Stress Tests

... You were on the crisis management team. ... What did that mean?

We were asked to put together the financial stability plan. What were we going to do to stabilize the banking system, reopen credit markets, and get credit flowing again? Remember, at the time markets were falling hundreds of points a day. If you step back, the overwhelming objective was to stabilize the economy, turn the economy around. ...

Many people were saying we're headed toward the second Great Depression, and that was a very real fear and a real possibility. So our piece of it was how do you stabilize the financial system? Not because anyone had any great love for banks. It was all about making sure that we stopped things from going off a cliff. ...

... The public's perception is often very different than the realities on the ground. Your team and the president need to define the programs, the answers to questions in a way that the confidence levels remain high so that you don't go down on the slope. ...

It was tremendously complex, and each individual problem was complex in its own right. The combination of them was kind of like three-dimensional chess. ...

Early on, the president said, "Look, we are going to have to make some very difficult, unpopular decisions, but we're going to do things in a way that we think has the best chance of success, even if those decisions are going to be unpopular."

So we -- and this includes the secretary and others -- started to figure out what each of the necessary elements was going to be and make sure that each of them worked with the others.

One of the great challenges we have in all this has been explaining what we did and why we did it. I've reflected on that, tried to figure out why is it so difficult, and the only thing I can come up with is that there's no magic bullet, there's no silver bullet, there's no one big thing we can point to and say, "We did X, and that's what turned things around."

It actually was a series of programs and plans and policies designed to feed on each other and reinforce confidence across the board.

For instance, we had a plan to get capital back into the banks. It involved the stress tests and a capital backstop that we made available. We had a plan to deal with reopening credit markets. Something on the order of 40 or 50 percent of all credit before '08 was generated by these markets. The third thing we had to focus on was how do we deal with the legacy assets, the old loans that were on the books and clogging up the system.

If we left out any of those three, our judgment was that the whole plan would fail. So you had to have detailed plans to deal with the banks, you had to have detailed plans to deal with reopening credit markets, and detailed plans to deal with the legacy assets.

We announced the broad outlines of our plan in I think the second week of February [2009], and continued to fill in details each week after that until the markets finally saw all the details and saw how it all fit together. I think when the markets realized that, confidence began to come back.

You can look at price charts of various instruments that indicate the health of the system, and you can see the announcement effect each time we put out details, how things would bounce off of that. That confidence started to feed on itself, and at a certain point it took hold. That's when the fever broke.

What was your opinion of the stress tests? Were you consulted?

I was not consulted. I was dubious when they were initiated, and I think this was as fine an accomplishment as the Treasury Department and the other agencies had. ...

[Why were] you were dubious about the stress tests?

I was very concerned that the announcements of the results would lead to even more pressure and even bank failures, but it was very carefully calibrated and handled by the government. I'm talking about the 2009 stress test, because there have been one set since then, and one just announced. But it was very thoughtfully done, very carefully done, and it worked out superbly. ...

What were the stress tests?

I think stress tests are a good idea. What you do in a stress test is you create an alternative scenario of terrible economic outcome, and then you test whether or not you think the bank or the institution can survive that severe, negative scenario. And so these are important things to do from a regulator standpoint.

Now it's also important for these institutions to do it to themselves internally. And I would expect any person that runs a decent business to always be thinking of what they would do under the worst outcome, under a surprise, what economists call an "exogenous shock," something that you didn't see coming. …

The ones that happened in May of '09, was there ever any doubt that all 19 banks were going to pass? …

I can't specifically recall May of '09, but you may realize that we recently did a new series of stress tests, and some made it and some didn't. And so now, of course, we heard some squawking from those that didn't quite make it. But this is part of just making sure that there is capital adequacy and that these institutions can survive an adverse scenario.

I want to come back to the Dallas Fed thesis, which is again, when you have institutions that are "too big to fail," then you have to make sure you are constantly watching them, constantly looking at them, constantly poking at them, making sure that they can survive the worst possible outcome. I think the best outcome is not to have institutions that are "too big to fail," period. …

How does it connect with the debate going on at the same time about the use of stress tests?

… I think some of the debate and discussion and argument back then was, is this stress test going to be as revealing as you think it is? Or is it going to be a way to make the banks look healthier than they really are to stave off any possibility of nationalization or taking them over in ways that would be distasteful in a capitalist economy?

And your thought?

At the time, I was skeptical. I worried that the stress tests wouldn't be as revealing as I think they ultimately were. The way this all played out had really interesting kind of path dependencies going on. I think the results of the stress tests signaled many in the investment community, in the economy that actually the banks were healthier than we might have thought. That led to more confidence in the system, and that in itself helped, I think, to generate a better credit flow.

So I was one who was critical of the stress tests and worried that the scale was going to be tilted in a way that you get the result you want. But looking back, I actually think they were pretty effective.

And so, your response to somebody like Professor [Joseph] Stiglitz, who we talked to, who still maintains that they were a con game and that the same stress tests were given to the Irish banks, who passed it, and then they went belly up.

