Tuesday, September 30, 2008

The memory hole is powerful

Shrill partisan Brad Delong attributes yesterday's market drop to the "no" vote in the house. Unfortunately, we were all alive during that time and the record of what actually happened is clear. The house "no" vote came in at 2pm EST, which is NOT where the big drop in the Dow happened.

If you cannot believe what you read in the newspapers about Sept 29, 2008 on Sept 30, 2008; how can you believe what they say about any time further back in history?

Update: You can see the memory hole in action on Brad's site. I posted a comment pointing out that his representation of reality did not, in fact, reflect reality, as both market prices and when the vote was confirmed are public knowledge. Brad deleted that comment. With Brad's "reality" being such a slippery thing, no wonder he needs to "grasp it with both hands".

Is the Paulson plan worse than doing nothing?

The Paulson plan is worse than doing nothing, yesterday's vote was a good one, but it looks like Congress, like European countries, are going to be asked again and again if they want to join the EU until they make the right decision. I heard new commentators today ascribe yesterday's decline in the Dow to the House rejecting the plan. Unfortunately, I was alive yesterday, and I know that the market opened over 200 points down, dropped another 400 points when the AIG list came out, and the final 200 when the House voted "no". Maybe the station I was tuned to (NPR) was for people who were born yesterday, and so did not know what happened.

Here are better suggestions for what to do:

1. Liquidity (short term deposits that fund long term debts)

Yves is on top of this. Essentially, extend FDIC cover to additional elements in the financial system, a move that Bernanke has already begun, but needs to complete.

2. Solvency (banks don't have enough money)

Hussman offers on approach, which is for the government to take super senior debt. This would both improve the quality of balance sheets, and expand (or maintain their size). Issue this debt at a high interest rate, and you have Bagehot's classic "lend freely, but at punitive rates" approach, which is perfectly fine here.

(Also, if FDIC insurance is extended to part of the shadow financial system, it makes total sense for them to borrow directly from the Government also, as that is essentially what FDIC insured term transformation organizations results.)

Waldman has a similar approach, but basis it on the Government taking an equity position instead (so is more Nationalization-like than Hussman).

Michael Lewis is right -- all the Paulson plan does is protect Goldman Sachs' bonus pool.

Enough with the metaphors

Megan cries uncle wrt to the metaphors swirling around the massive (and necessary) de-leveraging underway in the US. The reports on NPR this morning were excruciating, with "liquidity" "greasing the wheels" so the "cogs of business" did not "seize up". I would love for someone, anyone, to show me a cog that needs greasing. Even terms like "liquidity" and "solvency" are very hard to understand, and are used in a variety of different ways, most of which, I am sure, are wrong.

I'm going to try this rule -- any discussion of the liquidity crises needs to include an example balance sheet, so you can actually point to numbers changing. John Hussman shows how it is done here explaining why the Treasury must overpay for assets under TARP for it to do any good. While TARP 2 was more dishonest than TARP 1, TARP 1 was dishonest in that Paulson did not state plainly that his goal was to give banks money. Which he must do, and it must be, if they are insolvent. winterspeak reader Ironman created a web page for you to monkey with balance sheets of your own -- which is pretty cool -- over here. Thanks!

Monday, September 29, 2008

The revised bailout package is awful

The revised bailout package is terrible. It should not pass. Yves put it best: "turning Hank Paulson's three pager into a 110 page draft made for a nice fig leaf but made virtually no substantive difference." Those, like Paul Krugman, who said "no equity, no deal" should be rejecting this, but as you would expect from a good Progressive, he's backing the civil service plan.

The Paulson Plan was better than the current plan in that it was easy to understand. It would have been even better if it read "Give me $700B so I can give it to firms as I see fit". After Summer's terrible faux pas that got him thrown out of Harvard ("men are different from women, and this difference might be biological") it's not surprising to see him grovel to get back into the good graces of the permanent civil service, but his FT op-ed is drivel.

At heart, I think the problem is the utter failure of academic macro-economics. This is the best Arnold could dredge up
In a comment on my Where are the Macro Theorists? post, G. Martinez says that he teaches the current crisis as a surge in money demand, with Treasuries playing the role of money. Everyone else's borrowing rate goes up, and the rest happens according to a textbook.

That's a good analytical approach.
UPDATE: I'm going to put my original continuation below the fold. I've thought more since.

OK, so now we have a story that we can put into a textbook macro framework (I'll play that game for now). We've had an increase in liquidity preference. Why, if we don't act this weekend, will we get the Great Depression? I don't think that what has happened so far is going to do it. Do we really think that the lagged response of spending to the increase in liquidity preference that has already taken place will be that large? I don't.
What does "increase in liquidity preference" mean? Is a bank run an "increase in liquidity preference?" Can one distinguish between an "increase in liquidity preference" and an "increase in solvency preference", or even an "increase in not wanting to see all my money vanish preference?" Muddled language reflects muddled thinking.

