AIG, moral hazard, and “depression economics.”

I don’t have much to say about the initial financial crisis, other than lots of bankers made lots of bad decisions.  And through experience I have found that virtually nobody finds that bland explanation satisfactory.  At the same time I have always had a nagging feeling that moral hazard played a bigger role in the financial crisis than was apparent at first glance.   A very interesting recent post by James Hamilton shows how moral hazard contributed to the crisis.  He described the insurance giant AIG as a sort of hedge fund, which made an enormous bet insuring mortgage-backed bonds, even though:

AIG lacked the financial resources to make good on those contracts in the event that the housing downturn became as severe as it has now proved to be.

But then why would large investment banks have risked buying insurance from AIG?  Because they anticipated (correctly) that the U.S. government would never allow AIG to default on all of that bond insurance.  Just another example of “heads the Wall Street speculators win, tails the taxpayer loses.”

Hamilton also has some good ideas about how to handle this mess in a way that reduces the incentive of banks to engage in this sort of practice:

To the extent that the buyer and seller of the CDS were making a bet for which the taxpayers were implicitly picking up the downside, the CDS market, rather than helping institutions effectively manage risk, would have been a factor directly aggravating systemic risk.

I raise this issue not to be yet another voice clucking that we need to get tougher on AIG or others in order to prevent moral hazard, though that is one reasonable inference to draw from the discussion above. But the issue for me has always been not to exact retribution or instill market discipline, but instead the very pragmatic question of how to use available resources to minimize collateral damage. I accept the argument that a complete failure of AIG would have unacceptable consequences. The relevant question then is, what combination of parties is going to absorb the loss?

The concern I wish to raise is that any reasonable answer to that question would include Goldman Sachs, Merrill Lynch, Societe Generale, and Calyon, to pick a few names at random, as major contributors to this particular collateral-damage-minimization relief fund. But if they are to contribute, the plan must be something other than doling out another $100 billion every few months to try to keep the operation going a little longer, but instead requires seizing this bull by the horns. Split AIG into a core business we want to protect– with enough equity to be a viable operation, and a hefty fraction of the existing management team fired– and a derivatives business that’s going to be systematically liquidated in large part by abrogation of outstanding contracts.

Then there’s the domino effect to consider. What do we do when this brings down the next player who can’t continue operations without those payments AIG (or the taxpayers) were supposedly going to deliver? I say, we implement the parallel operation there.

That’s my proposal for how to dismantle the derivatives house of cards. One trillion at a time.

A very well thought out proposal, but unfortunately I just don’t see it happening.  It seems to me that ever since Lehman failed the policymakers in Washington have been doing whatever they can to prevent more shoes from dropping.  I think their logic goes something like this:  “The failure of Lehman triggered a severe financial panic due to a loss of confidence in much of the financial system.  Although bailing out institutions that behaved recklessly does incur some moral hazard problems in the long run, the first priority is to restore confidence and arrest the sharp decline in AD.”  You probably already guessed that I don’t agree with this line of reasoning (as I think excessively tight money caused AD to fall sharply, not financial panic), but I would like to use this example to address a series of questions that I have been getting from readers.

Many participants in this blog seem to be sympathetic to one of the various anti-inflation ideologies (the gold standard, Austrian economics, monetarism, etc.)  I am often asked why I want to pursue a policy explicitly aimed at generating a little bit of inflation.  One answer is that when AD falls sharply (as it has recently) the costs to the economy go far beyond unemployment.  Instead, all sorts of policies are adopted that in normal times, when the authorities are not worried about falling AD, would never even be considered.

These bad policies are justified on the grounds that they can boost AD, or at least prevent it from falling as quickly as it otherwise would.  Some of you may know that Mr. Krugman gives another name to this phenomenon, what I call “bad economics,” he calls “depression economics.”  Only he doesn’t consider it bad, he thinks it is good.  Like Keynes, he argues that when the economy has a lot of slack, policies that would otherwise be rejected on cost/benefit grounds, suddenly look much more desirable.  The core of classical economics, opportunity cost, becomes of doubtful value when there is slack in the economy.  (BTW, I am not saying that Krugman would oppose Hamilton’s proposal, my point is that the likely reason that it would be rejected by authorities is a variant of depression economics—that allowing AIG’s contracts to be abrogated would directly or indirectly depress AD even further.)

My own view is that Krugman’s “depression economics” is wrong.  Before explaining why, it will be useful to draw a distinction between a depressed economy, and an economy that is expected to be depressed in the future.  The reason this distinction is important is that there are significant lags in the impact of many “depression economics” policies, or indeed any stabilization policy.  For instance, consider a new light rail project that will take one year to be “shovel ready.”  This lag encompasses the time Congress spends debating the proposal, and also the time required to draw up blueprints, get regulatory approval, etc.  In other words, the normal lags in fiscal policy.

