« Martin Feldstein: The European $afety Net | Main | Easing Financial Market Stress »

Saturday, March 15, 2008

A Subprime Conversation

So what's the worst that could happen?

Earth could be hit by a asteroid wiping out life as we know it.

You know what I mean.

Just trying to put it in perspective. Where do you want to start?

At the beginning.

O.K. In the beginning, there was financial innovation, and it was good. Or so we thought.

You're not funny. So financial innovation caused the housing boom? How?

It wasn't just financial innovation. Several factors came together at once. Financial innovation created complicated financial instruments from mortgages that were sliced and diced - think of a salad of chopped up mortgages - and put together in ways that made it difficult for people to see if there was mold or something hidden in the pieces.

Huh? Mold?

These financial instruments were made up of little pieces of of lots and lots of different types of mortgages and because of that they had risks that were difficult to assess - we say they lacked transparency or that they were opaque - and this led people to assess that these financial instruments were safer than they actually were. People who bought these financial instruments were taking on far more risk than they thought.

Why did they make that mistake?

Several reasons, and this is what I meant by many factors coming together at once. One factor was people believing that the boom would continue, that housing prices would continue rising for the foreseeable forever. There were people, serious analysts, who said that we had entered a new financial world, one where new financial instruments coupled with the decline in the variability of GDP in the mid 1980s had created a new financial environment where risk had fallen considerably. There were plenty of people who wanted to believe this was true.

There were also stories about why housing demand would remain robust, and so on, that helped to keep things going. Another factor was the failure of regulators and bond rating agencies. These failures were big contributing factors as well.

O.K., let me see if I have this. Financial innovation, lack of transparency leading to misperceived risk, the belief the boom was a new trend rather than an aberration, and the failure of regulators and rating agencies fueled the housing boom?

So far, so good. But there's also the question of where the funds that fueled the boom came from. Here, I'd add two more factors to your list, high savings rates in Asia and in oil producing countries, and a low interest rate policy by the Fed needed to ease the fallout from the last asset bubble that popped in 2001. Both of these factors helped to create the liquidity needed to start and sustain the boom.

As this crisis developed, the Fed didn't seem to react very fast.

It took awhile for policymakers and financial market participants more generally to recognize and accept that there was a problem developing in financial markets. There was some turmoil during and before summer 2007, and there were certainly people who recognized that the coughs in financial markets meant the system had more than a mere cold, but it wasn't until late summer, around August 9th I think, that events finally caught the attention of policymakers and market participants generally.

What happened?

There was a "significant liquidity event," the suspension of a fund in Europe triggered widespread worry and a flight to safety. This caused banks to begin demanding reserves as insurance against potential problems, and this in turn sent the federal funds rate far above its intended target at the time of 5.25%.

I kind of remember that. Is that when the Fed first dropped rates?

Yes. Initially they used traditional monetary policy tools, textbook open market operations, to lower the federal funds rate by three quarters of a percent.

Did that work?

Temporarily at best, but it didn't get at the heart of the problem so it was not a long-term fix. The strains on the system were outside the traditional banking system, places the Fed had no authority to operate, or if they had the authority they didn't have the institutional structure in place, so the Fed had no way of attacking the problem directly by dealing "face to face" with the troubled institutions. As a result, commercial paper markets and credit more generally began drying up - a troublesome sign.

"Danger Will Robinson, Danger!!!"

Yep, and good job with the flailing robot arms. With credit markets drying up, the Fed needed to try something else so it took what I thought at the time was a fairly innovative action, though now it's easy to see that far, far more was going to be needed.

What did they do?

The Fed began accepting a wider array of assets at the discount window - the place where banks take out loans from the Fed. They didn't take just any assets, these were highly rated, fairly safe assets, but it was a broader set than before.

How did that help?

It allowed traditional banks to act as intermediaries between the discount window and troubled institutions. Here's how it works. Suppose you are a financial institution outside the traditional banking system holding highly rated financial paper that, because of the fear in the financial marketplace, the fear that assets may have hidden risks or that markets might collapse, cannot be traded for needed funds. Thus firms hooding these illiquid assets face a liquidity crisis.

The discount window helps, or was supposed to help, by allowing these firms to trade the illiquid financial assets to a traditional bank for cash, then the bank would immediately take the assets to the discount window and use them as collateral to replace the cash they had just given to the financial institution. In essence, the Fed trades cash for the illiquid assets, the bank is just a go between who makes a little on the spread. And to help even further and make the spread more attractive, the Fed also lowered the discount rate by half a percentage point to encourage banks to use the discount window and distribute funds to troubled institutions.