One clear description comes from U Chicago's Bob Shimer. He acknowledges the potential of information assymetry being a potential problem in the MBS market (I don't agree that this is a big issue) and points out that TARP does not fix that:
This program does not solve the lemons problem. The government purchases a lot of lemons at an inflated price. This improves the balance sheet of the firms that can sell their worst securities. It also improves the balance sheet of firms that own better securities because the market price of those securities will increase. (Of course, it cannot increase too much, or no one would sell to the government. They would wait to sell at the higher market price. I have not worked out the equilibrium of an auction with an option to resell later. It seems complicated.) But this is fundamentally no different than giving taxpayers' money to owners, managers, and debt-holders of firms that made the worst decisions.
He also points out, albeit in passing, the key problem
What else can the government do? First, it can establish stable rules and play by them. Holding out the possibility of distributing vast sums of money in an unspecified manner does not help market participants value the securities or value the firms. Second, it can prevent panics, i.e. Diamond-Dybvig bank runs. This is what it did when it offered insurance for money market mutual funds, an important source of funding in the commercial paper market. So far, that market appears to be holding up. Third, it can reduce the risk that its current actions encourage future misbehavior. We have already seen evidence of moral hazard in these markets, for example in AIG's decision to turn down a $8 billion offer from J.C. Flowers during the weekend before AIG collapsed.
The informality of the Paulson plan is toxic, and it is why I prefer outright nationalization. If we extend FDIC to non-commercial banks, then it's easy just to get rid of fractional reserve banking altogether, keep deposits in vaults, and just have the US Government lend to firms directly. There is, in fact, one stable equilibrium in the Diamond-Dybvig model and it is the bank run state. Prices, if left to their own devices, will settle at their term matched state.

Looks like the plan is dead. I hope it stays that way.

Friday, September 26, 2008

Micro is from Earth, Marco is from Cuckoo land

I attribute Steve Landsburg's complete failure to understand the credit crises to the fact that he's well versed in microeconomics (which is sane) and therefore feels that macroeconomics is too (which it is not -- it is fundamentally nonsensical)
That's one reason I feel squeamish about the official pronouncements we've been getting. They tell us bank failures will make it hard to borrow but never that bank failures will make it hard to lend. But every borrower is paired with a lender, so it's odd to state the problem so asymmetrically. This makes me suspect that the official pronouncers have not entirely thought this thing through.

In the 1930s, a wave of bank failures did make it hard for borrowers and lenders to find each other, and the consequences were drastic. But times have changed in at least two relevant ways. First, the disaster of the 1930s was caused not just by bank failures, but by a 30% contraction of the money supply, which is something today's Fed can easily prevent. Second, as any user of match.com can tell you, the technology for finding partners has improved since then. When a firm wants to raise capital, why can't it just sell bonds over the web? Or issue new stock? Or approach one of the hedge funds that seem to be swimming in cash? Or borrow abroad?

Steven, in our fractional reserve system, every borrower is NOT matched to a lender. A borrower who wants a long term loan is NOT matched to a lender who wants some place to stash their money for 30 years. Instead, short term lenders (who aren't really lenders at all -- they think their money is in cash) are, through the magic of term transformation, converted to nine times that number of long term lenders. This is an inherently unstable state of affairs and will end in a bank run, which is what we are seeing in the shadow financial system. When these banks vanish, we will see a world where every borrower IS matched to a lender, and it will be very different from this world. Key differences:
1. Long term rates will be MUCH higher, as the current supply of long term money is created almost entirely through term transformation. And only banks can term transform.
2. A much smaller money supply, as this 9x multiplier (which, if it flows through another bank, gets an additional 9x boost, ad infinitum) vanishes. This will be true even if the Government steps in and keeps M0 whole. At this point, the US Government can always run the printing press to create M2 etc., but banks failing are inherently deflationary because banks are inherently inflationary.

Make that the second biggest

John Berry from Bloomberg contends that the Paulson Plan would be the biggest carry trade in history.
The government will get the $700 billion by selling a range of Treasury securities to the public with yields of 3 percent to 4 percent. With investors around the world clamoring to buy risk-free Treasuries, the market should be able to absorb the jump in supply without a significant increase in yields.

Contrast that with likely yields on the troubled assets for which there currently is no market. No one can be sure how big a haircut there will be on the assets Treasury buys, though if it's 50 percent or more, their yields should be 10 percent or higher.

That is, the government will be borrowing at 3 percent to 4 percent to buy assets yielding 10 percent or even 12 percent. Conservatively, that spread on an investment of $700 billion should generate income of $40 billion to $60 billion annually.
Got that -- since the US Government can currently borrow cheaply (3%), it will make money if it buys any asset that yields a higher interest rate. This is awesome -- the Government should buy ALL assets, make money off the carry, and then write checks to us, the taxpayers.

It also does something like this in the retail banking sector, where FDIC insurance means that the current system is indistinguishable from deposits sitting in a vault, and the banks borrowing directly from the Government (at 3%) and lending out at more (10%+). If we're to have fractional lending with a Government guarantee, we should really just have the Government make all loans.

Quick Links

Michael Lewis: America Must Rescue the Bonuses at Goldman Sachs (funny). Michael Lewis really is the Jeremy Clarkson of the business world.

Johnny Debacle: The Bailout, a play Too many great lines to pick just one. Paulson is fantastic.

Greg Mankiw: If I were a member of Congress... I would listen to Ben Bernanke. Disappointed by Mankiw's Appeal to Authority.

