Puzzling Germany
The Economist has a recent article about the German “banking disaster”:
IT PUZZLES many in Germany that the country’s punctilious parsimony and restrained housing market have not saved it from a banking crisis that seems every bit as bad as those suffered by spendthrifts abroad.
I guess this is a puzzle for those who still think the American banking crisis caused the sharp fall in AD last fall. For those of us that have always seen the causality going the other way—from falling AD to a weakening banking sector—it comes as no surprise at all.
I’m guessing some of you will tell me that:
1. I know nothing about German banking. True.
2. There are all sorts of specific problems that explain the condition of German banks. True.
3. That some of the problems are a spillover from bad American mortgage bonds. True.
But how likely is it that lots of banking systems that seemed to come through the original sub-prime crisis in decent shape, simultaneously hit a major crisis due to a variety of special local factors? Or could banking crises in markets with no housing bubble have been due to falling worldwide AD? And if so, isn’t it likely that falling AD also dramatically worsened those banking systems that were already reeling from the sub-prime fiasco?
Tags:
4. May 2009 at 15:03
so the phrase “falling worldwide AD” means what, now? 🙂 Sorry, Google is no help here.
4. May 2009 at 15:26
4. Butterflies and dominoes: http://www.youtube.com/watch?v=T5fe6nZx6cw
I still think it’s hard to even DEFINE what a cause can be in a self reinforcing leveraged process. Yes, the scale of the crisis was NOT foreseeable with the tiny american real estate blowup, but would anything have happened if we didn’t have a real estate bubble first? Don’t think so.
4. May 2009 at 16:28
I guess this is a puzzle for those who still think the American banking crisis caused the sharp fall in AD last fall.
Huh? This is entirely consistent with the Austrian models ( or if you prefer a neo-classical economist, see the Diamond-Dybvig model of bank runs) In a maturity mismatched banking system, sunspots can cause the bank run. Germany has a maturity mismatched banking system just like the U.S., so its collapse is not surprising.
4. May 2009 at 16:29
Alex, You are absolutely right. The sub-prime fiasco was probably a necessary condition. But my focus is policy–and it wasn’t a sufficient condition with good policy. I apologize for taking so long to do the causality post, but until I get through exams all I have time for is posts where I don’t have to do much thinking, like the stuff I have been recently churning out.
There are lots of recent events where a slight tweaking of initial conditions would have changed things radically (think 9/11.) But I believe that causation goes much deeper than necessary conditions.
4. May 2009 at 16:30
Devin, Even if I bought your model of the German system, why did it fail in late 2008?
4. May 2009 at 17:11
Scott,
I am not an expert on the German banking system, and I do not know the details of the crisis over there. In a previous comment I laid out a detailed timeline of the U.S. crisis, trying to show cause and effect. A maturity transformation scheme is naturally unstable. Without government backing it can collapse at any time. The system was starting to fall in late 2007 and late 2008. But until late 2008, only a few equity holders in a few financial companies had been wiped out (Bear Stearns shareholders, Fannie Mae shareholders, etc). Creditors had mostly been bailed out. But when Lehman was allowed to go bust in September, and creditors got wiped out, the bank run began in earnest. I suspect, given the interconnected nature of the banking system, that the bank run on the shadow banking sector in the United States in starting in September of 2008 prompted similar runs in Germany.
From the article: “DEPFA’s big error was to try to boost margins by raising a large chunk of the money it loaned out in shorter-term money markets.”
Looks like it’s our old friend maturity transformation. Any closer to being convinced that maturity transformation played a role in the crisis?
4. May 2009 at 18:46
From what I’ve read in the FT, German Banks funded speculative mortgage investment throughout the world. They lent in South America and in Eastern Europe, and they did these things to escape parsimonious German regulations that curtailed the domestic market.
But your sense of timing is right. The German banks showed their distress after the global economy imploded not because house prices collapsed but because foreign borrowers were suddenly caught-out by the collapse their economy and sudden shifts in exchange-rates.
4. May 2009 at 20:06
re: Devin Finbarr
This brings to mind a quote of Krugman’s – that the run on the shadow banking system is the core of what happened:
“As the shadow banking system expanded to rival or even surpass conventional banking in importance, politicians and government officials should have realized that they were re-creating the kind of financial vulnerability that made the Great Depression possible–and they should have responded by extending regulations and the financial safety net to cover these new institutions. Influential figures should have proclaimed a simple rule: anything that does what a bank does, anything that has to be rescued in crises the way banks are, should be regulated like a bank.” He referred to this lack of controls as “malign neglect.”
