Thursday, March 12, 2009

Krugman’s illogic extended

I am flattered that Paul Krugman thought to reference this blog on his.  I have been a fan since undergraduate days at the Australian National University where I read (as a second year) some of his then recently published trade papers and thought that he was a seriously smart guy.  He later won the Nobel Prize for that work.  

I have kept a copy of Currency and Crises on my shelf for years too – and – like Paul am surprised that currency has played such a small role in this crisis.  The post on Swedbank (still the most visited post on this blog) was a direct application of Krugman’s book to the business of stock picking.  (Paul please read the post.  You will be amused.  Given what has happened in Latvia since I wrote that post it looks pretty good.)

I purchased seven copies of Pop Internationalism – Krugman’s mid 1990s popular work – and gave them to people who spouted all sorts of (Lester Thurow) crap about trade.  I even reviewed Krugman's book on Amazon – you can find that and my other reviews here.  

To put it bluntly – I am a Krugman fan almost to the point of idolatry. 

So I am getting puzzled at some illogic.  Paul has not (yet) agreed that the Swedish bullet (effective guarantee of all banking liabilities before selective nationalization) is going to be the way the US goes – but he acknowledges that a large amount of banking is effectively non-recourse finance with the losses going to the taxpayer.

He agrees with the obvious – that when banks are guaranteed they need to be regulated so they have more than zero capital.  Presumably they want to have quite a bit of capital (amount open to debate) or the incentives on the management to bet the money on red at the casino – or even loot the bank – is pretty high.  They also need to be regulated.

But he appears vehemently opposed to the (still not well detailed) Geithner plan to price toxic assets by giving a bunch of hedge funds non-recourse finance to purchase them.  He even refers to this a zombie financial idea - one that will not die.

This an illogical position.  Banks are non recourse.  Krugman acknowledges that and says that banks require adequate capital or they should be confiscated.  (No disagreement there – but I am sure we would disagree about the quantum of capital and how to measure it and what the confiscation process should be.)  

Anyway if there is new capital put in banking it will also be levered up non-recourse.  

I hope (and maybe I read Paul wrong here) that he wants new capital in banking and he would prefer it be private.  

Well isn’t that the Geithner plan?  Lets finance the tough stuff (scratch and dint mortgages, some commercial property) with new capital – as per any other bank – leveraged non-recourse to the taxpayer.  

The objection Paul makes is to it being “non recourse” – an objection which is illogical.  He has made this objection repeatedly.

 The debate should be – as this blog has made clear before – about the numbers (and possibly the confiscation rules).  It is not whether the Geithner idea is good or bad – it is whether Geithner demands enough private capital to be a reasonable outcome for taxpayers.

If the government requires enough capital then hey it is real capital and it reduces risk to the taxpayer.  If they require too much capital then the returns won’t be attractive enough.  There is not much economic difference between just establishing new banks and the Geithner idea – spelt out in some detail in the "long post" – for providing non recourse finance to several toxic asset funds.  

Just because it is non-recourse doesn’t mean it is evil.  If there is adequate capital then – hey – its new capital to the banking system – something that many (but not all) of us strongly desire now.  The issue is how much new capital for how much taxpayer risk.


PS.  Paul - I know you are a busy guy - but I would love an email.  I once sent you an email (on Iceland) via Joe Nocera of the NYT.  I have no idea whether you got it - but it also looks pretty good...


Anonymous said...

I don’t understand your logic.

The Geithner plan as I understand it is for private capital to provide the “capital” for the purchase of toxic assets, and for the Fed to provide the “leverage” for the benefit of that capital.

The purpose of the leverage is to improve potential returns to private capital.

The private capital is obviously non-recourse.

Why should the leverage be non-recourse?

(Other than to sweeten the deal for private capital with a form of extraordinary subsidy and potential taxpayer loss absorption - but what does that have to do with your logic?)

John Hempton said...

Come on. If I set up a new bank and borrow with brokered deposits I can lever 12 times non-recourse. If I win I keep the profit. If I lose the FDIC pays the losses.

I was thinking 6.5 to 7 times was about right for the GEITHNER funds.

Geithner lends the money to the special purpose fund. Non against the pool of purchased assets - but with private capital pitched in.

Sounds like banking to me.

Anonymous said...

The FDIC has recourse on seizure. That's what matters ulimately to the capital provider.

Why shouldn't the Fed be in the same position directly as the FDIC?

Anonymous said...

The other variable is the rate of interest charged on the debt financing provided by the Fed. Any views on that?

Also, the reasoning behind your 6.5/7x comment below would be interesting to know.


Anonymous said...