Understanding Pinata

Watching Tyler struggle with the credit crisis is like watching a blindfolded man swinging at a Pinata. He sometimes gets close, but then he's just flailing at the air again. His latest swing:
The proximate cause of the credit crisis (as distinct from the housing crisis) was the interplay between two choices made by banks. First, substantial amounts of mortgage-backed securities with exposure to subprime risk were kept on bank balance sheets even though the “originate and distribute” model of securitization that many banks ostensibly followed was supposed to transfer risk to those institutions better able to bear it, such as unleveraged pension funds. Second, across the board, banks financed these and other risky assets with short-term market borrowing.
In other words, one problem was not enough (!) securitization. They also call for counter-cyclical capital requirements. They like mandatory capital insurance -- with payments triggered by capital disasters -- even better. My main worry, of course, is how we should regulate (or not) the entities which offer this insurance. Will they too engage in liquidity transformation and if so who ensures them?

And, going back to banks, part of the governance problem was this:

...it is very hard, especially in the case of new products, to tell whether a financial manager is generating true excess returns adjusting for risk, or whether the current returns are simply compensation for a risk that has not yet shown itself but that will eventually materialize.
Their phrase "recapitalization as a public good" should not soon be forgotten. And it is leverage which is dangerous, a lesson becoming clearer every day.
Sadly, Tyler just keeps being wrong in exactly the same way again and again and thinks he is making progress. This is particularly interesting in his latest post because the paper he's refering to actually nails the ultimate cause:
Second, across the board, banks financed these and other risky assets with short-term market borrowing.
Tyler sadly ignores this because it describes the entire world banking system. While the proximate cause of this crises is bad mortgages and insufficient securitization, there have been crises in the past that look exactly the same from a credit perspective, but have different proximate parameters. Instead of focusing on the most recent horseless barn, perhaps we could consider that fact that the barn, indeed all barns, are missing an entire wall. Go back to the mock balance sheet, you'll see that equity is $5 and debt is $15 -- so 3x debt leverage, but that $5 of equity is floating $100 of price-unstable, mismatched securities! Even if GS had 0 debt (which is traditionally what I've understood "leverage" to mean) it would still have a much larger balance sheet that it's equity could support in the event of a bank run. Reducing operating leverage in banks will have no effect, because the difference in value for long term assets when maturity transformation works, and when it does not, is massive, and will overwhelm any possible equity cushion.

Thursday, September 25, 2008

Not quite there

Bryan Caplan struggles to understand the credit crises:
My conjecture: If Paulson's bail-out were funded by a permanent tax increase sufficient to raise $700B in present value terms, it wouldn't stand a chance. The minimal public outcry, therefore, hinges on "debt illusion" - the mistaken view that debts, unlike taxes, never really have to be paid.

Am I right?
Caplan is wrong, and suffering from the "'debt illusion' illusion". Debts never really have to be paid, they can be defaulted on, or inflated away. Caplan does not fully realize that the dollar is a fiat currency. But, people who care about such things, have totally noticed.

Arnold is struggling too.
1. It's not the 1930's.

In the 1930's, the economy had many fewer industries than it has today. If the auto industry slumped, there was no computer industry to pick up the slack. Much of the economy was still agricultural, and the farm economy was in the process of transition. The dust bowl and advances in transportation displaced many farmers, who joined the ranks of the unemployed.

Today, we have a very diverse real economy. Some sectors are in a downturn, but many other sectors are not. The ability of any one sector to bring down the rest of the economy is much reduced.
Except when that sector if the financial sector, because the financial sector controls money supply. Suppose m1 collapsed to m0 because the financial sector folded up all the credit it had been extended. Does Arnold really believe it would have no impact on the real economy? If anything, today's economy is more susceptible to credit crises simply because more of the economy runs on credit. It gets worse.
2. The bailout blends finance with government.
Pray tell what does the Fed do? And the Treasury? And the SEC? FDIC? Are they part of the Government?Are they part of the Financial system? Also, who has been buying US Treasuries, and holding the long end of the yield curve down? Is that part of the Financial system? Are those buyers part of A government (albeit not THE Government)? The current financial system is entirely a product of regulation. The financial system IS part of the government--that's what fiat currency means.

Wednesday, September 24, 2008


The most excellent Steve Waldman blew my mind today in a response yesterday's question about whether the debt of an insolvent financial institution had any value or not.
But I am all for converting counterparty liabilities to equity (along the lines suggested in my post, define $1 stock and let the firm pay some of its obligations in stock), if necessary. I'd start with straight, long-term, unsecured debt (obviously the most-equity like), but I'd go as far as possible up the chain as necessary.
As Keanu said, "whoa".