4. May 2009 at 20:06
(sorry for the double post)
5. May 2009 at 02:21
Finbarr:
Why does the inability of a firm that plans to refinance but cannot cause other firms that had planned to refinance to also be unable to refinance? If all firms that had planned to refinance cannot, why does this adversely impact the flow of income, expenditure, and output?
I think you will find that implicit in your thinking is that money that is not lent is held. And that the quantity of money is inelastic. There is an assumed monetary disequilibrium that is not explored.
If, instead, we look at this in terms of the allocation of resources, then allowing for maturities mismatch provides for for more saving and, further, more rather than less round about projects to be financed. People are always free to hold long term securities. Freeing people to hold short term securities while firms fund longer term projects allows increased production–increased real return on investment. If, on the other hand, people are less willing to do this, then these productive options are not avaialable.
There is a shift in the allocation of resources between the production of consumer goods and capital goods, or among different types of capital goods.
The reason a firm cannot refinance is that its prospects are less good. When it comes time to refinance, it is like financing the first time but with a shorter term project. What is the problem? The project it is financing is less likely to pay off. Why is there contagion? Just because one firm’s prospects are less, and so it cannot refinance, why does that make other firms prospects less good so that they cannot refinance?
If you assume that monetary disquilibrium is being created (money not lent is always held and the quantity of money is fixed,) then contagion is natural.
I suppose that having a policy of never lending to any firm that plans to refinance might be a personal strategy that you want to follow. Trying to somehow prohibit other people from benefiting from planned refinancing would be an value destroying intervention in the market.
The basic mindset that sees credit building up from base money, using the money multiplier relationship between base money and deposits, is wrongheaded. Income is generated by output and expenditure is funded by income. Lending pulls money out of the stream of output and lending puts it back in. As time passes, an outstanding stock of debt can be built up. If it is repaid, the repayment takes money out of the stream of income, the receipient of the debt puts it back in.
Bankruptcy means that the bankrupt debtor takes less out to pay off, and the creditor who receives partial payment, adds less to the stream of income.
For credit markets to clear, interest rates may need to adjust.
And imbalance between the supply and demand for money can interfere with this process. But if the quantity of money accomodates the demand to hold money, there can be no monetary disequilibrium and so no problems with aggregate expenditures.
If changes in the willingness of people to finance any investment or else only invest shorter projects (or less risky
projects for other reasons) occur, this will impact the real economy. So what? Reallocations of resources to better reflect what people want happen all the time. It is a good thing.
As Scott has explained, the fundamental problem is that surpluses or shortages of the medium of account require changes in the price level to clear and that requires changes in the prices of goods and services, including the ones that make up output, and the resource prices that make up income.
I think processes where the demand for financial assets shift to shorter terms to maturity and that impacts back to bank liaiabilities used as money and on to base money are entirely plausible. But this is disruptive to the flow of income, output, and expenditure, only if the quantity of base money fails to rise enough to meet the demand.
5. May 2009 at 03:51
Bill, did you read my earlier response to you previous comments? In it I try to clarify why maturity mismatching is different than other kinds of risk taking. I’ve you read it and still have questions, I can respond further.
Why does the inability of a firm that plans to refinance but cannot cause other firms that had planned to refinance to also be unable to refinance?
Let’s say that lots of people lend money to various banks for one year. The banks then lend out the money for five years. The bank can only repay my loan if it finds a new borrower for one year. Let’s say at one point the bank fails to find a new borrower. It cannot pay back my loan and it defaults. Now, other one-year lenders observe the default, and become less willing to make that one year loan. But this constant influx of new one year lenders is needed to pay off the former one-year lenders. The bank defaults in paying off more maturing one year loans. This makes new one year lenders even less likely to lend. Now all the banks start collapsing as no one can find new one year lenders to pay off all the maturing one year loans.
The collapse of a maturity transformation scheme has the same contagion as a Ponzi scheme. It pays old investors from finding new ones. Once the scheme starts to break, no new investors want in, and then it collapses very fast.