Now consider when you would want to undertake a public works project that would not normally meet cost/benefit criteria.  When the economy is at full employment, but expected to decline sharply, or when the economy is in recession, but expected to recover over the next year?  Obviously the former, you’d like to have the project actually underway when the economy was in recession, and thus when the opportunity cost of using labor and capital was lower than normal.  So it’s not really “depression economics” it’s “expected depression economics” that Krugman is discussing.

Now let’s suppose that the Fed was following Lars Svensson’s maxim of always setting the stance of monetary policy in a way that equated the target growth in inflation or NGDP and the forecast growth in those aggregates.  This doesn’t mean the economy would never be in recession, but rather that the expected future level of spending would never be at recessionary levels, at least over a time frame where monetary policy could be expected to impact the level of nominal spending.  So if you want nominal spending in the first quarter of 2010 to be running at an annual rate of $15.2 trillion, you set monetary policy (interest rates or OMOs) at a level expected to produce that level of spending.  And then what auxillary role does fiscal policy play?  None.  (Unless the lags of fiscal policy are significantly shorter than monetary policy, which I consider highly unlikely.)

I started this post talking about how recessions lead to bad economic policy, that is, policy that ignores classical principles such as opportunity cost, the value of thrift, or moral hazard.  Then I argued that Paul Krugman had defended this sort of unorthodox approach to policymaking, arguing that in the world of “depression economics” we had to throw out part of the classical rulebook.  Now I am arguing that depression economics isn’t just “bad economics” but much more specifically it is “bad monetary policy economics.”  He is talking about policies that can only be justified if the central bank is not doing its job.

Even if there are unemployed resources, there is no justification for dropping the principle of opportunity cost if the central bank is targeting the forecast.  By the time any public works project is ready to implement, the economy should already be at full employment, or at the very least, closing in on full employment as quickly as the monetary authority thinks is prudent.  The future expected opportunity cost of labor is then roughly its wage rate, exactly as in the classical model.  And there is no excuse for bailing out reckless Wall Street tycoons, because we have nothing to fear if their firms collapse—the Fed will adopt a policy stance expected to offset any impact on AD.

The sort of policy I envision may seem hopelessly utopian given the recent failures of monetary policy.  Ideally the Fed would adopt something like the NGDP futures targeting approach that I advocated in an earlier post.  But even last October, when no such system was in place, if they had simply cut rates to zero, set an explicit inflation target, and engaged in aggressive enough QE to keep the indexed/nominal bond yield spread (i.e. expected inflation) in the 2-3% range, we never would have had to resort to “depression economics.”  We will be paying for this monetary policy failure for many years, and in ways that we don’t even fully understand.

In 1999 I published a paper in Economic Inquiry that argued the General Theory was a gold standard model.  The gold standard is one form of “bad monetary policy,” that is, a policy that does not keep the expected growth in NGDP stable.  I found that the gold standard assumption was the only way I could make heads or tails of the General Theory.  Try reading the GT under the assumption that you have a fiat money regime and the central bank is always targeting expected NGDP growth at 5%.  See if you can find even one insight in the entire book that makes sense under that sort of regime.



28 Responses to “AIG, moral hazard, and “depression economics.””

  1. Gravatar of Jon Jon
    14. March 2009 at 20:27

    “But then why would large investment banks have risked buying insurance from AIG? Because they anticipated (correctly) that the U.S. government would never allow AIG to default on all of that bond insurance.”

    Scott: AIG made CDS contracts on super-senior tranches of South American MBS held primarily by European commercial banks not investment houses. As I understand it, AIG has still not paid any significant amounts as a result of those contracts, but has been compelled to post 100B+ in collateral under the terms of the CDS. It has been this collateral call that has been hoovering all the cash. No actual losses have been incurred and no significant payouts have been made.

    “The word” has been that an AIG bankruptcy would have ruptured those CDS contracts and compelled a massive (and impossible) ~400B capital raising by European bank counterparties. Hence the ‘systemic risk’ and the intervention by the NYFED.

    AIG’s 10K filing to the SEC lays it out:
    “Approximately $379 billion (consisting of the corporate loans and prime residential mortgages) of the $527 billion in notional exposure of AIGFP’s super senior credit default swap portfolio as of December 31, 2007 represents derivatives written for financial institutions, principally in Europe, for the purpose of providing them with regulatory capital relief rather than risk mitigation. In exchange for a minimum guaranteed fee, the counterparties receive credit protection in respect of diversified loan portfolios they own, thus improving their regulatory capital position.”

