentives that they may trigger within crises. The “exemplary punishment” approach is one example of a misguided policy along these lines, letting Lehman go under may have been another one, but there are many post-crisis regulatory responses that could deal with moral hazard without backfiring during the crisis.
Ricardo J. Caballero is professor of economics at MIT. Pablo Kurlat is a PhD student in the department.