What Steve's suggesting is that not only do the bondholders get converted to equity, but all the entire liability side of the balance sheet does (in some order)! So, going back to our hypothetical insolvent bank, GS:


Good assets: $75

Assets gone bad: $0 (written off)



Liabilities to customers/counterparties: $80

Debt to bondholders of company: $15

Shareholder equity: -$20


Steve would convert all bondholders AND some liability holders to equity holders until the desired amount of capitalization was reached. Say we think 3x leverage is "right". The final balance sheet would look like this:


Good assets: $75

Assets gone bad: $0 (written off)



Liabilities to customers/counterparties: $50 (37% haircut)

Debt to bondholders of company: $0 (wiped out)

Shareholder equity: $25


I have no idea what the nature of "Liabilities to customers/counterparties" is -- some might be brokerage accounts which have some government protection, so there the US Government would take the equity share and pay out the customer in cash, others may be other banks who would now own a slice of GS, etc. The Zingales plan explicitly talked about debt-for-equity swaps, not liability-for-equity swaps, so I'm pretty sure they were not talking about the entire right hand side of the balance sheet (if I'm wrong, please let me know). So I think that the Zingales plan is still a non-starter, because balance sheets are worse than they imagine, but if you're going to be open to converting *everything*, then banks can still re-capitalize themselves and yes, high leverage makes this easier (as there is more to convert).

Two additional thoughts:

1. Who the counterparties are is important. If they include the Chinese, then this could remove the last prop supporting the Greenback as China decides to stop its lender financing program for US Consumers. This would be a Significant Event as their generosity has been a powerful countervailing force against the deflation the US is currently experiencing.

2. This plan re-capitalizes banks, but it does not inject more money into the financial system. Balance sheets still shrinking, so the money supply is deflating. Paulson and Bernanke would have to inflate (if they want to do that) via another route. I would recommend mailing $1M checks to households.

As my bleg worked out so well yesterday I'll try another. Can anyone tell me who the counterparties are, and what liabilities GS might be holding?

Tuesday, September 23, 2008

Is the debt worthless?

A number of readers have sent me notes saying that a debt for equity swap would help re-capitalize insolvent (and illiquid) financial institutions, contra my assertions here and here. I've outlined my thinking below -- let me know what I'm missing.

Here's an example (riffing on a simple model by Hussman)

Take a simple balance sheet for a hypothetical bank called, say, GS. It's 20x levered, and suffered losses of 5x equity.


Good assets: $75

Assets gone bad: $25 (but not yet written off)



Liabilities to customers/counterparties: $80

Debt to bondholders of company: $15

Shareholder equity: $5


So, they are leveraged 20x (floating $100 of assets on $5 of equity), and they have losses 5x equity ($25 of assets that are actually worth zero, but have not yet been marked to zero).

Say they do that writedown, and bad assets are now zero. Their balance sheet shrinks to $75. The liabilities side shrinks by equity going from $5 to -$20. This bank is insolvent. Here's the new balance sheet.


Good assets: $75

Assets gone bad: $0 (written off)



Liabilities to customers/counterparties: $80

Debt to bondholders of company: $15

Shareholder equity: -$20


The Paulson plan takes those $25 of bad assets, and substitutes them for $25 of good assets (or maybe $20 of good assets). The point is, for the bank to re-capitalize, he *HAS* to pay more than the assets are worth. By swapping good for bad, the balance sheet does not have to contract, and part of the Paulson/Bernanke plan is to combat deflation, and keep balance sheets from shrinking so lenders can keep borrowing.

Zingales, to my understanding, says that instead of public gifts of cash, bondholders should have their debt converted to equity. If shareholder equity was say, -$10 (GS did not lose that much money) then this would work, and such a debt to equity conversion would reduce debt to $5, bring equity to $0, and debtholders would take an immediate 67% haircut and now have participation in any upside. I can see how this works if debt was larger than negative equity, but I don't think that's the case as firms were so leveraged, and floated on equity cushions that were so tiny. At any rate, I can see how that *might* be the case, which, to my understanding, would render the Zingales option as not feasible.

I also cannot see GS reducing liabilities by going after that "counterparty" line item, as that would trigger exactly the kind of contagion that Paulson & Co. wants to avoid.

Finally, to keenly illustrate how bank runs work, look at the Good Assets line item at $75. In maturity transformation stops working, the value of these fall to pennies on the dollar -- remember, maturity transformed long term debt has two stable prices. Pure liquidity interventions work by bringing this back up to the "maturity transformation working" price, rendering the firm solvent.

Please let me know if I'm getting something wrong.

Monday, September 22, 2008

Getting warmer

James Surowiecki has a nice piece in the New Yorker where he almost, but not quite, gets that, in a fractional reserve banking system, liquidity = solvency for financial firms:
That’s because the entire edifice of Wall Street is built on confidence. Investment banks rely on short-term debt to run their businesses, and their businesses consist of activities—trading, dealmaking, money management—that depend on people’s faith in their ability to honor their obligations. As soon as the customers and creditors of a company like Lehman start to wonder whether it might collapse, they become less willing to lend or to trade, and more likely to demand their money back. The perception of weakness exacerbates the reality of weakness. And although there are myriad measures of a company’s health, nothing looks scarier than a stock price that’s heading toward zero.
I don't like the word "confidence" when describing the technical willingness to rollover short term debt, because it makes it sounds as if the problem can be solved by some trust building exercises, perhaps some group singing, instead of highlighting that the system is fundamentally, unpredictably unstable, and that there are two prices for long term maturity mismatched securities, not one. This isn't a problem of confidence in the same way that a pencil refusing to balance on a point isn't a problem of "gravity". In retail banking, the problem was solved by essentially having the government lend to banks while they kept depositor's money in the vault, but there is no formal solution to this in the investment banking world.

