Remembering the Suits vs. Geeks Divide

I’ll provide a post-mortem on my appearance at this panel on whether or not to break up the banks after I’ve had more time to reflect.

Prior to the panel, I Googled one of the other panelists, and I found that he had hopes for the Volcker Rule, which would try to keep banks from doing proprietary trading. I am not a fan of the Volcker Rule. In fact, in recent years, I have not been a fan of Paul Volcker, because I think he has what I call a low geek quotient.

What I call Geek Finance has emerged over the past thirty years. It is used extensively in derivatives markets and in mortgage finance. It involves very complex probabilistic simulation models that are used to assign values to long-term, deep out-of-the-money options. Some thoughts.

1. Is Geek Finance a good thing? On the one hand, I would say that we need some rational way of valuing these sorts of options. On the other hand, it is important to be aware of the assumptions that go into such models and not to have too much faith in their precision. In the case of mortgage default risk, for example, it matters whether you assume that house prices across different locations are highly correlated or nearly independent. It matters whether you assume that a large nationwide house price decline is practically impossible or just somewhat unlikely.

2. Shortly after the financial crisis, Robert Merton, who shared the Nobel Prize for developing option price theory, gave a lecture in which he suggested that many top corporate executives did not really understand what was being done by the practitioners who worked for them. I termed this the Suits vs. Geeks divide. Many CEOs, and also many top officials in Washington, had low geek quotients.

3. Whether you love or hate geek finance, whether you want to tolerate it or would seek to get rid of it, you have to understand it. I think that Suits with low geek quotients are dangerous.

4. In 2003, Freddie Mac’s Board ousted a CEO with a high geek quotient and replaced him with a CEO with a low geek quotient. The main change that resulted from that was that Freddie Mac greatly increased its exposure to risky loans.

5. In 2006, Goldman Sachs lost a CEO with a low geek quotient. Subsequently, and this may have been purely coincidental, Goldman was relatively good at reducing its exposure in the mortgage market.

6. The original idea of TARP, which was to use government funds to buy toxic assets, had a low geek quotient. As a geek, I did not think it was workable. Of course, this idea was never implemented. Instead, the TARP money pile was used to inject capital into banks, to restructure GM and Chrysler, and ….well, whatever the President and Treasury Secretary felt like spending it on, it seems.

7. Back to the Volcker Rule. I don’t think it can fly. My prediction is that they will get it off the ground to save face, then it will wobble at low altitude for a bit, and within a few years you will find it resting on the ground. I imagine a sequence of conversations going something like this:

Volcker rulers: Banks, you cannot touch securities.

Banks: But you do want us to hedge our risks, don’t you?

Volcker rulers: Hmmm. OK, but you have to hold your hedging instruments until they mature.

Banks: But when interest rates change, you do want us to rebalance, don’t you?

Volcker rulers: Hmmm. OK, but…

etc., etc., until there is no rule left

People, Countries, and Debt

James Kynge writes about China,

Total debts owed by the government, companies and households have ballooned to 240 per cent of gross domestic product, virtually double the level at the time of the global financial crisis.

This ratio, it is true, remains modest next to some in the west; US debts stand at 322 per cent of GDP, Ireland’s at more than 400 per cent, while Greece and Spain are at about 300 per cent each.

Pointer from Tyler Cowen.

To me, it seems careless to add up the liabilities of government, companies, and households. I object to taking the sum of these numbers and saying “China owes ___.” The debt is not owed by an entity called the country of China. It is owed by disparate entities within the country of China. And much of it is owed to disparate entities within China.

Speaking of which, it also seems careless to ignore assets. Suppose somebody has a $300,000 mortgage and a $30,000 income. You might say, “wow, their debt is 1000 percent of their income!” But that is not so alarming if they have $1 million in assets (maybe the house itself is worth $1 million).