Hi John,

There is still a risk that private capital might game the system under the Geithner plan. Suppose you can contribute $10n of capital to buy toxic assets. You might be tempted to split the $10bn into five different funds of $2bn each. Then, each $2bn funds can be leveraged to purchase $20bn of assets, for a total of $100bn. Suppose two of the funds are wiped out by losses on toxic assets, the remaining three funds may make outsized returns to compensate for the losses. However, from the taxpayers' perspective, tax payers will need to absorb the additional losses on the two funds that were wiped out, instead of using capital from the remaining three funds to cover the losses...

Anonymous said...

I agree that getting hung up on "non-recourse" is more than a little bit strange. You don't even need to go the quasi-bank analogy to see why "non-recourse" is a non-problem. It's simply the way these asset disposition arrangements work in banking crisis resolutions.

A simple way to look at the Geithner proposal is that he's setting up a bunch of Asset Management Companies (like multiple RTCs) without the intermediate step of the Government first taking over the assets in an FDIC-type resolution of the banks whose assets are being purchased.

This has several advantages. It can work for those banks that we don't want to completely take over (either because they're too-big-to-swallow or they're sickly but viable). It takes advantage of private investors rather than government employees to select and price assets. It saves the time of aggregating assets in a government entity and then packaging them for sale to bottom fishers. It pre-clears the eligibility of bottom fishers. And the government doesn't have to come up with the entire amount of the price it pays for the assets as a sale (which would be carried on the books of the government?) but rather can leverage its own contribution by providing participating loans to the AMCs.

If we added the intermediate step of setting up a bad bank or an RTC, the government is on the hook for the entire price it paid. So the whole price tag is "non-recourse" and the government only recovers as it sells off assets (almost certainly below what the government paid for them).

In countries where AMCs are set up, they often operate as quasi-agents of the government anyway, putting up no or very little capital of their own, and with performance incentives for balancing the interests of maximizing the government's recovery with speed of resolution. Even if the AMC owners have to put up some of their own capital, "non-recourse" for the price the government paid for the assets is the default solution in these sorts of arrangements.

Now, it's true that we're still facing the dilemma we had under TARP V1.0 -- can we find a "goldilocks price" that is high enough to attract participation of banks and low enough for it to not be a complete giveaway. But in this case, the anti-giveaway discipline is that the price has to attract private investors, even with cheap government leverage.

But Geithner's plan has a lot of advantages over TARP V1.0 -- the government doesn't have to try to price the assets itself through a to-be-invented reverse Dutch auction, it doesn't have to pay swarms of consultants to dig through the assets of participating banks to value them since I assume that would be part of the job of the now-properly-incentivized AMCs, it skips the intermediate step of the government buying and then selling off assets, and it gets more bang for the government's buck.

Geithner's plan may not work, but not because of the "non-recourse" feature.

m said...

"Come on. If I set up a new bank and borrow with brokered deposits I can lever 12 times non-recourse. If I win I keep the profit. If I lose the FDIC pays the losses.
This is not the good bank, investing in sound assets, but the Bad bank.

Toxic waste worth? 20% of face amount or ..?

How can you justify high leverage in that case?

John Hempton said...

I was fairly careful here. The recourse I suggested (to the "bank" not the "shareholders") was the same as for FDIC guaranteed banks.

The leverage I allowed was a third of FDIC guaranteed branch deposit institutions - but I would prefer a sliding scale depending on the price the asset is purchased at. Personally I think the 6.5 of the long note is slightly generous.

The problems with looting are similar to banks. You need a level of independence, some force to diversifiy - and a few other protections. But banks will be looted more often than these funds because banks have lots of locations and these funds will be few, centralised and contractually watched.

But - this is NOT a perfect solution by any stretch. None exist here.

Anonymous said...

"I was fairly careful here. The recourse I suggested (to the "bank" not the "shareholders") was the same as for FDIC guaranteed banks."

I'm sorry, but I remain confused on the whole thing. I'm just stupid, I guess. I’d like to understand what this difference is with Paul Krugman.

The Fed's Bear Stearns (Maiden Lane) loan was criticized for being non-recourse. What would a recourse loan have looked like in that case? Recourse to what? Would it have been parallel to what Krugman apparently wants here? You say that “Banks are non-recourse”. I don’t know what that means. Recourse to what? Are there no recourse loans in banking? If a recourse loan in the case of Bear Stearns was unthinkable, why were people talking about it and why was it different from what you’re talking about here?

I have my own idea as to what recourse and non-recourse is, but apparently it is completely wrong. I’m willing to admit that. But I would much appreciate your elaboration if possible, as in a very brief recourse/non-recourse 101 summary vis a vis this debate.

John Hempton said...

Maiden Lane was a subsidy - no question.

It was a simple practical question - JPM would buy Bear Stearns provided the Fed took some assets.

$30 billion - on which the Fed will lose maybe 5 may 25 but not 30 was the cost of making Bear not into a Lehman.