Dean Baker usefully assembles a load of anodyne pablum and labels it the Progressive conditions for a bail out. Having been clearly labeled, it is now easy to avoid. The Reactionary solution is in the comments, here, and it's quite possible moldbug would include some public beheadings to deal with the Moral Hazard problem.

The Dodd plan comes closer to my preferred solution -- complete nationalization of the financial sector -- by having warrants that convert to an equity stake if taxpayers lose money (which they must, since this is a transfer of capital) AND recognizing that both the debt and equity for financial firms is worthless, so a straight Zingales style swap will not help matters. The Treasury does not want taxpayers to be protected, and is rejecting the Dodd proposal. I hope we are all clear now on who really runs things in DC.

Sunday, September 21, 2008


Yves has the inside line on the $700B bailout. It's, as is obvious, a mechanism to recapitalize investment banks using taxpayer money (with no taxpayer ownership). Banks need to be recapitalized, as they are insolvent.
Anyway, I wanted to let you know that, behind closed doors, Paulson describes the plan differently. He explicitly says that it will buy assets at above market prices (although he still claims that they are undervalued) because the holders won't sell at market prices. Anna Eshoo pressed him on how the government can compel the holders to sell, and he basically dodged the question. I think that's because he didn't want to admit that the government would just keep offering more and more.
If the government does not buy these assets at above market price, the Paulson plan is pointless, because it does nothing to address the fundamental insolvency that is at the heart of the financial crises.

On the topic of fraud, here's a NYTimes article on the wealthy, disabled, and retired LIRR employees. Note that the word "fraud" appears nowhere in the text.

Saturday, September 20, 2008

Quick links

U Chicago prof. Luigi Zingales is unimpressed with the Paulson bailout of the financial sector.
As during the Great Depression and in many debt restructurings, it makes sense in the current contingency to mandate a partial debt forgiveness or a debt-for-equity swap in the
financial sector. It has the benefit of being a well-tested strategy in the private sector and it leaves the taxpayers out of the picture. But if it is so simple, why no expert has mentioned it?

The major players in the financial sector do not like it. It is much more appealing for the financial industry to be bailed out at taxpayers’ expense than to bear their share of pain.
But what good does debt for equity swaps do when the firm is insolvent, and both debt and equity are worth zero?

Also, a warning to those who think "more regulation" is the answer. Financial services is not an unregulated industry, and it this thicket of regulations that gave rise to the complex derivatives we see blowing up. An old post my Mindles Dreck states:
A CBO is just one example of a credit rating-driven transaction, but most of them achieve the same thing - they decrease frequency of loss but increase the severity. So they blow up infrequently, but when they do it's often a big mess. Ratings-packaged instruments are less risky than the pool of securities they represent but often riskier and less liquid than the investment grade securities for which they are being substituted. As a result, they pay a yield or return premium (even net of high investment banking fees). That premium may or may not be enough to pay for their risk. But they pass the all-important credit rating process and are therefore sometimes the only choice for ratings-restricted portfolios reaching for yield.

...[Frank] Partnoy is a former derivatives salesperson, and he clearly suggests that regulation is often the derivative salesman's best friend. Complicated rules encourage complex transactions that seek to conceal or re-shape their true nature. Regulated entities create demand for complex derivatives that substitute proscribed risks for admitted risks. If a new risk is identified and prohibited, the market starts inventing instruments that get around it. There is no end to this process. Regulators have always had this perversely symbiotic relationship with Wall Street. And the same can be said for the ridiculously complicated federal taxation rules and increasingly byzantine Financial Accounting Standards, both of which have inspired massive derivative activity as the engineers find their way around the code maze.
FInally, Krugman continues to demonstrate just how far behind the eight-ball he is.
The Treasury plan, by contrast, looks like an attempt to restore confidence in the financial system — that is, convince creditors of troubled institutions that everything's OK — simply by buying assets off these institutions. This will only work if the prices Treasury pays are much higher than current market prices; that, in turn, can only be true either if this is mainly a liquidity problem — which seems doubtful — or if Treasury is going to be paying a huge premium, in effect throwing taxpayers' money at the financial world.

And there's no quid pro quo here — nothing that gives taxpayers a stake in the upside, nothing that ensures that the money is used to stabilize the system rather than reward the undeserving.

I hope I'm wrong about this. But let me say it again: Treasury needs to explain why this is supposed to work — not try to panic Congress into giving it a blank check. Otherwise, no deal.
The financial system is insolvent, and even if maturity mismatch turns back on (liquidity) the underlying solvency issue is still a problem. To re-capitalize the system, the government needs to take money from those who have it, and give it to those who don't. The Paulson plan does exactly that.

Friday, September 19, 2008

Liquidity = Solvency

I'm struck by how, for financial firms in a fractional reserve system, there really is no difference between "liquidity" and "solvency". It's hard to nail down what "liquidity" means. With all this talk of dollar injections I imagine a sort of green juice in a giant hypodermic needle that will jab the rump of the financial system -- wherever that is -- and somehow make everything good again. But this is bogus.