Sounds like a good deal for banks, but you said "supposed to help," so I gather this didn't work as planned?

Banks were hesitant to visit the discount window because of the stigma attached to doing so. The worry was that when investors found out that a bank had visited the window, they would worry the bank was in trouble, and that would cause the trouble investors feared as they withdrew assets from the bank. So banks were unwilling to risk their own health just to make a few bucks on the spread, and because of this the liquidity didn't get to the institutions that were in trouble. The problem was that the fear in the markets made people wary of hidden risks. That made it difficult, even for banks that appeared healthy on paper, to visit the discount window without triggering wariness among investors.

So why not just stop announcing who gets loans?

That's essentially what the Fed did when they created the Term Auction Facility, called the TAF. This allowed banks to bid for a preset amount of funds, but the important part was that the winners were anonymous. Banks could borrow funds from the Fed and funnel them to where they were needed without risking their own health by having it revealed that they are borrowing from the Fed.

But even that wasn't enough?

Problems continued, bond insurance markets, student loan markets, municipal bond markets - the list goes on and on - the troubles have not stopped and they appear to be widening.

So is that it, has the Fed run out of options, are we all doomed?

The Fed doesn't seem to think so. The TAF wasn't the last attempt to create an institution that would be able to get funds to the points where they are needed and, more importantly, to remove risk from the system and ease the fear that has frozen these markets. Recently, they created the Term Securities Loan Facility, or TSLF. This will accept a broader array of assets, in particular highly rated mortgage backed securities that have been illiquid, and deal with a wider array of customers that go beyond traditional bank boundaries.

How will this help, by providing liquidity to these banks?

I don't think liquidity is the main point this time, at least not through increasing the overall quantity of assets, though they are liquefying mortgage backed securities and other assets they take in trade. That is, this is more of a trade in assets - illiquid for liquid - rather than a change in the quantity of assets. The hope is that this will provide temporary relief to markets by removing some of the risk, relief that lasts long enough for markets to begin recovering on their own.

What do you think? Will this work?

I think the Fed will need to assume more risk than they have to date to unfreeze markets, and even that might not be enough. Remember, there are unrecoverable losses that must be suffered. When we talk about relief we mean that we hope things aren't worse than they need to be, that a small fire doesn't jump borders and become a large devastating fire, but so far the fire has continued to spread so it's too early to say we have this thing fully contained - and most signs point the other way. For example, we saw the troubles with Bear Sterns, though I found it encouraging that the Fed showed it was willing to take aggressive action if and when it is needed.

If it were my choice, If I were king of the Fed, I'd do more. First, I'd trade for any financial assets at a price that fully reflects the risk of holding that asset. The Fed should trade a non-risky assets, money or government bonds, for risky private sector assets at a risk adjusted price. And to promote these trades, the risk adjustment could be reduced - the amount the discount is reduced would be just like adding reserves to the system.

Can you explain that last part?

Suppose the Fed takes a mortgage backed security off the hands of a bank that wants to reduce its risk profile, and pays the bank 80% of its value, the 20% "haircut" is to compensate for the risk of holding the asset. The Fed could pay 90% instead. If it were an asset worth $100, for example, the Fed would give the bank $90 in reserves in trade for the asset. That would break down into $80 on a fair risk-compensated trade, and $10 extra in reserves.

What if the Fed loses money on these trades?

That could happen. But these assets could also increase in value as well, precisely because the Fed is holding them and reassuring markets. If the Fed makes a fair trade, e.g. pays 80% in the example above, it is as likely to make money as lose money. The assets could, of course, be mispriced meaning that the Fed has more or less risk than it thinks, but even in the very worst case - every single asset they hold falls to zero - I'm not sure it would be a disaster.

And I'm not sure you aren't nuts.

Think of the example above. Suppose that the value of the asset the Fed holds falls to zero after the trade. The trade was permanent, so there s no margin call or anything like that - the Fed owns the asset and it is worth nothing. Then, in the end, it is no different than the Fed simply printing that same amount of money and giving it to the banks as new reserves, it is an increase in the money supply. It is a large increase, and it could surely be inflationary, but inflation is not the main worry when financial markets appear on the verge of a downward spiral that could drag the economy down along with them.

