Tuesday, February 03, 2009

Bad Bank/Worse Bank

So what do we think of Geithner's Bad Bank idea? Is this the solution that will finally fix the economy? Will this mark a Bottom (tm)?

First of all, the universal bad bank idea is much better than how the TARP is currently being utilized, namely equity injections into private banks. When you have the government actually owning private companies, it opens up any number of Pandora's boxes. Already the government is trying to influence how banks operate by forcing them to lend out TARP injections. That's a terrible precedent.

On the other hand, if the government simply buys certain assets from banks, that could be the end of it. Congress could attach certain rules and regulations surrounding the asset purchases, for example, forcing banks to agree to executive pay restrictions. But once the purchases have happened, that could be the end of it.

If you want to some day return to real capitalism, then we need to figure a way to get through the current crisis. But we also need to do it in such a way that government interference in private business is minimized. As long as government owns banks, that isn't happening.

How should the purchases of bad assets be handled?

I'd like to see it done something like this. We form a new company, call it LoanCo. The Federal Government capitalizes LoanCo with some amount of money, say $500 billion. LoanCo agrees to hold weekly reverse auctions. Each auction is held with specific types of mortgages or mortgage securities. For example, one week might be OptionARMs with a certain FICO range, original loan size range, vintage year and interest rate. The next week would be a different set of characteristics.

Each bank would offer to sell their block of loans at some price, expressed as a percentage of original loan amount. LoanCo would have a pre-determined total amount they will be buying. The purchases would occur at the lowest price that "cleared" the market. Essentially, banks would be giving LoanCo limit sell orders. Bank of America might say they'd sell some set of loans at 60% of par or better. If LoanCo gets all the loans they want by paying 30%, then B of A is left out in the cold. If LoanCo winds up paying 70%, then B of A simply gets better execution.

But rather than get cash for the bad assets, the selling bank gets stock in LoanCo. All interest received by LoanCo is initially retained, but all principal is immediately returned to shareholders. The government guarantees half of the principal in these loans. In exchange, the government keeps all interest payments until LoanCo starts winding down and keeps any principal payments over the initial purchase price. So for example, if a loan was sold to the government at $60 but they eventually recover $80, taxpayers keep the $20.

(Note there is a somewhat similar plan being proffered in today's WSJ. Robert Pozen's idea is similar to mine in many ways, but I like mine better).

The advantage of my system is that banks would get capital relief, as LoanCo stock has a guaranteed value of at least $0.50 cents on the dollar. It would also make investors in banks feel more confident, knowing that the value of distressed mortgage assets can't be any worse than half of its current value.

This would also create a somewhat market-based price for the "bad" assets, at least more so than creating some model to determine a price. There is a good chance that LoanCo would suffer from selection bias. Banks would have to submit loans with certain criteria, but they would clearly pick the "worst" loans that fit that criteria. But in the scheme of things, this shouldn't be a deal breaker.

The downside of this plan is that banks get very little in fresh cash, only certainty as to the downside on their assets. But I argue that's not all bad. If we just give banks cash, they are essentially allowed to grow earnings on the backs of taxpayers. By issuing stock, the banks get the capital certainty they need, but have to figure out how to grow earnings on their own.

And I argue that banks will start lending once the fear of another round of bank runs diminishes. The margins on new loans should be excellent. I don't think we need to dole out free cash in order to incent banks to lend.

And the best part about LoanCo is that its clearly a one-time deal. The moral hazard and long-term government intervention problem is limited.

11 comments:

PNL4LYFE said...

I'm skeptical that this approach would engender any confidence in bank solvency. Bank tangible equity ratios are incredibly thin, and leaving them exposed to 50% further downside from the auction price will leave us right where we are now; we'll have a bunch of banks that can't raise any private capital because they are very likely insolvent. It also seems odd that you would have banks participate equally in any downside but taxpayers get all the upside.

Depending on where the auctions settle, it's possible that the banks will get no capital at all. If I have an asset market at 70 cents, why would I sell it for 70 cents worth of LoanCo shares that can go down 50% but can never go higher? It does nothing to help my current capital position. I'm also losing the interest income on the assets I sold. What is the motivation for any bank to do this?

Maybe I'm just misinterpreting the mechanics, but it seems like this won't help at all.

Alex Morrow said...

I have to congratulate you for being emblematic of the United States capitalist system. This is why we will work our way through this problem.

This is an entrepreneurial solution and whether its absolutely correct or partially correct it shows that we can find an answer.

The government is a not for profit operation. In this case it would be a market maker.

I'd like to post your solution on my blog. calculatedwhisk.blgspot.com My blog is dedicated to good news and your proposal is GOOD NEWS, IMHO.

Accrued Interest said...

I don't think we want to create a system where banks automatically get additional capital. Fact is a lot of big banks have plenty of capital. Its more capital uncertainty that is the problem, at least with big banks.