I think the Irish stress tests were a good example of the kind of tests that I was worried about at the time. I thought that they wouldn't provide us with the diagnostics we need. I mean, you could do a phony stress test or a real stress test. And it looked to me like ours were ultimately pretty realistic. …

I think the stress tests suggested less insolvency and more liquidity, such that more temporary injections, like we got from the TARP, would ultimately help. And I think the evidence has borne that out. I mean, what did Stiglitz say about that?

He has a different opinion. … He basically thinks that we got over a hump, but the reality is that they never gave us a true understanding of the nature of the problems that some of the banks had. Of course, he's still saying that the "too big to fail" banks are only bigger now.

Well, that is true. … Part of this is your definition of success. What worked? I mean, the credit system is up and running. Beyond, obviously, Lehman, we haven't had large bank failures that have cascaded through the economy.

And if anything, the financial sector is highly profitable again. I think the problem is that they're about the only ones that are highly profitable right now. And so, that leads to, I think, very justifiable anger at bailouts that didn't help the middle class enough.

When we talked to Professor [Joseph] Stiglitz, his attitude about the stress test is that they were a con game, that in fact they were basically a softball sort of test, that all they were doing was trying to create confidence and then kick the can down the alley. ... What's your response to that?

With all due respect, it's just patently false. The notion that we could have arranged these stress tests and pulled the wool over everyone's eyes is nonsensical.

What we did is we put the banks through the stress tests which made certain assumptions. It assumed that each of these, I think it was 14 different types of loans, 14 asset classes, would suffer losses due to nonpayment, worse than at any time since before and including the Great Depression.

The Fed put out a white paper when we released the results of these stress tests which has this wonderful chart which shows what the losses would have been in this adverse economic scenario for each one of these asset classes.

The fact of the matter is, it was a two-year, forward-looking study. Guess what? Those two years are over. We know the results. They weren't nearly as bad as what was discussed in the stress tests. So the facts are the facts; the losses are the losses.

We released these results, institution by institution, asset class by asset class, because we knew there would be skeptics out there. We knew there would be people who would say, this is just a whitewash; it's ... a con game. We knew there would be that kind of skepticism and cynicism.

So we said: "Here it is. You take a look. Don't take our word for it. Here are the results." And trust me, everyone, every hedge fund on the face of the earth was plowing through that, trying to find the holes, wanting to see, aha, this was a cover-up.

It just wasn't the way it was, and the actual results, with the two years having expired, show that without a doubt. The markets believed it. When the results came out, everything bounced off those results, not because we fooled anybody, but because we said: "You do the calculations. You figure it out. If you think we were too easy, you're going to go short." And that is not what happened.

In today's global, transparent world, there's no chance even if you wanted to fool people that you could, and that never crossed our minds.

If you remember, the plan we announced in February had as one of its components a stress test on the largest banks. Basically all banks that had assets in excess of $100 billion had to go through a detailed stress test that was designed to show us -- and the world, frankly -- how much capital each of these institutions might need in an economic scenario much worse than what we were already facing.

That was a process that when we launched it, we knew there was going to be some uncertainty injected into the markets, because markets tend to expect the worst.

It was a three-month process. You had all the examiners, all the bank supervisors working together in an unprecedented fashion, digging into the books of each one of these banks.

They were coordinated. They made sure that similar assets at different institutions were treated the same, and they made draconian assumptions as to what would happen to those assets in the unlikely scenario [that] the economy turned out to be even much worse than we expected.

We needed the results of those stress tests, and we felt that we could give a clear picture to the world as to what really was in these institutions in detail, not hide anything, disclose everything and let the world judge for themselves whether or not these institutions were strong or weak.

But when we launched it, it did start this period of uncertainty, because no one knew what the results were going to be. We didn't know; the world didn't know.

Alongside those stress tests, we announced the capital plan. This is one of the least appreciated pieces of the overall financial stability plan. What we said was, we're going to do these stress tests. At the end of those tests, we'll find out which banks, if any, need injections of capital.

Then we're going to give them a six-month window where they can go out to the private markets, private investors, and raise that capital from private sources; i.e., not the government. If at the end of that period they have not been able to do it, then they have to take an investment from the government.

It was a multistep process, because we were of the view that if we can do this without government money, or with as little government money as possible, that's a better outcome for the taxpayer. We felt that the banking system is healthier and more effective if held in private hands. Governments historically have been terrible owners of banks, and we thought this was the best chance to allow that to happen.

At the end of the day, the Obama administration did not have to put one penny into any one of those large institutions, not a dime. Most people you ask today think we put hundreds of billions of dollars into each of those institutions. It's not true. ... They raised hundreds of billions of equity capital privately.

If you had asked people -- market participants, economists, folks who were arguing for nationalization, other side, us -- in January or February or March of '09, "Would these banks be able to raise hundreds of billions of dollars' worth of capital on their own?," if you asked 100 experts, I don't think you'd have gotten two yes votes. But they were able to. ...