这篇文章发表在WSJ的网站上。John H. Cochrane 是芝大经济和经融学教授。我上过他的课，是个真正得才兼备的学着。他的这篇文章深入浅出，解释了目前危机存在的原因，为什么刚刚通过的bailout plan是个灾难，以及怎样才能解决危机的核心问题。建议大家阅读。

The Monster Returns
By John H. Cochrane

Like a monster from an old horror movie, the Treasury plan keeps coming back from the dead. Yes, we are in a financial crisis that needs urgent, determined, and clear-eyed help from the government. But this plan is fundamentally flawed. It won’t even work if we leave aside its horrendous cost and long-lasting damage to the financial system. The additions and sweeteners in the Senate version, and those on the table in the house, make it even less likely to work.

A workable plan has to be based on fundamentally different principles: recapitalizing banks that are in trouble, including allowing orderly failures, and providing liquidity to short-term credit markets. These are not new and untested ideas; these are the tools that governments have used for 100 years to get through financial turmoil. However, they have to be used in forceful and decisive ways that will step on a lot of powerful toes.

The Problem

The heart of the problem now, as best as anyone I know can understand it (we are all remarkably long on stories and remarkably short on numbers), is that many banks hold a lot of mortgages and mortgage-backed securities whose values have fallen below the value of money the banks have borrowed. The banks are, by that measure, insolvent. Credit market problems are a symptom of this underlying problem. Nobody really knows which banks are in trouble and how badly, nor when these troubles will lead to a sudden failure. Obviously, they don’t want to lend more money.

A credit crunch is the danger for the economy from this situation. Banks need capital to operate. In order to borrow another dollar and make a new loan, a bank needs an extra, say, 10 cents of its own money (capital) — so that if the loan declines in value by 10 cents, the bank can still pay back the dollar it borrowed. If a bank doesn’t have enough capital — because declines in asset values wiped out the 10 cents from the last loan — it can’t make new loans, even to credit-worthy customers. If all banks are in this position (a much less likely event), we have a credit crunch. People want to save and earn interest; other people want to borrow to finance houses and businesses; but the banking system is not able to do its match-maker job.

Solving the Problem

O.K., if this is the problem, then banks need more capital. Then the people, computers, buildings, knowledge, and so forth that represent the real businesses can borrow money again and start lending it out. The core of any plan must be to recapitalize the banking system. How?

Issue stock — either in offerings, in big chunks as Goldman Sachs famously did with Warren Buffett last week, or by merging or selling the whole company. There are trillions of dollars of investment capital floating around the world, happy to buy banks so long as the price is good enough. Banks don’t want to issue stock because it seems to dilute current stockholders, and it might “send bad signals.” Lots of sensible proposals amount to twisting their arms to do so. In many previous “bailouts,” the government added small (relative to $700 billion!) sweeteners to get deals like this to work.

Let banks fail, but in an orderly fashion. When a bank “fails,” it does not leave a huge crater in the ground. The people, knowledge, computers, buildings, and so forth are sold to new owners — who provide new capital — and business goes on as usual; a new sign goes in the window, new capital comes in the back door, and new loans go out the front door. Current shareholders are wiped out, and some of the senior debt holders don’t get all their money back. They complain loudly to Congress and the administration — nobody likes losing money — but their losses do not imperil the financial system. They earned great returns on the way up in return for bearing this risk; now they get to bear the risk.

We saw this process in action last week. On Monday, we heard many predictions that the financial system would implode in a matter of days. At the end of the week, JPMorgan took over Washington Mutual. Depositors and loan customers didn’t even notice.

As someone who argued publicly against the Treasury plan on Monday, I felt vindicated.

This process does need government intervention; “in an orderly fashion” is an important qualifier. Our bankruptcy system is not well set up to handle complex financial institutions with lots of short-term debt and with complex derivative and swap transactions overhanging. Until that gets fixed, we have to muddle through.

An important long-run project will be to redesign bankruptcy; delineate which classes of creditors get protected (depositors, brokerage customers, some kinds of short-term creditors) and how much regulation that protection implies; and design a system in which shareholders and debt holders can lose the money they put at risk without creating systemic risk. But not now.

What is simple to describe economically — wipe out shareholders, write down debt, marry the operations to new capital — is not straightforward legally and institutionally. If we just throw everyone into bankruptcy court, the lawyers will fight it out for years and the operations really will grind to a halt. In the heat of the crisis, we need the same kind of greasing of wheels and twisting of arms that went into the last few bank failures.

Fancy ideas. The main point of any successful plan is to marry new capital with bank operations. There are lots of creative ways to do this, including forced debt-equity swaps and various government purchases of equity. (My colleagues at the University of Chicago are particularly good at coming up with clever schemes.)

The second part of the solution is to maintain liquidity of short-term credit markets. The Fed is very good at this. Its whole purpose is to be “lender of last resort.” We are told that “banks won’t borrow and lend to each other.” But banks can borrow from the Fed. The Fed is practically begging them to do so. Even if interbank lending comes to a halt, there need not be a credit crunch. If banks are not making new loans, it is because they either do not have capital, or they don’t want to; not because they can’t borrow overnight from other banks. (And the “other banks” are still there with excess deposits.) If the Fed is worried about commercial paper rates, it can support those.