  2. Gravatar of bill woolsey bill woolsey
    15. March 2009 at 05:11

    Jon’s report here is interesting and horrifying.

    It is fundamentally about a fraud.

    Either the European banks need more capital or they don’t. Making sure that AIG continues to meet its obligations so that we can pretend that these securities on the balance sheets of European banks are extra safe (compared to ordinary mortgages or commercial loans) is just a fraud.

    What is the purpose of capital requirements? How can there be any purpose to them if they are being met by window-dressing?

    It seems to me that the best hope for Wall Street to return to a model in which banks hold securitized loans in preference to ordinary loans (in other words, allows Wall Street to collect fees for these services) is to prop up the house of cards as best as possible.

    Anyway, I have held the view that the Fed should maintain nominal income on its past growth path using quantitative easing. It seemed to me that the financial crisis would distrupt credit markets, but that as long as nominal income was kept on target, the result would be stagflation. Slower growth in output and higher inflation than normal. Perhaps output would actually shrink.

    It is interesting to imagine how this would “work.” Demand remains strong. Firms can sell. But some firms cannot get credit to bring goods to market. Or to finance operations. Shortages appear. People buy “second best” products. Firms with strong cash flow don’t have problems. Firms that are good credit risks, don’t have problems. They produce goods and that is what people buy. (Credit markets are not playing the real role of moving funds to finance the most productive uses of resources. Less valuable goods are produced because of failures in finance.) Banks gaining deposits and willing to park loans on their balance sheets expand. Banks that make loans and try to sell them, can’t. Wall Street’s loan securitiziation business is gone. People who were doing the securitization are out of a job.

    Compounding this problem is that there was a bubble in housing, and resources need to move from housing to elsewhere. This will result in lower production while the resources shift.

    To the degree that problems in the credit markets result in overshooting (too few housing purchases), then this is exacerbated.

    One can certainly understand how it would seem desirable to avoid this and “fix” the problems in credit markets.

    And, of couse, Wall Street is politically powerful, and fixing up their problems, covering their pass losses, and making sure that they can continue with business as usual ASAP is in their interests. And so, in the politicians interests.

    Further, it seems sensible that if they fix up all the problems in the credit market, spending should rise again.

    So, not only do we not get losses in output because of inefficiencies in credit markets, we keep nominal income on target. Isn’t this so much better?

    But, it isn’t working. So, what not quantitative easing? Why aren’t short term interest rates zero (or less?)

  3. Gravatar of ssumner ssumner
    15. March 2009 at 06:08

    Jon, I know little about high finance, and so I won’t have much to say here. But a couple of small points:

    1. Hamilton suggests that investment banks like Goldman Sachs were counterparties that stand to gain from the bailout. Is that so?

    2. If we let the bond insurace part of AIG fail, and European banks were about to fail as a result, can we assume that European taxpayers, rather than American taxpayers, would pick up the tab for any bailout?

    Bill, Again, I don’t know enough finance to intelligently comment on this. I’ll just reiterate two points that may have gotten lost in my rampling post:

    There are two big advantages to maintaing a sound monetary policy expected to produce roughly 5% nominal growth:

    1. Any financial crisis is much less likely to have a strongly negative impact on the real economy.

    2. The financial crisis itself would have been far smaller if NGDP was growing at 5%, than if it shrinks at 7% (as it recently did.)

    It is always important to keep both benefits in mind. This doesn’t directly address your suggestions about financial reform, but I am not knowledgeable enough about banking to have a useful opinion. Anything that reduces moral hazard, however, sounds good to me.

  4. Gravatar of P.E. Bird P.E. Bird
    15. March 2009 at 07:16

    I have a different criticism of Keynesian approaches. To the extent they worked, it was because no one expected them to – they were so counter-intuitive at the time that their intention was allowed to take hold prior to public (and private) realization of what was occurring.

    The best these approaches can hope for is some sort of catalytic effect. Confidence is a very fragile thing and extremely difficult to engineer. It cannot be manipulated into existence.

    The problem now is that everyone “knows” how Keynesian approaches “work” – it can be pre-gamed by participants with enough firepower – so any catalytic effect is lost. Instead of confidence being restored it is further destroyed.

    I also think that moral hazard has and CONTINUES to have a huge effect. But we should call it what it is – immoral refuge – and not simply a lack of moral hazard.

    We would try one set of cures if we believe we didn’t have “enough” moral hazard. But if instead we have immoral refuge, we have different problems indeed.

  5. Gravatar of Jon Jon
    15. March 2009 at 09:05

    While I get your visceral reaction, and while I think history has now shown us that the Basel II accord is idiotic, this stuff was considered genus only a couple years ago.