The most meaningful sense of the word "liquidity" is the ability to use short term deposits as collateral to make long term loans. This is at the heart of fractional reserve banking, and it is a system that is inherently unstable, like a naked singularity, or a pencil balanced on its point. It is inherently unstable because it has two equilibria: 1) everything works (lenders do not pull out their deposits en masse) and 2) the bank collapses, because there is a rush for the exit. The slightest stochastic fluctuation can trigger a switch reversal from state 1 to state 2. There is no single, Nash equilibrium. The switch from 1) to 2) is discontinuous and chaotic -- it falls into Taleb's fourth quadrant -- so it cannot be effectively hedged against. FDIC insurance does not eliminate this liquidity risk, it just takes the depositor side of it and moves that to the Government. As Mencius Moldbug points out in this excellent discussion:
our present deposit system is equivalent to one in which Y (the bank) stores its checking deposits as cash in the vault, and then borrows money from X (the Fed) to make its 30-year mortgage loans.
In other words, given the inherent instability of fractional reserve banking, de facto, the government carries all loans on its balance sheet. This is why I applaud the Paulson plan of having the US Government buy all bad mortgage debt -- it formally acknowledges reality.

Brad DeLong, when he isn't mindlessly shrieking his devoutness and piety, has some useful graphs that show this two-state equilibria in action.

The quantity of risk assets is fixed (at least in the short term -- these would be the 30 year mortgages) but their price can get locked into a high or low equilibrium state. The high state corresponds to maturity transformation working, and the low state corresponds to a bank run. Remember -- bank runs can happen for NO REASON. So, while Brad gets the two-state equilibrium right, he gets the causality wrong:
But there is another mode of operation: if financial intermediaries are poorly-capitalized they themselves will have great problems borrowing--savers will fear the moral hazard problems that arise when those who manage their money don't themselves have a large stake in the game, and a financial intermediary without a large equity cushion leads savers to ask the American question "if you're so smart, why aren't you rich?" and shy away. So if financial intermediaries are poorly-capitalized, supply and demand looks very different [and you switch to the bank run state]:
Here, Brad says that a lack of solvency triggers the bank run, but we know that the bank run can start spontaneously -- ie. for no reason at all. So, bank run = maturity transformation stopping = "liquidity dries up" = lower price of long term risk assets because risk assets have one price when maturity transformation works, and another when it does not. In other words, embedded in the price of a long term risk asset, is the value of maturity transformation. The ability of maturity transformation makes long term risk assets seem great and at the same time permanently destabilize and undermine the system, paving the way for its collapse. If this makes you think of cocaine, you are not the only one.

Think of this another way -- how much would a 30 year CD need to pay to make you happy to sign away $ for three decades? Would 4% cover it? If maturity transformation turned off, and 30 year bonds really needed to be held for 30 years, they would need to be sold for pennies on the dollar. Maturity transformation creates the feedback loop between the price of risk assets and the willingness of short term depositors to lend, so the liquidity of a financial institution and its solvency are one and the same.

Steve Waldman, who I think the world of, also gets this relationship wrong, I think.
There is no question that we are going to spend a lot of public money to address the current crisis. We have already put a very extraordinary amount at risk. The question we should be asking is not whether or how much, but to whom and for what. The financial crisis we are facing is a symptom of a much larger economic and social crisis. Wall Street is not the source of the pain. On the contrary, the financial sector has been put this decade primarily in the service of hiding, literally of papering over, unsustainable trends in the current account, income distribution, human and physical capital deterioration, and the sectoral composition of the American economy. The conventional wisdom is that this is a financial crisis, and that so far "Main Street" has been largely insulated from the catastrophe. That is rubbish. The cancer is on Main Street, and the tumor has been growing there for years. Wall Street provided drugs to hide the pain and keep us going, palliative but not curative. What is happening now is those drugs are wearing off. The American economy is fundamentally unsound, and has been for some time. We would have noticed sooner, were it not for financial methamphetamine conjured by mad scientists in lower Manhattan from a whirlwind of foreign central bank money.
I completely agree that the American economy is unsound, but I believe the root causes are strictly monetary. American's are good at business. But we are lousy at controlling the money supply. And this is the difference between asset bubbles and credit bubbles. Asset bubbles debauch the asset class when they ultimately pop, while credit bubbles debauch the currency -- and therefore the economy as a whole. Greenspan's actions to treat the internet bust in 2000 as if it was a credit bubble is going down in history as one of the worst decisions by a central banker, ever. His conversion from Maestro to Goat is complete.

To wrap up an overly long post, it is good that the Government is recognizing its role as lender for risk assets, and simply taking all the loans onto its balance sheet effectively nationalizing the mortgage market, and perhaps all of the financial system. However, their actions will be focused on propping up what needs to be propped up to get us back into the maturity transformation game. Regulations around mortgages, investment banks, CDSs etc. are meaningless because the inherent instability in the system will simply find a new avenue to express itself, resulting in a larger liquidity implosion in the future. The world will not get off the dollar until China decides to get off the dollar, and the PBoC, as a matter of policy, does not seem interested in doing that this time.