I’m not trying to dismiss the issue. Just the other night, one of my favorite economists pointed out that if the debt burden on the Chinese government starts to pinch, then it might have to stop buying (or even start selling) American government bonds. That could fuel a rise in interest rates, and then the debt burden of the American government would spike up. If that happens, have a nice day. But a high ratio of total liabilities of all of the entities within in a country to its GDP is at best a very imprecise indicator of financial distress.

Bank Regulation, Left and Right

One possibility I am considering for this panel discussion is to give a spiel on how free-market economists have been more hawkish than mainstream economists when it comes to bank regulation. I was inspired by listening to Robert Litan recount some of the history at a talk last night on Trillion Dollar Economists. You may recall that may take on that book is that it has great material, but I would have liked to see different organization and emphasis. I was reading it with my high school economics students in mind.

Anyway, here is the spiel.

Long ago, two groups of free-market economists decided to “shadow” the Fed. The Shadow Open Market Committee, which I believe started in the 1970s, issued pronouncements critical of monetary policy. The Shadow Financial Regulatory Committee (SFRC) , issued pronouncements critical of regulatory policy starting in 1986.

The SFRC had both a deregulation agenda and a regulation agenda. Their deregulation agenda was mainstream. Observers of banking regulation across the political spectrum recognized that deposit interest ceilings were no longer workable, that the Glass-Steagall separation of banking and securities was no longer workable, and that prohibitions against interstate banking and branch banking were no longer workable. People had known since the late 1960s that those regulatory boats were sinking, and the laws that Congress passed in the 1980s were just the long-awaited permission to abandon ship.

Take Glass-Steagall, for example. In 1968, the distinction between a loan and a security was obliterated by GNMA. A few years later, the distinction between a deposit and a security was obliterated by money market funds. Even though some people try to blame the financial crisis on the repeal of Glass-Steagall, the fact is that it collapsed 20 years before it was repealed and nobody has said that you could bring it back.

It was the SFRC’s hawkish regulatory agenda that was out of the mainstream. In particular, if you read through its old statements, you will see that the SFRC issued many warnings that bank capital regulations were inadequate. They pleaded for tighter regulation of Freddie Mac and Fannnie Mae, for higher capital requirements for banks, and for regulators to require banks to have a thick layer of subordinated debt which would put the onus for failure on the private sector rather than the taxpayers. These calls went unheeded. The SFRC economists were viewed as cranks, whose judgment on these matters was impaired by an irrational distrust of government agencies.

The Financial Supermarket Bubble and Banking History

Here is a chart, using the Google ngram tool, showing the frequency of the appearance of the term “financial supermarket” over time.

Note the spike in the mid-1980s. Given that these are books, which appear with a slight lag, I would say that the spike in the media was in the early 1980s.

At this panel, I don’t know whether I will have time to get into the history of bank concentration in the U.S., but here it is.

1. The market share of the largest banks follows a hockey stick pattern since 1950. It stayed very low until the late 1970s, and then around 1980 it started to grow exponentially. Growth of banks had been retarded by ceilings on deposit interest rates, branching restrictions, and Glass-Steagall restrictions. Banks had been trying to find loopholes and ways around these restrictions, and regulators had been trying to close the loopholes. Then, during the period 1979-1994, the regulators stopped trying to maintain the restrictions, and instead cooperated in ending them. That was when the hockey stick took off.

2. The regulators thought that this would bring more competition and consumer benefits. What the banks had in mind was something else. That is where the chart comes in. The bankers all thought that “cross-selling” and “one-stop shopping” would be killer strategies in consumer banking. In 1981, when Sears bought Dean Witter, many pundits thought that putting a brokerage firm inside a department store was going to be a total game-changer.

3. It turned out, though, that consumers did not flock to brokerage firms in department stores, or to any of the other one-stop-shopping experiments in financial services. The economies of scope just weren’t there.

4. Meanwhile, concentration in banking soared thanks to mergers and acquisitions. I’ve read that JP Morgan Chase is the product of 37 mergers and Bank of America is the product of 50. All of these took place within the past 35 years.