    The basic element is under the Basel II frame work, capital requirements are based on an underlying estimation of the risk of the assets. Banks are required to have very little capital against Treasuries and more as you go down the rating scale. One nasty element though is that the rules are based on historic loses–which makes the rules very pro-cyclic. Just as loses start appearing and you start draining into capital, the Capital-Adequacy-Ratio increases.

    One way to combat that is to buy an insurance contract; i.e., a CDS with AIG. Now you get to both treat the underlying asset as being higher grade and declines in the quality of the underlying asset are decoupled.

    You don’t hear about this too much in the US because the US is very late in adopting the European model–2007 versus 2004 in Europe.

    Scott: GS is an AIG counter-party. I believe they were one of the first firms to recognize the risk at AIG and demand collateral. As to whether the Europeans would have been on the hook. The answer is “yes”. The banks would have been insolvent and conceivably seized by the appropriate national regulator, but its easy to see why that would have been devastating.

    Background: (I encourage you to at least read the introduction)

  6. Gravatar of JimP JimP
    15. March 2009 at 10:57

    On expectations –

    Bernanke is going to be on 60 minutes tonite – and I am awaiting it with horror. My expectations could not be lower.

    He will be a complete contrast to Roosevelt – with his confidence, big smile and cigarette holder. You know what Roosevelt said about that holder – “my doctors told me to stay just as far away from cigarettes as I can”. Roosevelt was an inflationist – a happy and optimistic man.

    Bernanke is sad and glum. He will no doubt tell us that he expects the economy to be weak for quite a while longer.

    He will cause that weakness by telling us that the central banker expects that – the one man and institution that could actually do something about it is going to continue to take a pass. The thought of it makes me gag.

    It has been a continuing and onging horror show – a complete collapse of monetary policy. Krugman is happy – because soon Obama will be able to nationalize everything.

    The thing that got to me about that Noonan piece is that she has noticed that people are hording both gold and cash. Monetary expectations are completely unanchored. Will this have an inflationary or a deflationary outcome? No one has the remotest idea.

  7. Gravatar of JimP JimP
    15. March 2009 at 11:15

    I am bitter – of course. And I sound bitter. We are going to get miles of Krugman – whom I detest.

    Here is what Scott said on Nick Rowe’s blog:

    “My prediction is that no matter what the Fed buys, if they have an expansionary enough and credible enough policy stance to send the stock market soaring, long yields will almost certainly rise. And I think a truly dramatic and credible initiative could cause an extraordinary increase in equity prices, despite the liquidity trap-types who say money no longer matters.”

    I sure do believe that to be true. And I would rather have higher stock prices than having to listen to Paul Krugman tell us how smart he is to the end of my days.

  8. Gravatar of ssumner ssumner
    15. March 2009 at 13:47

    P.E. Bird,

    When Keynesian economics was first tried you could argue that it had a bigger effect because a given increase in AD, or nominal spending, will tend to be more expansionary in real terms if it is unexpected. The 1960s might be an example of this (and then we had the hangover in the 1970s.)

    I agree that moral hazard is very important, whatever it is called.

    Jon, Thanks for all the info. It was a struggle for me to even read the intro you attached. My brain is not wired correctly to pour through regulatory writing. Let me throw out a couple observations, which are really questions to you, and get your reaction.

    1. The intro showed a probability distribution with a point estimate of losses that would occur only one time in 1000. Am I right that the actual losses are probably larger than that estimate, and that one problem was the regulators (or banks) simply incorrectly estimated how risky certain assets were. In other words, it’s not just that we had a one in 1000 event, but in retrospect this was always much more likely than we imagined?

    2. The intro talks about the problem of international competition. The impression I get is that although you could force American banks to be super-conservative, there is a fear that they would lose a lot of business to London. Thus even “liberal” senators in states like NY (normally pro-regulation) would be reluctant to support draconian regulation. Does this mean (if we can’t solve the moral hazard problem) that we need an international agreement on regulation?

    JimP, Thanks for the comments. For what it’s worth, I’m not sure if I’d invest in gold right now. Gold might do well in the extremes (1930s-style depression, or 1920s-style hyperinflation.) But we are more likely to muddle along with 1% inflation and slow growth. But I think your point wasn’t that gold is necessarily a good investment, but that the fact that people are hoarding gold and cash is a sad comment on our monetary policy—and I agree with that.

  9. Gravatar of Scott Sumner’s public-choice inflationism « Entitled to an Opinion Scott Sumner’s public-choice inflationism « Entitled to an Opinion
    15. March 2009 at 14:22

    […] In keeping with Bryan Caplan on Hillary’s gas-tax holiday and Tyler Cowen on the stimulus, Scott Sumner supports inflation contra the Austrians/gold-bugs if only to distract the public from pushing for Paul Krugman’s […]

  10. Gravatar of Winton Bates Winton Bates
    15. March 2009 at 14:40

    Scott: I wonder if it is a coincidence that your expection of 1% inflation is in line with market expections (TIPS market etc). Should the Fed be participating actively in that market to try to raise inflation expectations?