RIP Mr English

O Mister English
You mixed coffee and Japan
I'm sad you are gone

Thursday, September 18, 2008

Still in the dark

This article by U Chicago professors is supposed to explain the current market turmoil. Unfortunately, it makes no sense to me, but your mileage may vary. First Fannie/Freddie:
Fannie and Freddie were weakly supervised and strayed from the core mission. They began using their subsidized financing to buy mortgage-backed securities which were backed by pools of mortgages that did not meet their usual standards. Over the last year, it became clear that their thin capital was not enough to cover the losses on these subprime mortgages. The massive amount of diffusely held debt would have caused collapses everywhere if it was defaulted upon; so the Treasury announced that it would explicitly guarantee the debt.
Does this make any sense to you? Fannie and Freddie were always government sponsored entities. The US Government always guaranteed their debt. Which is why it traded as if it was a Treasury bill. Why do Diamond and Kashyap ignore the most salient point about these two entities? Also, if you have a financial institution that borrows short and lends long, won't it eventually experience a bank run? Wasn't it inevitable that eventually, the Government would have to make its implicit guarantee explicit? What part of the Government backing up Fannie and Freddie debt was not inevitable, given that the Government backs up Fannie and Freddie debt?

Now, Lehman

Lehman’s demise came when it could not even keep borrowing. Lehman was rolling over at least $100 billion a month to finance its investments in real estate, bonds, stocks, and financial assets. When it is hard for lenders to monitor their investments and borrowers can rapidly change the risk on their balance sheets, lenders opt for short-term lending. Compared to legal or other channels, their threat to refuse to roll over funding is the most effective option to keep the borrower in line.

This was especially relevant for Lehman, because as an investment bank, it could transform its risk characteristics very easily by using derivatives and by churning its trading portfolio. So for Lehman (and all investment banks), the short-term financing is not an accident; it is inevitable.
Isn't borrowing short to lend long a recipe for a bank run? Isn't that why we have FDIC insurance? If this structure is inevitable for investment banks, then aren't bank runs inevitable for investment banks? And why is it so critical for banks to maturity transform at any rate? Why can't we have a financial system with no maturity transformation, and hence, no liquidity risk?

Finally, AIG:

The credit-rating agencies looking at the potential losses downgraded A.I.G.’s debt on Monday. With its lower credit ratings, A.I.G.’s insurance contracts required A.I.G. to demonstrate that it had collateral to service the contracts; estimates suggested that it needed roughly $15 billion in immediate collateral.

A second problem A.I.G. faced is that if it failed to post the collateral, it would be considered to have defaulted on the C.D.S.’s. Were A.I.G. to default on C.D.S.’s, some other A.I.G. contracts (tied to losses on other financial securities) contain clauses saying that its other contractual partners could insist on prepayment of their claims. These cross-default clauses are present so that resources from one part of the business do not get diverted to plug a hole in another part.
So, AIGs liquidity insurance was worthless because when it needed to supply others with liquidity, it could not access liquidity itself.

Most nonsensically:
6) What does this mean for the markets going ahead?

Letting Lehman go means that the remaining large financial services firms now must understand that they need to manage their own risks more carefully. This includes both securing adequate funding and being prudent about which counterparties to rely upon. Both of these developments are welcome.

If the remaining investment banks, Goldman Sachs and Morgan Stanley, do not get more secure funding in place, they may be acquired or subject to a run too. In the current environment, relying almost exclusively on short-term debt is hazardous, even if a firm or bank has nothing wrong with it.
"In this current environment, relying almost exclusively on short-tern debt is hazardous?" What makes this current environment so special -- that it's on planet earth? A fractional reserve financial system is hazardous in *any* environment, because it can experience a run for *no reason*, at *any time*. That is an inescapable feature of fractional reserve banking. If you define liquidity as "the ability to make long-term loans against short-term deposits, secure in the knowledge that you can roll those short-term deposits over", then bank runs, and systematic, contagious bank runs, are an inevitable feature of the system. If you put a lender of last resort behind the whole system (who has a printing press) then you will end up nationalizing the financial system those years that it has a run. This is no way to run a railroad.

Nowhere in the article do Diamond and Kashyap explain that bank runs are inevitable in a fractional reserve banking system. And no where do they explain why we need to run on a fractional reserve system, and why we cannot have a system where for every borrower, there is a single lender, and their maturities are matched. In other words, a world of fully banked cash accounts, and CDs. "Banks cannot make much profit" is not an acceptable reason.

If the US Government simply nationalized all bad debt, they will "solve" this crises, and set us up for the next one in 5-10 years, which will be even worse. Regulation in the CDS and mortgage lending world is not helpful, as the shadow financial system will find new instruments to borrow short and lend long. The problem is the "borrowing short and lending long", not sub prime, or CDS, or SIVs, or CDOs. But I can bet that the God Given right for banks to maturity transform will emerge from this crises unscathed, and begin setting up the next, bigger, one.

Wednesday, September 17, 2008

Hail Paulson

Megan asks how the Fed has any authority to do what it's been doing (as does Paul Volker)? I think her conclusion is right on: something needs to be done, and no one else is interested in stepping up to the plate. This NPR interview is also bogus -- we're talking about actual governance here and elected officials have no business being any part of it. Caesar knows who is Caesar, Circus should know it is Circus.