The one good part of the current proposals is a temporary extension of federal deposit insurance. The last thing we need is panicky individuals rushing needlessly to ATM machines.

By analogy, we are in a sort of “run” of short-term debt away from banks. We have learned in this crisis that the whole financial system is relying to an incredible extent on borrowing new money each day to pay off old money, which leads quickly to chaos if investors don’t want to roll over. It doesn’t make sense to threaten that overnight debt winds up in bankruptcy court, which is at the heart of the need for government to smooth failures.

In the short run, guaranteeing new short-term credit to banks as a sort of deposit insurance could stop this “run.” If we do that, of course, we will have to limit how much banks and other financial institutions can borrow at such short horizons in the future.

Banks vs. the Banking System

Banks can fail without imperiling the crucial ability of the banking system to make new loans. If a bank fails, wiping out its shareholders, and its operations are quickly married to the capital of new owners, the banking system is fine. Even if one bank shuts down — so long as there are other competing banks around who can make loans — the banking system is fine.

I think many observers, and quite a few policy makers, do not recognize the robustness that our deregulated competitive banking system conveys. If one bank failed in the 1930’s, a big out-of-state bank could not come in and take it over. Hedge funds, private equity funds, foreign banks, and sovereign wealth funds didn’t even exist — and if they did, there’s no way they would have been allowed to own a bank or even substantial amounts of bank stock. If a bank failed in the 1930’s, a competitive bank could not move in and quickly offer loans or deposit and other retail services to the first bank’s customers. JPMorgan could not have taken over WaMu. But all those competitive mechanisms are in place now — at least until a new round of regulation wipes them out. This is, I think, the reason why we’ve had nine months of historic financial chaos, and only now are we starting to see real systemic problems.

There is a temptation for regulators and government officials to hear stories of woe from failing banks, their creditors, and their shareholders, and mistakenly believe that these particular people and institutions need to be propped up.

The Treasury Plan

The Treasury plan is a nuclear option. The only way it can work to solve the central problem, recapitalizing banks, is if the Treasury buys so many mortgages that we raise mortgage values to the point that banks are obviously solvent again.

To work, this plan has to raise the market value of all mortgage-backed securities. We don’t just help bad banks; we bail out good banks (really their shareholders and debt holders), hedge funds, sovereign wealth funds, university and charitable endowments – everyone who made money on mortgage-backed instruments in good times and signed up for the risk in bad times. This is the mother of all bailouts.

There is a storm out on the lake, and some of the boats are in trouble. Commodore Bernanke has been helping to bail water from some boats until they can patch themselves up, encouraging other sound boats to help, and transferring passengers on sinking boats to others. But it’s getting tough and the storm is still raging.

Someone had a great idea: let’s blow up the dam and drain the lake! O.K., it might stop the boats from sinking, but there won’t be a lake left when we’re done. That’s the essence of the Treasury plan.

Short of that, it will not work. Suppose a bank is carrying its mortgages at 80 cents on the dollar, but the market value is 40 cents. If the Treasury buys at 40 cents or even 60 cents on the dollar, the bank is in worse trouble than before, since the bank has to recognize the market value. Unless the Treasury pushes prices all the way past 80 cents on the dollar up to 90 cents or even 100 cents, we haven’t done any good at all; and $700 billion is a drop in the bucket compared what that would take.

There is a lot of talk about “illiquid markets,” “price discovery,” and the “hold to maturity price.” The hope is that by making rather small purchases, the Treasury will be able to raise market prices a lot. This is a vain hope — at least it is completely untested in any historical experience. Never in all of financial history has anyone been able to make a small amount of purchases, establish a “liquid market,” and substantially raise the overall market price.

Since the Treasury will not be able to raise overall market prices, it will end up buying from banks that are in trouble, at prices fantastically above market value. This is transparently the same as simply giving the banks free money. Make sure the taxpayers get a thank-you card.

There is other talk (reflected in the Senate bill) of abandoning mark-to-market accounting — i.e., to pretend assets are worth more than they really are. This will not fool lenders who are worried about the true value of the assets. If anything, they will be less likely to lend. Conversely, if prices are truly artificially low, then potential lenders to banks will know this and will lend anyway. We might as well just ban all accounting if we don’t like the news accountants bring. No, we need more transparency, not less.

Many of the changes in new versions of the bill make matters worse, at least for the central task of stabilizing financial markets.

The Senate adds language to protect homeowners: “help families to keep their homes and to stabilize communities.” That’s natural; a political system cannot hope to bail out shareholders to the tune of $700 billion dollars without bailing out mortgage holders on the other end. But it makes the bank-stabilization problem much worse. Mortgages are worth a lot less if people don’t have to pay them back. This will directly lower the market value of the mortgages that we’re trying to raise.

Yes, we need to do something. But “doing something” that will not work — with potentially dire consequences — is not the right course, especially when sensible and well-understood options remain.