  11. Gravatar of JimP JimP
    15. March 2009 at 14:44

    The Bernanke interview –

    A promise to “take the money out” after the emergency is over. A deflationist. Charming. An explicit rejection of price level targeting. At a time when he knows rates are six percent too high. The man is insane.

  12. Gravatar of bill woolsey bill woolsey
    15. March 2009 at 15:30

    I think promising to “take the money out” after the emergency is over, is exactly correct. The money supply needs to rise to offset the drop in velocity. When velocity rises again, the money supply shoud drop. An equivalent statement is that the quantity of money should rise to meet the increase in the demand for money. When the demand for money falls, the quantity of money should fall again.

    Currently, the CPI is less than 1% below its long term trend. Most of the deflation in the CPI since July 2008 was a return to long term trend.

  13. Gravatar of JimP JimP
    15. March 2009 at 15:47

    Real rates are too high. He needs to address that. Inflation needs to be unexpectedly higher than it is. He could do so with an explicit price level target. I agree that people are frightened by the idea of future inflation – too frightened I think. Rates are six percent too high and he did not address that at all.

  14. Gravatar of Jon Jon
    15. March 2009 at 19:33

    Scott: “In other words, it’s not just that we had a one in 1000 event, but in retrospect this was always much more likely than we imagined?”

    Yes, although this could have been a ‘black-swan’, there is a consensus now that model-data that was was not representative. Still, the picture is fairly complex. I happened to be on-board with the position that FASB157, mark-to-market has exacerbated the situation. As evidence: the delinquency rate is actually a good deal lower now than it was in the 1991 recession but charge-offs are about as high.


    “The intro talks about the problem of international competition. The impression I get is that although you could force American banks to be super-conservative, there is a fear that they would lose a lot of business to London.”

    Basel II is an international accord; the rules are mostly uniform now in the world, but the US has been very slow in getting on board. You are right about the international competition aspect: there is the potential for a race to the bottom because lower capital requirements essentially mean lower costs. Presumably this applies, as you suggest, to any form of banking regulation.

    Sorry to bring up the old saw again, but the ECB used to defend paying interests on reserves as compensation for the higher reserve requirements European banks face–thus forestalling a need to race-to-the-bottom as has happened in almost every other country on the reserve-ratio.

  15. Gravatar of Jon Jon
    15. March 2009 at 19:43

    I guess the news is out widely now: “The list released Sunday of “counterparties” that benefited from the bailout is topped by European banks Societe Generale and Deutsche Bank”

  16. Gravatar of Já que há uma crise, o que o governo deveria fazer? « De Gustibus Non Est Disputandum Já que há uma crise, o que o governo deveria fazer? « De Gustibus Non Est Disputandum
    16. March 2009 at 00:25

    […] os textos de (sorry pals, apenas alguns links) Larry White, DeLong, Krugman, Mankiw, Chinn e Scott Sumner. Um breve passeio por este blog é um exercício de paciência – útil para os apressados dias de […]

  17. Gravatar of ssumner ssumner
    16. March 2009 at 04:22

    Winton, No coincidence, in an earlier post I said my motto is “I don’t make predictions, I infer market predictions.” I was thinking of replacing the word “I” with “Good economists,” But then I thought that SOMEONE has to make the predictions that I infer.

    JimP, I agree with Bill that much of the money will eventually have to be removed. At the same time you are also correct in your intuition. Krugman pointed out that a monetary injection that is perceived as being temporary will have almost no effect. Thus Bernanke should say he will remove only a part of the money, and keep enough in circulation so that nominal GDP 3 years out will be at least 15% higher than today. (Or set a price level path, as you propose.) Some people advocate faster than normal growth for the next 12 months in a sort of “catchup.” This can be justified because unemployment is 8% headed for 10%. But I have chosen not to take that path for reasons of clarity. I want to make it clear that I am not calling for a risky untried new policy, it is those that support the current Fed stance who are calling for a risky untried new policy—a policy of abandoning the normal practice of targeting the forecast, of setting policy at a level expected to achieve on target growth in prices and NGDP. By relentlessly sticking to my 5% EXPECTED nominal income targeting approach, which I first developed in the 1980s, it is easier for others to see how I could have concluded that policy went sharply off course last September/October, when most economists thought the Fed had adopted an “expansionary policy stance.” So I think you and I are on the same page, although you are a bit more aggressive in your conclusions.