If you could have done the same with Lehman you would have.

You would have dealt with the moral hazard by having the buyout at 10cps not $2 a share (let alone $10 a share where it wound up).


I am talking about living institutions selling bad assets for liquidity and to test their capital.


Anonymous said...

Some big flaws in your approach.

Compare the TALF and a bank's balance sheet.

With the TALF, the government lends every penny. In other words, it provides all the liabilities and thus is at-risk for all of them.

It's very different at a bank. Why? The private creditors provide most of the liabilities. The government is only on the hook for any deposit insurance payout. And in a seizure, it does have something akin to recourse -- it can go and seize assets before other creditors can claim them. With the TALF, there are no other creditors it can push to one side.

That is the very simple reason why Mr. Krugman is right on this one.

As for whether we should guarantee all bank liabilities, your mistake isn’t one of logic, but what I'd call financial chauvinism -- a worldview that says the all other interests in society should come second to that of the financial system. This is bad politics and bad economics.

Take this very simple counterfactual step: Imagine there were no taxpayer -- would the financial system be alive today in its current form? Nope. Much bank debt would likely be worth nothing if it wasn't for people paying taxes to Uncle Sam.

Taxpayers haven't received a single payment from bank creditors for ensuring they are made whole. I doubt they ever compensate us appropriately for the very risky risk-capital we paid in to save them.

But, the financial chauvinists respond, the system has always been set up so that the entire banking sector gets to access taxpayer support in times of crisis.

If so, why did we need special legislation to provide TARP money? Legislation that was wildly unpopular, and remains so.

Consider this instead. Maybe this assertion that the taxpayers must always support the entire banking system was made "true" only recently -- with the same sort of scare tactics that the Bushies used to get us into Iraq.

Cries that "We-Can't-Have-Anymore-Lehmans" are hardly different from cries that "Saddam has thousands of WMDs."

But, you might respond, we should seize banks and hurt creditors when "due process" has occurred to see who's weak.

Sure, but if we'd waited for that to happen, a liquidity crisis would have killed many of the banks.

So, be honest, what you really are asking for is for the taxpayers to stave off a liquidity crisis to buy time to see which ones have too many bad loans and are insolvent.

But what if a bank is bad not because of too many bad loans, but because its funding structure is vulnerable -- too much short-term market debt?

Does such a bank get a pass?

By calling for a full banking-sector liability-guarantee you are by definition saying they must get a pass. At taxpayer expense.

As can you see, Mr. Krugman is correct on this one, too -- the rage would be immense if we guaranteed all bank liabilities. And that rage would be economically and politically correct.

John Hempton said...

The last comment nails the issue. The right question is whether the policy response should be


If you decide that there should be no more lehmans you have decided that you should guarantee all liabilities".

If you do that then you should have no issue with non-recourse funding for the buyout funds.

So is the issue no more Lehmans?


Anonymous said...

Anonymous, that was very eloquently said. I couldn't agree more.

Anonymous said...

Dear Mr. Hempton,

You have undoubtedly seen Krugman's "Stressed for Success", published in his blog on the 14th of February.
Here is an excerpt: “it’s extremely hard to rescue these banks without either (a) giving a HUGE handout to current stockholders or (b) effectively taking ownership on the part of we, the people. Of these, (a) would be politically unacceptable as well as bad policy — but the Obama administration isn’t ready to go for (b), because it’s not in our ‘culture’.”

In his "Banking on the Brink" (of the 22nd of February) he kills the guarantee option: "For example, the administration initially floated the idea of offering banks guarantees against losses on troubled assets. This would have been a great deal for bank stockholders, not so much for the rest of us: heads they win, tails taxpayers lose."

To my mind, there is a middle way between handout to stockholders and taking full ownership. Its centerpiece is a three-year deal, subject to the following principles:

1. Buy now the toxic assets at their book value, but instead of paying cash -

2. Create a new asset in the (selected) banks’ balance sheets, owed and guaranteed by the government.

3. Let the existing and/or appointed managers utilize the three-year period for selling, as trustees, the toxic assets at the best favorable terms (under, possibly, economic recovery conditions). The amount of guarantee will be reduced as the proceeds come in.

4. At the end of the period, the ownership issue will be settled according to an agreed formula, the core of which is the size of the residual amount of guarantee – the higher the amount, the lower the prospects of stockholders' salvation, and vice versa.

5. Cash transfers from the government to the banks will be made (subject to the terms of the guarantee) only in cases of liquidity problems and will be taken into account in arriving at the final ownership apportionment.

This, I believe, is the most expeditious and cheap (no taxpayers’ money upfront) solution, giving the current stockholders the fairest possible chance of recouping at least part of their investment.
On the other hand, the guarantee will not be given without strings attached re management, so that practically, the healing process could be initiated as if ownership was effectively taken.