In the pas de deux between the Treasury and the Fed, it's interested to see Paulson eclipse Bernanke. I guess if you mix a bunch of ex-bankers with a bunch of academics, it's clear who will end up on top. So the Treasury will give money to the Fed, so the Treasury can then give it to Wall Street.

Monday, September 15, 2008

Lehman, RIP

I don't have too much to add regarding the bankruptcy of Lehman Bros. I'm not sure anyone really knows what the consequences are, but they will make themselves clear in due time. I hope that it shifts the conversation from about "liquidity" to one about "solvency" as that will accurately reflect the reality of what is happening the financial sector. As I've posted earlier, the US has about $6T in losses from the credit bubble, and the question is who will get stuck with the bill. Since various US Agencies are involved, the US taxpayer is a candidate, and I'm sure that banks are aggressively lobbying for that solution.

Currently, money is being destroyed (through the collapse of credit) than the Fed is pumping it out, so the overall economic story remains one of deflation.

Friday, September 12, 2008

You cannot get there from here

Arnold suggests that "instead of handing a low-income home buyer a subsidized mortgage with a low down payment, I'd rather hand the buyer a check to use for part of the down payment." This exists -- they're called Downpayment Assistance Programs (DAP) and they are a powerful font of corruption that does not help poor people at all.
Those who claim that DAP loans provide a benefit to borrowers without funds are making no sense even if you grant that making loans to people without even minimal skin in the game is a good idea: the DAP programs simply keep contract sales prices inflated, channel fees into the pockets of "nonprofits" who provide no other service than laundering money, and result in lower insurance premiums than FHA should be getting for loans with riskier profiles. If you care at all about the long-term survival of the FHA program, you would be doing everything you can to protect it from this kind of damage.

Supporting DAPs means supporting property sellers--particularly but not limited to builders and developers--and the "entrepreneurs" who form "nonprofits" to extract fees from naive homebuyers, not to mention loan originators who pocket higher commissions, with the risk being carried by government insurance. It is, precisely, the kind of sleazy, conflict-ridden, self-serving "initiative," overtly "faith-based" or its sort-of secular equivalent "dream-based," that thrives in an environment where regulation is dismantled or unenforced and "government" is bashed with one hand and milked with the other.
Maybe, instead, we should just accept that the Universe is a Tragic place where people who cannot afford to purchase homes, cannot afford to purchase homes.

Tuesday, September 09, 2008

Rubbish Roubini Column

Nouriel has strayed from his usual beat of ultra-bearish economics and wandered into politics. Oh dear. Maybe his high profile appearances on the NYTimes and elsewhere have gone to his head. It's always ugly when that happens. Anyway, this latest column is a dog.
Today instead the US has performed the greatest nationalization in the history of humanity. By nationalizing Fannie and Freddie the US has increased its public assets by almost $6 trillion and has increased its public debt/liabilities by another $6 trillion. The US has also turned itself into the largest government-owned hedge fund in the world: by injecting a likely $200 billion of capital into Fannie and Freddie and taking on almost $6 trillion of liabilities of such GSEs the US has also undertaken the biggest and most levered LBO (“leveraged buy-out”) in human history that has a debt to equity ratio of 30 ($6,000 billion of debt against $200 billion of equity).
Rubbish. Fannie and Freddie were always part of the US Government -- that's how they operated and that's why they were able to issue debt at par with Treasuries. All this has done has formalized what was a widely accepted and only slightly informal arrangement. Therefore, the US Govt has not changed its public assets or liabilities by one iota. They are exactly what they were before, because nothing has changed.

Also, Nouriel claiming that the US has turned into the world's biggest hedge fund is also rubbish. A country that has obligations far in excess of its productive capacity and a monetary system leveraged to the hilt has been an enormous hedge fund for quite a while.

Saturday, September 06, 2008

Mr 10%

In other news, the most crooked man in all of Pakistan is now PM. The State Dept must be thrilled -- they have a new project.


The US Government is finally going to nationalize Fannie Mae and Freddie Mac, formalizing a relationship that has been clear since the inception of those two firms. Both Brad DeLong and Chris Bowers do not get it. First Chris:
The problem I have is with the incredible cognitive dissonance surrounding "big government" in our national political discourse. Even as we have reached national consensus on nationalizing industries, which is the literal definition of socialism and big government, politicians of every party keep talking about "small government" as though it were a virtue. I mean, the day after the Republican convention, which included countless attacks on big government, the Republican administration goes out an nationalizes a major industry.
Fannie and Freddie were always parts of the Federal Government. The mortgage industry is as "socialized" as it ever was, and this move has not changed the size of the government one iota. Now Brad:
I don't see the necessity for nationalizing Fannie and Freddie right now. They both are still cash-flow positive, right? If they fail to rollover their bonds and become cash-flow negative, the Treasury can finance them with preferred plus warrants, right?
Wrong. They need to roll over their bonds at a price that treats them as being part of the government, not any price, for them to remain in business. The ambiguity over that position has finally been cleared up, exactly as everyone knew it would.

Thursday, September 04, 2008


For reasons that I never understood, my family prefers to send items via friends and relatives instead of posting them. A new service in the UAE gives us all the chance to be couriers.