    Bill, I agree with you that inflation is only a bit below trend due to the high inflation of 2007-08, but I would just add that it is another indication of why NGDP is a better indicator than the CPI (as we both argue.) I also notice that some people look at the CPI question as a separate issue for the whole stimulus debate. I’m not saying anyone on this blog thinks this way, but have you noticed that some people sort of automatically assume that there is an AD problem, and the question is whether Obama’s stimulus will work, at the same time in press articles about inflation they speculate that inflation may even be a bit below the Fed’s comfort zone, as if inflation and AD are separate issues! I want to shake those reporters and point out that if inflation were not too low, that Obama’s stimulus plan could only “work” by “not working” Because if inflation is not too low, then we don’t want to boost AD, so we should hope Obama’s stimulus plan fails. Has anyone else noticed the tendency for press people to treat inflation and AD as separate, unrelated issues?

    Jon, I wonder whether some of the difference between now and the early 1990s, and indeed the early 1980s (when balance sheets also looked bad due to high rates on short term deposits, compared to the rates on 30 year mortgages), is due to pessimism about monetary policy. There were no liquidity trap fears then, so there was always the feeling that the Fed could cut rates sharply if necessary. Indeed when rates fell sharply in 1982-83, I seem to recall that bank balance sheets improved quickly. I think we could do the same today, obviously not with lower rates, but with higher NGDP growth.

    Good point on the ECB. Just to clarify for other readers, I am not necessarily opposed to interest on reserves. I have talked up Hall’s proposal for a policy of tying the interest rate on reserves to price level deviations (in a way that would make money demand move in the direction needed to stabilize the price level. What is so ironic is that, as JimP points out, interest rate would now actually be negative on Hall’s plan! So yes, let’s pay interest on reserves, but at a negative rate. In normal times, a positive rate might be appropriate.

    I missed the Bernanke interview (will try to catch it on you-tube) but I heard that he said he slammed the phone down several times when working on the AIG case. Any future memoirs written about this episode should be very interesting.

  18. Gravatar of JimP JimP
    16. March 2009 at 07:07

    Aggressive – that would be right. It really amazes me that this is happening. We are living through history here – right in front of our nose the biggest monetary error in many years is being made. I find that stunning.

  19. Gravatar of JimP JimP
    16. March 2009 at 07:15

    The FT yacks on about all the fundamental changes we will have to make in the world economy and apparently in the whole universe. What are the changes we will be able to do that? Zero. As you said in that early post on the Puritan attitude toward monetary stimulus – if we have to do all that stuff to get out of this we are really cooked. Lets set ourselves a task we can actually DO. And then do it.

  20. Gravatar of StatsGuy StatsGuy
    16. March 2009 at 10:07

    Here is an excerpt from a comment I’d made on baseline scenario…

    “Governments (particularly when working within a multilateral framework) cannot possibly move as fast as international capital markets. Particularly with incentives to defect (this is a Prisoner’s Dilemma game)

    So far, all of the major US international efforts – fund the IMF, pass 2-3% GDP stimulus programs, keep trade barriers low, etc. – all of them have incentives to defect.

    The only one that does not is monetary policy.

    The world’s major currencies are all killing debtors by preserving high currency valuations when the world is de-leveraging.

    The reason: Terror that when deleveraging stops, inflation will kick in. All of this comes about because of dependence by governments on private institutions (banks, hedge funds, etc.) to maintain the size of the world money supply.

    Perhaps the solution to deleveraging is not to encourage re-leveraging. Perhaps it’s to replace privately created money with publicly created money (so the Fed can have more direct control over the money supply, and is less reliant on Velocity, which can radically change when risk perceptions change).

    So, pump in money, and to prevent a “rebound effect” when confidence is restored, reduce capital-asset ratios proportionately to the amount of money pumped in. This will force long-term deleveraging, remove systemic risk, and avoid a massive short-term credit crunch.

    This should actually prevent long-term inflation (which will really destroy cash based savers) and prevent government defaults (particularly if led by world major currencies).

    Who suffers? Honestly, banks and certain hedge funds, by losing the ability to leverage as highly (and therefore make one-sided bets).

    If the G20 was _really_ interested in the prosperity of the world, they would do all of this at the same time:

    1) Open up the Basle Accords, and create a fixed schedule (phasing in over ~3 year period) to reduce capital/asset ratios to levels which are less prone to causing systemic risk. This would prevent inflation by permanently taking down worldwide leverage levels.

    2) Set monetary inflation targets, to be achieved through global quantitative easing to allow member countries to both reduce debt _and_ temporarily hold to existing expenditures. This would pick up the slack from deleveraging.

    3) Regulate large pools of international money – whether they be money havens or hedge funds – to prevent excessive leverage. Remove all of the loopholes.