5. Just five years into this exponential growth process, Continental Illinois became insolvent, and that was when “too big to fail” began. So out of the 35 years where we were on the exponential part of the hockey stick, 30 of them have taken place under a “too big to fail” regime. In short, the concentration in banking got started during the “financial supermarket” bubble, and from then on was supported, if not propelled, by “too big to fail.” But the market share of the biggest banks is not something that grew naturally and organically out of superior business processes.

6. As another historical point, when the S&L crisis hit, the government set up the Resolution Trust Corporation. Each failing institution was divided into a “good bank” and a “bad bank,” with the good bank merged into another bank and the assets of the bad bank bought by the RTC. While this was a somewhat distasteful bailout, it was conducted under the rule of law. When TARP was enacted in 2008, Congress and the public were led to expect something similar to the RTC, with TARP used to buy “toxic assets” in a blind, neutral way. Instead they ended up calling the biggest banks into a room and “injecting” TARP funds into them. They also spent TARP funds on restructuring General Motors. It was the opposite of government acting in a predicable, law-governed way. It was Henry Paulson and Timothy Geithner making ad hoc, personal decisions. I think that in the U.S., that is what bank concentration leads to–arbitrary use of power. That is why as a libertarian I do not think that allowing banks to become too big to fail is desirable.

Financial Report of the U.S. Government

The report is here. It looks interesting, but I find it difficult to parse. Liqun Liu, Andrew J. Rettenmaier, and Thomas R. Saving parse it this way:

The liabilities reported in the FRUSG at this time last year included $12 trillion in debt held by the public, $6.5 trillion in federal civilian and military employees’ accrued pension benefits and other retirement and disability benefits, and $1.3 trillion in other liabilities, producing total liabilities of $19.9 trillion.

They point out that the liabilities for Social Security and Medicare seem suspiciously small, because the report acts as if these could be erased quickly with the stroke of a (legislative) pen. Technically, that is true, but realistically it is not. Instead, Liu, et al, propose to include benefits payable to current retirees.

Adding the $16 trillion in accrued Social Security and Medicare benefits payable to current retirees produces a total of $35.8 trillion in federal liabilities. These accrued Social Security and Medicare benefits are larger than the debt held by the public and are 45 percent of the total.

Pointer from James Pethokoukis.

This is still not very satisfying.

1. The liabilities to pay benefits to those of us not yet eligible ought to be included.

2. If we are going to include future government expenditures as liabilities, then we ought to include future tax revenues as assets.

3. We ought to use a discounted present value concept, rather than treat dollars that will be spent or received 10 years from now as equal to dollars that will be spent or received today.

Conceptually, I believe that what we want is a present discounted value of assets (including future tax revenues) and liabilities under current law (or what CBO projects law to be under its more-plausible “alternative scenario”). You can then look at the change in these values from year to year as an accrual-accounting measure.

Resolving Illiquid Institutions

Noam Scheiber brings up the AIG bailout, once again.

Which leaves only two possible explanations for the overly solicitous treatment of Goldman and the others. The first is that their own financial position was so precarious that accepting anything less than the billions they expected from A.I.G. would have destabilized them, too. Which is to say, it really was a backdoor bailout of the banks — many of which, like Goldman, claimed they didn’t need one. Alternatively, maybe Mr. Geithner simply felt that Goldman and the like had a more legitimate claim to billions of dollars in funds than the taxpayers who were footing the bill.

Five years ago, AIG had more liquid liabilities (“collateral calls”) than liquid assets. There were a number of ways this could have been resolved.

1. No government action, AIG’s creditors go to court, they win a quick judgment, and AIG has to sell off assets in order to pay the creditors.

2. No government action, AIG’s creditors go to court, things stay tangled up for a while, meanwhile AIG’s liquidity position improves, and creditors get paid out without AIG having to sell assets.