But if the quantity of money accomodates the demand to hold money, there can be no monetary disequilibrium and so no problems with aggregate expenditures.
If a Ponzi scheme starts to collapse, and the government prints out money to bailout the investors who want out, but cannot get out, then the collapse will not be contagious. That doesn’t mean its good policy to continually bail out Ponzi schemes. For one, it will just mean more people will invest in them, and the bail out will just have to be bigger next time. This is basically what happened with our financial system over the past thirty years.
5. May 2009 at 04:17
Record number of responses in the last 24 hours, so I’ll be a bit briefer today (with much grading to do.)
Tanksley, It means NGDP was falling in most developed economies, and slowing sharply in China. I believe monetary policy determines aggregate demand and thus NGDP growth.
Devin, When you say “without government backing” I think in terms of macro policy. Without a stable NGDP path, our banking system is fragile (I agree it has some flaws due to moral hazard.) But we should try to get stable NGDP growth anyway, for other reasons. (So I am not proposing a monetary bailout of banking, by the way.)
Bill also has some good points in response.
Regarding the bank run, my understanding is that the key problem was solvency, not liquidity. Thus fallind AD led to loan defaults, which led to worsening bank balance sheets. Once this happened banks did begin hoarding, and I agree that was an additional factor reducing AD. But the Fed could have offset it.
Jon, Good point about Eastern Europe. And also about the timing issue being the key. I think we both agree there are always specific local issues, but when it’s worldwide, you look for deeper reasons. BTW, I could also have mentioned the Great Depression. Individual banks may have screwed up, but why so many failures after 1930?
Anonymousaurus, Krugman doesn’t seem to understand that it was regulation that caused the 1930s banking crisis. In the early 1930s The US had strict branching regulations, and 1000s of undiversified banks failed. Canada did not, and no banks failed (one merged under duress.)
He seems to think banks acted recklessly back then. Actually they were quite conservative, compared to modern banks, and probably even compared to Krugman’s ideal regulatory scheme.
BTW, this blog should discuss Canada at some point. Perhaps Nick can explain why their banks did relatively well. I am glad to hear that Obama is looking at Canada as a model. I am a pragmatist who believes in learning from the successes of others.
5. May 2009 at 04:52
Finbar:
Forget banking. Suppose a business has a project that will only earn enough money to pay off all of its debt in 10 years, but borrows only for 5 years. It plans to refinance after 5 years.
Now, can the firm actually refinance after 5 years? It depends on whether or not lenders believe that it will be able to pay off after five more years. That is, will it earn enough on the project for the next five years.
You are claiming that this plan to refinance is inherently distablizing. No it isn’t.
If the project had been funded for 10 years, and it will not be able to be paid off, and this is discovered early on, the value of the bonds will fall to reflect the expected default. It is true that the firm will not have to default until year 10, but the real situation is the same.
If the project needs to be refinanced after year 5, then it is like there is a new project. Will it pay off or not?
Now, if banks lend for 10 years and fund the the loans with 5 year CD’s it is the same thing.
Just like the money earned by the firm’s project is what is important, so is the ability of the banks to collect on the loans when they come due.
The problem isn’t that banks are planning to refinance. It is that their loans aren’t going to get repaid.
Your contagion theory is that one bank fails and then people won’t lend to other banks. (One nonfinancial firm doesn’t pay off its debt and then people won’t lend to other firms.) This only makes sense if you assume poeple foolishly thought that banks couldn’t fail. Or, of course, that the assume that all banks have the same asset portfolios.
Oh, what if they do have the same asset portfolios? If the portfolios are bad, then that is the problem.
Anyway, I will grant that it is possible that firms that plan to refinance will be unable to do so because lenders wrongly believe that the funds cannot be paid off. In general, if lenders wrongly believe that some project will not be paid off, they won’t fund it, and it cannot be carried out. From a God’s eye perspective, that is bad. And when God comes down and manages the economy, we can solve that problem.
Of course, a firm that plans to refinance will not be able to pay off its current creditors. What should happen is that firms that cannot refinance as planned, even if the projects are really sound and then just cannot find anyone to fund them, should go bankrupt.
Banks too. If a bank cannot refinance, then it should go bankrupt. The depositors take losses.
Now, explain why that adversely impacts the flow of output, income, and expenditure.