    Alas, this shall not happen – the Wicked Witch of Western Europe wants her money back. Now.”

    Gold-enthusiasts lay the roots of the problem at the feet of Fiat currency. Quite the contrary – the problem is not that govts. are printing too much money, it’s that they are printing too _little_ money and letting banks print most of it. This abdication of the power to control currency means less control over the overall money supply as bank default rates soar (due to small losses being exacerbated by leverage) and money slows down (due to risk perception).

    Both of these problems would be substantially ameliorated by increasing cap-asset requirements. Fewer bankruptcies (due to better capitalized banks) and changes in money velocity would have less impact on overall money supply.

    This would imply a permanent reduction in total world money supply – but this can be easily countered by simply increasing the base money supply (which, btw, allows countries to pay down debt while maintaining consumption).

    Who loses? Banks. And, I suppose, fraudulent peddlers of Ponzi schemes.

    Will such a proposal see the light of day? Nope. The die has been cast – we are committed to restoring the current system.

    I suppose I could swallow this (with difficulty), except that in order to prop up the current system we’re incurring huge taxpayer debt precisely when unemployment is rising, entitlements are increasing, boomers are retiring, etc.

  21. Gravatar of JimP JimP
    16. March 2009 at 11:09

    StatsGuy –

    All true.

    So: – Do what we can do. Set a price level target – and put Volcker in charge of hitting it. Toss Geither under the bus and give Treasury to Mr. Inflation Fighter. People would believe him when he says that the aim is moderate inflation – but committed inflation – not deflation – as well.

    And with the direct support of Obama. This will give him the decisive policy he needs. The leadership he has so far not shown. He is fading faster than a cut flower.

  22. Gravatar of Carl Futia Carl Futia
    16. March 2009 at 14:29

    Here in an interpretation of events along the lines of Minsky-Schumpeter.

    The Panic of 2008 was a consequence of a three coincident, revolutionary, technological innovations and the credit inflation that typically follows such innovations.

    What were they? First we have securitization of loans, exemplified initially by mortage backed securitiesin the late seventies and then later by various asset backed securities, CDO’s etc.

    But securitization would not have been able to do the damage we have seen without two other enabling innovations. These were the exponential growth of computing power (desktop computing from the early eighties) and the pathbreaking insight of option pricing theory which began with the Black-Scholes theory (1973).Computing power enabled the packaging and tracking of vast numbers of loans into complex securities. Pricing theory told bankers how such securities could be rated and valued (something that needed computing power too!)

    These innovations opened up legitimate new and previously unexploited profit opportunites in the banking business. Like all such opportunities, especially those that create new ways of doing business, they seduced people into believing that the stream of profit they created would never run dry.

    But there was a special demon at work here that is easy to overlook. The mathematization of the banking business enthroned “numbers” and “data” as the ultimate arbiters of expectations about the future. Technology empowered the frequentist interpretation of probability. Bayesian subjective priors (which might for example have attached positive probability of the return of housing prices to the level of construction costs) could not survive economic competition in the banking business.

    Bayesian beliefs were starved of profit, for they would NOT have permitted AAA ratings of what eventually proved to be “toxic waste”. Only the frequentist derived AAA rating (based only on historical housing price data) could sell asset based securities.

    When the bubble in housing prices stopped inflating, and default experience started to falsify expectations built from the last 25 years of data, the entire edifice, which was initially constructed to exploit real profit opportunities, collapsed.

  23. Gravatar of ssumner ssumner
    16. March 2009 at 15:43

    Carl, I’ll defer to your expertise with regard to the financial crisis, but I’m not sure whether Schumpeter and Minsky have a theory of the monetary problem. Here is my question: Do those individuals have a theory of nominal income determination? If so, what is that theory? Does it explain why our GDP is roughly $15 trillion, rather than $15 billion or $15 quadrillion. I will admit to not having looked closely at their monetary theory, but my impression is that they really don’t have one. And without a monetary theory, any explanation of the financial crisis is fatally flawed–because even with all those factors you mentioned, we wouldn’t be in this predicament if the Fed had provided enough money to keep NGDP growing at a stable rate.

    Statsguy, I am not opposed to regulation of the financial system where there is a moral hazard problem, but I have very little confidence that our future attempts at regulation will be any more successful than our past attempts, unless we fix monetary policy.

    You are right on the mark with your observation that all the major currencies are too strong right now.

    JimP, I am equally stunned that people don’t see what is going on here. I just watched the 60 minutes interview, and heard Bernanke say the problem got much worse than expected, requiring more and more bailouts, because the economy got dramatically worse. That’s what I was saying in October, and even now I don’t think Bernanke understands that the economy got worse because monetary policy allowed NGDP to fall precipitously. Having said that, Bernanke seems like a very smart, likable guy, so I am more perplexed than angry. Maybe he sees something I don’t, be he hasn’t communicated what he sees.