3. What I advocated, which was that the government tell creditors that they could get most of their money now or all their money later, but not all of their money now. I called this the “stern sheriff” solution.

4. A pure government bailout, which ensured that creditors could get all of their money now, courtesy of the taxpayers.

5. What we got, in which creditors received their money, but the government made sure that AIG shareholders suffered in the long run.

Note that (5) ended up close to (1), and (3) would have ended up close to (2). Had the government done nothing, then the courts would have effectively decided which path to head down. The advantage is that we would have gotten there by the rule of law, not by arbitrary exercise of power.

I think that the lesson we should draw is that in future cases of liquidity problems, officials should stand back and let nature take its course. I think that the number of prominent economists who agree with me on that approaches zero.

Question: Suppose that the top officials involved in dealing with the financial crisis had been forced to wear cameras and an audio recorders during all of the meetings during the crisis, with the stipulation that they could delay the release of the recordings for 90 days if they determined that immediate release would be harmful to financial stability. Do you think that this would have changed either their decisions or the public perception of those decisions?

Government Accounting

Jason Delisle and Jason Richwine write,

the government’s official method for estimating cost is incomplete. It fails to incorporate the cost of the market risk associated with expecting future loan repayments. So-called “fair-value accounting,” an accounting method favored by the vast majority of finance economists as well as the CBO itself, factors in the cost of market risk. The difference transforms the official student-loan “profit” into a loss, for a budgetary swing of $279 billion over ten years. That figure demonstrates why the stakes are so high in the debate about fair-value accounting.

I recommend the entire essay. I would like to make changes to government accounting a top economic priority, because I think that avoiding a debt crisis ought to be a top priority.

If you ignore risk, then the government can appear to make a profit with all sorts of loans and loan-guarantee programs. I would go beyond fair-value accounting and subject the government budget to stress-testing, to give a measure of risk exposure.

Social Security as a Public Bad

Robert Fenge and Beatrice Scheubel write that they provide,

an empirical confirmation of the negative relationship between statutory old-age insurance or more broadly statutory social insurance and fertility. The effect amounts to a total reduction of approximately 1.7 marital births per 1000 between 1895 and 1907… [the] impact of pension insurance is comparable to the impact of an increase in urbanisation by 10-20%.

Pointer from Brian Blackstone

Off hand, it would seem to me that any form of capital accumulation by the elderly could have this effect. Private pensions certainly, and perhaps even private savings for retirement. But it might be argued that there are huge negative externalities in public pensions, in that they allow you to benefit from my having children.

New York City Pensions

The NYT reports,

Next year alone, the city will set aside for pensions more than $8 billion, or 11 percent of the budget. That is an increase of more than 12 times from the city’s outlay in 2000, when the payments accounted for less than 2 percent of the budget.

I could see the same thing happening where I live, in Montgomery County, Maryland. At some point, citizens will be paying much of their taxes not for current services but instead to try to keep unsound pension systems afloat.

Megan McArdle writes,

The core problem is that returns have not tracked with the city’s optimistic projections. In 2012, the city finally lowered its projected return to 7 percent from 8 percent, but after decades of excessive optimism, that left it with a giant hole; the payments had to be stretched out over more than two decades in order to minimize the fiscal hit. Yet this still may not be enough; it’s possible that 7 percent is still too rosy.

And of course, as many people have pointed out, a private firm’s auditors would not sign off on this sort of pension accounting.

Europe Still in Trouble

Barry Eichengreen and Ugo Panizza write,

For the debts of European countries to be sustainable, their governments will have to run large primary budget surpluses. But there are both political and economic reasons to question whether this is possible. The evidence presented in this column is not optimistic about Europe’s crisis countries. Whereas large primary surpluses for extended periods of time did occur in the past, they were always associated with exceptional circumstances.

Pointer from Mark Thoma. Read the whole thing. Shorter version: Have a nice day.