And I don't think we need to create a world where there is zero bank failures. There is a small bank here in Baltimore that I personally think is on the brink. I don't have insider knowledge, just knowing they were heavy lenders to developers, both real pros and rehabers.

That bank probably should go down, if indeed their losses are too large. There is no big contagion if this bank goes under, or for that matter if 20 or so other banks of that ilk go under.

The banks have to give something here.

Lockstep said...

Oh boy AI, you swung for the fences on this one.

Super SIV, TARP, FHLB, Big Bad Bank... they all amount to a hill of beans.

This is going to be nasty and we need to prepare for a 1 1/2 more of frozen credit. Just look at the auto industry unwind, no "Bad Car Company" to save them there. There is no bottom in sight and don't talk about one because it is not there. Can you imagine what will happen to assets when the banks have to mark down any loan related to areas where the auto industry predominates? How many banks will the autos take down themselves?

If anyone on this board has not referred to the events of the Great Depression to get a context of what is going to happen to the banking industry, I recommend they do so. AI mentions 20 banks disappearing, at least 150 banks across the land will disappear this year.

The only risk is if one of the big counterparty banks (BAC, C, JPM) or one of the "shadow banks" (Citadel, All hedge funds with MS and GS as a prime broker) are unable to settle trades.

This said, I am an optimist. I believe we will get through this and prosper but a "Bad Bank" provides no added value to the flushing process.

Advant Guard said...

I prefer getting six bank together and have each bring a certain amount of assets to sell. The data on the assets are circulated to all the participants. Then let the other banks bid on the assets. So Citi and Wells Fargo would bid on Bank of America assets. Bank of America and Wells Fargo would bid on Citi assets. You do three rounds of bidding and then you give the bank owning the asset the choice of whether it will accept the high price in the bidding for the assets.

All "purchased" assets go in to a pool and the selling banks are paid in a mix of cash (no more than 10% of the price), debt (B rated equivalent yield) and equity. But if a bank "buys" more assets than it sells, the mix is skewed to equity. If a bank sells more than it "buys" the mix is skewed more towards debt.

Each bank has the incentive to bid correctly on the other banks assets because it improves the value of the equity. Buy having three rounds of bidding, banks can see how other similar assets are priced and adjust their own bidding in the next round.

wrightak said...

PNL4LYFE got it right but here's my take.

A la Krugman:

Let’s suppose that a bank has a balance sheet with 1 trillion of assets and 900 billion worth of liabilities, giving it a net worth of 100 billion. However, of the 1 trillion of assets, 300 billion of them are ‘bad’ and are actually only worth 150 billion, making the bank insolvent. The only way that the bank can become solvent again is if it manages to sell those bad assets for at least 200 billion.

In your scheme, the bank doesn’t get 200 billion in cash. Instead, it gets stock in LoanCo. The first problem is, will it get enough stock? LoanCo only guarantees 50% of principal, which isn’t enough for the example (2/3 please). Which brings us to the second problem – where does 50% come from? Is this figure completely arbitrary? A guarantee of 50% might allow some banks to scrape through, but for others it might not be enough. What do you do then? Like you said, the banks only get a limit on the downside, but what if this limit still isn’t enough?

The fact is that banks have failed and if they were small enough, they would be allowed to go bankrupt. Since they’re huge and part of the system, the government needs to come in, clear out management, the tax payers need to take the hit after everyone else, and new, smaller banks need to be created.

Boon said...

Excellent idea, I think this is the clearest plan for the "bad bank" that I have heard, and it is so simple that most people will be able to understand it as well.

The double boost here will be that once the banks sells all their loans to LoanCo, although they will have to declare massive M2M losses, their shares should get an immediate boost as investors regain confidence that the worst of the writedowns is over. This will in turn make it easier and cheaper for the banks to recapitalize on their own.

Nutt ja Hala said...

Back in 1991-1993 crisis in Finland (after Soviet Union collapsed and Finnish exports with it), Finland made so called Bad Bank. While the bank waited several years (in fact until 2000s), it eventually made money to taxpayers.

of course, Finland is tiny compared to US, Finland still remained and still is undebted and they devaluated currency.

Accrued Interest said...

The problem a couple of you have mentioned is that this plan doesn't improve a bank's balance sheet. But that's true only if the bank hasn't properly marked their assets. Otherwise selling the asset in exchange for a 50% principal protected securitiy has to be an improvement. Adding the 50% backing has to improve the value of the asset.

The 50% number is totally arbitrary. I was trying to pick a number that is large enough to matter but also small enough that a loss would remain painful.

And when I said 20 bank failures, I wasn't making a prediction, merely saying that a relatively large number of smaller bank failures won't sink our system. If there is 1 or 2 banks that go under in every community but no more of the big firms go under, I think we get through this.

Bill Moore said...

How is this plan not just one big CDO with the participating banks as the equity holders?

hamariweb-fans said...

nice infoo...

bank alfalah