  24. Gravatar of StatsGuy StatsGuy
    16. March 2009 at 17:35

    @ Carl

    Holy holier-than-thou Batman!

    You are trying to blame the frequentists for the financial crisis? I am stupefied!

    It’s not as if both the Bayesians and the frequentists didn’t have FOUR HUNDRED YEARS of data to look at… Oh no, don’t chalk this up to statistical philosophy.

    It’s bad assumptions, bad (corrupt) incentives, bad system design, bad human brains (from evolutionary pressures that did not optimize our brains for analyzing CDS contracts), and a small dose of bad luck.

    But blaming the frequentists? For all intents and purposes, Bayesian and frequentist models say the same thing 99% of the time. If you have a Bayesian model that assumes independent stochastic processes when everything is massively correlated, you get the same wrong estimate of risk. How is a frequentist who makes bad assumptions different from a Bayesian who assumes bad priors?

    @ Scott

    Reduce my proposal to this: Let credit shrink. Keep it shrunk (permanently). Replace it with more base money. Have government spend this money (preferably to pay down debt).

    It is a perfectly rational way to maintain money supply. _And_ it has the advantage that it won’t risk an inflationary rebound, AND it reduces future systemic risk (by increasing banks’ capital buffer, permanently).

    So why is it not being discussed? Answer: distributional consequences. This means it’s a question of power, not free exchange. That’s the “political” part of “political economy”.

  25. Gravatar of Carl Futia Carl Futia
    17. March 2009 at 04:49


    Calm yourself! Appearances to the contrary, I have no axe to grind in the Bayesian-Frequentist Methodenstreit. If you prefer to speak of assumptions, not priors, fine with me!

    My essential point in the paragraph that so offended you is that a kind of Gresham’s law of statistical modeling was at work. The “bad” models which, through their choice of data assumed that the 20 year housing boom would continue, drove out the “good” models which took a broader historical view. Why? The bad models allowed the ratings agencies to rate the CDO’s top tranches as AAA, a rating unobtainable if you put any significant probability on the event of a housing bust. And you could only sell a CDO to a pension fund or an overseas bank and earn a big commission if the CDO was AAA rated.

    However, taking a further step back, I do think there is a big difference between our two perspectives, but that this difference has nothing to do with the foundations of probability theory. You seem to have far more confidence than I do that events like the Panic of 2008 can be foreseen and avoided. All that is needed is well-intentioned, smart people to design the correct mechanisms and set up the right incentives.

    I have a more pessimistic assessment of this possibility. The history of financial innovation shows that governments are always one big step behind events, and that markets always find new, unexpected ways to “fail”.

    P.S. Please send me your 400 years of sub-prime default data.

  26. Gravatar of StatsGuy StatsGuy
    17. March 2009 at 12:15

    I’m referring to money supply and housing data. In the US it goes back at least 120 years…

    In Europe, I’ve seen data going back 4 centuries.

    I think you’re concern with models is not so much that modellers were too dull, but that they had every incentive to overlook problems with their models and underrate risk. Every incentive. Including preserving their jobs.

    As to markets finding new innovative ways to fail… perhaps, but in this case many of the ways they failed were tried and true old ways – reopened for abuse after striking down regulations from the 1930s. I suspect financial innovators would have had a harder time getting around 20% down payments, particularly as housing prices rose higher and incomes stayed flat.

  27. Gravatar of Saad Saad
    17. March 2009 at 12:52

    I’m new to this blog, and sorry if this question is silly:

    What does AD stand for ?


  28. Gravatar of ssumner ssumner
    17. March 2009 at 14:43

    Statsguy, Although I think it is a mistake to focus on credit markets as a cause of this crisis, I do have views in other areas that you might find appealing (if you think there is too much credit in the U.S. I am attracted to Singapore’s system of forced saving and ultra-low taxes as a substitute for the Western-style welfare state. One side effect of such a system might be to reduce the amount of debt defaults, as most borrowers would have far more assets than do American borrowers. I should probably add (since I frequently mention Singapore) that I don’t endorse everything in their system. But I do think replacing high taxes with forced saving has a lot going for it. I think they are forced to save about 33% of their income, and actually save about 40%–of course such a system would have to be phased in very gradually in the U.S. But it’s not quite so hard as it seems, because much of what now goes into Social Security, unemployment comp, and health insurance benefits in the U.S., would go into the forced saving accounts.

    Saad, No need to apologize, I shouldn’t use so may abbreviations. It is aggregate demand, which can be defined many ways, but I prefer to think of it as nominal spending.

Leave a Reply