Basically, the whole crisis is about loss apportionment (between stockholders, uninsured creditors and taxpayers) rather than nationalisation. To my mind, uninsured creditors should be made to "contribute" something after stockholders are wiped out, and before taxpayers' cash is poured in.

Anonymous said...

Sensible, workable plans like that laid out in the last post have been drowned out by the hysterical shreeking of the no-more-Lehmans crowd.

This bunch of course runs the Fed and the Treasury.

Incidentally, why does the no-more-Lehman crowd always use absurdly unindicative data to show that Lehman's BK caused the crisis?

They use Libor and "spreads" from the asset-backed market.

But Libor is a useless indicator when the central banks are willing to act as a huge, alternative source of liquidity for banks.

And the shadow banking system spreads mean just as little. The market was already smashed when Lehman went bust, so it was always going to show most stress. Just as speculative small-caps will get hit badly when the wider stock market gets hit.

The Lehman shreekers never point out just how much bank lending was going on in the period afterward.

In reality, the thing that scared the policymakers most was the impact on money-market funds. The problems there were taken care of.

But the wider hysteria meant we could never hurt bank creditors again. We must all bow to them.

Merc said...

"Well isn’t that the Geithner plan? Lets finance the tough stuff (scratch and dint mortgages, some commercial property) with new capital – as per any other bank – leveraged non-recourse to the taxpayer."

It seems to me that the difference is that the banks' individual securities aren't backed by non-recourse loans, but that the entire bank is.

We don't mind the bank making a bunch of uncoupled risky investments because the ones that fail are paid off by the ones that succeed. The only time the fact that the bank itself is "non-recourse" is if they all fail.

Geithner's plan splits the bank's risky investments into a bunch of individual investments backed by non-recourse loans. Instead of the winning investments paying off the losers, the winners keep their winnings but the government eats the losses.

Am I missing something obvious?

Kien said...

Hi John - perhaps the way to make sense of Krugman's logic is that: (a) he wants to distinguish between toxic/legacy assets acquired prior to the GFC ("bad bank assets") and new loans/assets made post the GFC ("good bank assets); and (b) Krugman thinks that the US government is too fiscally weak to spend scarce resources buying up bad bank assets which will just suck up resources and do nothing to facilitate new loans/assets. I believe Willem Buiter makes this point well. Buiter thinks that spending money on shoring up the value of bad bank assets achieves nothing in terms of facilitating fresh loans and, in any event, doing this (i.e., buying toxic assets) at the scale required for it to be effective is beyond the US government's fiscal capacity. Perhaps you think just establishing a market for some toxic assets is enough to kick start the banking system?

I understand you also think that price discovery is a pre-requisite to nationalisation. I don't understand why nationalisation is so difficult to do. Why couldn't the US government set up a Sovereign Wealth Fund ("SWF") (like Australia's Future Fund) with a mandate to acquire any US bank on condition that: (a) the board and senior management of the bank is replaced as soon as it is acquired; and (b) the SWF is to be wound up in (say) 20 years. (The Swedish experience is that it is a mistake to sell too quickly.) Someone eminent like Paul Volcker could be appointed to chair the SWF. There is no need for the US government to be involved operationally in running a bank (although I'm not necessarily opposed to greater operational control by the government). In fact, the SWF could even be established by Congress, and be answerable directly to Congress rather than the US Treasury. This may be more palatable to a Congress that is distrustful of the US administration.

Mr Volcker could be charged to acquire as many banks as he thinks is under-valued, and can be profitably resold within a 20 year time frame. The US government should limit its guarantees and non-recourse loans only to banks that are majority owned by the SWF. Tying fiscal support to banks owned by the SWF will minimise the risk to tax payers as well as conserve the US goverment's fiscal resources.

General disclaimer

The content contained in this blog represents the opinions of Mr. Hempton. You should assume Mr. Hempton and his affiliates have positions in the securities discussed in this blog, and such beneficial ownership can create a conflict of interest regarding the objectivity of this blog. Statements in the blog are not guarantees of future performance and are subject to certain risks, uncertainties and other factors. Certain information in this blog concerning economic trends and performance is based on or derived from information provided by third-party sources. Mr. Hempton does not guarantee the accuracy of such information and has not independently verified the accuracy or completeness of such information or the assumptions on which such information is based. Such information may change after it is posted and Mr. Hempton is not obligated to, and may not, update it. The commentary in this blog in no way constitutes a solicitation of business, an offer of a security or a solicitation to purchase a security, or investment advice. In fact, it should not be relied upon in making investment decisions, ever. It is intended solely for the entertainment of the reader, and the author. In particular this blog is not directed for investment purposes at US Persons.