The idea is to trade less risky assets, money and government bonds, for risky assets at a fair price, perhaps with a premium attached to encourage these trades, remove as much risk as possible from the system, and hopefully ease the fear that has been so pervasive. The amount of the trades could be capped to avoid taking every single risky asset in the economy in trade, and there are ways to set the incentives so that only the riskiest assets will be traded. The Fed might make money if things turn out better than expected and the assets go up in value, they might lose as well - that is the risk they assume and that is by design. But unlike they private sector, they have already replaced their losses by printing the money used to purchase the assets. They are gambling with money created out of thin air.

Yeah, you're smarter than the Fed. The Fed prints money, buys risky assets, problem solved. Yep, that'll work. Don't be offended, but what other ideas are out there?

Well, I've tried to talk in fairly general terms, but the actual details - involving repos, etc. - would be quite a bit more complicated. Most of the analysis we are hearing lately says that we are not going to be able to avoid using public money in one way or another to bail out troubled institutions. I agree with that, I think that is becoming increasingly likely. Our first allegiance has to be to protect those who did nothing to cause the troubles we are experiencing, and it may be necessary to use public money to minimize the chances of a widespread downturn. The actual details of each proposal differ, and most proposals pay more attention to the actual details and intricacies of credit market institutions than I did in my example, but the point is generally the same - the Fed tries to erase fear by making it clear that steps will be taken to avoid worst case outcomes.

One more thing, though it's playing second fiddle right now and getting to be a bit late to use it, there's always fiscal policy - I argued for more aggressive fiscal quite awhile ago, but that didn't happen and I doubt we'll get much more in terms of fiscal policy than what is already planned.

Well, I don't mean to - yawn - say I'm getting bored or anything, but, well, I should get going. But one more thing. Is the meteor about to hit. Again, are we all doomed?

I'm an optimist, and when I look at the data we are not there yet - we haven't passed into the zone of no return - but there's no doubt we are headed in that direction, and unfortunately slamming on the policy levers as hard as we can may not be enough to keep us from slipping into a recession. Deep down, I still have hope it won't be so bad, that suddenly the trouble will pass and we'll pull out of this, but it's not looking good - our celebrated robustness and flexibility may not be enough to deal with a shock this large - and analysts everywhere are increasingly gloomy.

That's not a real answer to the question. You didn't actually say what you think will happen.

I wouldn't bet my house on things getting better.

Real funny.

I thought you said you have to go.

[Update: Later...]

Oh, hi again - didn't expect to see you here.

Hey. I was thinking - about our earlier conversation - I saw a proposal from Brad DeLong after I talked to you, but it seemed different than what you were saying. What does he want to do?

I should have talked about that proposal as it's something I thought was a good idea several weeks ago when Alan Blinder first proposed it. Brad wants to, as quickly as possible, set up an institution to purchase mortgages and reissue them at more attractive terms. The goal is to limit the number of foreclosures and restore stability to financial markets.

How is that different from what you proposed?

His proposal differs from mine in two ways. First, the government buys mortgages directly rather than buying mortgage backed securities, then reissues those mortgages at more attractive rates. Second, Fannie Mae rather than the Fed purchases the assets since that is the fastest way to put the mortgage purchase plan into action.

Would this prevent defaults and foreclosures, is that the idea?

That's the hope, and if it does restore confidence in financial markets, then there would be no reason for the Fed to do as I proposed and remove risk from the market by purchasing risky mortgage backed securities (since they would no longer be as risky).

Hope? Is that all it is?

It's more than that. I think purchasing mortgages would help to stabilize the mortgage market, so, as I've said in the past, I am in favor of the mortgage purchase plan, particularly since it helps homeowners directly. I am just not sure that it will be enough by itself to get credit flowing again.


One reason is that loan characteristics are different today, e.g. they are longer, so the experience of the past may not be a very good predictor of what will happen this time around. Because of the uncertainty about whether purchasing mortgages will be enough to stabilize markets, I would like to see a combination of both policies, purchase troubled mortgages and purchase troubled financial assets at the same time, particularly since the purchase of financial assets can be put into place very quickly, an important consideration right now. I should also add that other measures, e.g. regulatory change, could also be put into place so his is not all that can be done. For example, as Richard Green has noted, it will be important for regulators to ensure that there is no incentive for borrowers and lenders to repeat the behavior that led to the need for intervention.

I"ll keep my fingers crossed.

    Posted by on Saturday, March 15, 2008 at 12:34 AM in Economics, Financial System, Housing, Monetary Policy, Policy | Permalink  TrackBack (0)  Comments (46)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference A Subprime Conversation:


    Feed You can follow this conversation by subscribing to the comment feed for this post.