If there is no credit at all, firms self-finance out of retained earnings and households consume all their dispoaable income. There are lots of gains from trade that are sacrificed. That is why credit is valuable.
Here is where I think you go wrong. More lending means more nominal expenditure. A breakdown in banking means less lending. Therefor less nominal expenditure.
This ignores what the depositors do with their money. This ignores that if there is a bankrupcy, the debtors are paying less. For every borrower there is a lender. The only excepti8on is if there is monetary disequilibrium.
Bad loans and bad investments are… bad. Imagining that mandating that maturities always match will fix things… well, it is like Keynes arguing against stock markets, that owners should be married to their firms. The costs are less than the benefits.
5. May 2009 at 06:22
“necessary but not sufficient”. it’s a deal! As far as grading exams, i believe this will be of great help to you: http://www.concurringopinions.com/archives/2006/12/a_guide_to_grad.html
5. May 2009 at 06:26
Bill W,
>Your contagion theory is that one bank fails and then
>people won’t lend to other banks. (One nonfinancial firm
>doesn’t pay off its debt and then people won’t lend to
>other firms.) This only makes sense if you assume poeple
>foolishly thought that banks couldn’t fail. Or, of course,
>that the assume that all banks have the same asset
>portfolios.
An alternative to the irrational contagion story is the debt overhang story. Banks facing insolvency have perverse incentives to hide losses, which can turn small solvency problems into a full blown credit crunch affecting all banks. I don’t know if you have access to NBER, but a relevant paper by respected Chicago economists can be found at:
http://www.nber.org/papers/w14925
The importance of MT is so much more relevant to banks than other firms for the obvious reason that banks have FAR more leverage and a much more significant maturity mismatch.
This isn’t to say money doesn’t matter, but it suggests the form of monetary intervention matters, since one intervention may be more effective at eliminating overhang than another.
BTW, is Devin F really arguing MT is flatout inefficient, or just inherently unstable, which leads to a tradeoff between having a low growth stable economy or having higher growth with a low frequency boom-bust credit cycle?
5. May 2009 at 06:36
Not just in the 1930s. Canadian banks seem fine today, after initial losses due to subprime.
5. May 2009 at 07:19
I have read that Europe’s banks were much more highly leveraged than even US banks. According to this link (http://blogs.wsj.com/economics/2008/09/22/european-banks-too-big-to-rescue/), Deutsche Bank, for example, is also very highly leveraged. It doesn’t matter how parsimonious you are initially, if you invest foolishly (because of moral hazard, amongst other things) and take a lot of risk, you suffer, particularly when the economy goes south.
5. May 2009 at 07:42
Thruth:
I looked at the abstract of the paper you cited.
How exactly does the debt overhand and the unwillingness of banks to lend impact the flow of output, income, and expenditure?
Superficially, it looks like this paper, like so many others in this vein, is assuming that an increase demand for liquidity generates an excess demand for base money and hence monetary disequilibrium.
We have a cure for that problem.
So now, we are left with a model where problems in the banking system result in an inability to produce goods and services that can be sold because someone in the production process cannot get credit.
It is an imbalance problem where some firms make profits and can fund investment. Other firms have production oportuntities and cannot get funding. So, with nominal income growing on target, the impact is shortages of goods and services. A supply side problem and “stagflation.”
Well, let’s see if that is really a problem.
All the evidence I have seen (as opposed to partial models) is that credit problems are reducing demand–sales. In other words, aggregate demand is falling. Further, we are not seeing strong expansion in some areas with shrinking in others.
This is a serious problem. It appears that quantitative easing is not being used because many people, including a susbstantion portion of the FOMC believe that the “problem” is with credit markets and they are trying to direct credit into securitization.
I don’t disagree that if they can rebuild the house of cards that the demand for base money will fall again.
(Scott, if the market clearing short term rates are back up to 4%, the .25% payment on reserves will be irrelvant.)
My view is that the market should handle credit markets. That trying to have policy makers fix them is a fools errand. But that this should all occur within a framework of nominal income growing that trend growth rate of the real economy.
Would there be some kind of real business cycle with stagflation because of credit problems? And then “booms” assocated with price deflation? Maybe. Let’s see.
5. May 2009 at 09:00
Scott: I don’t really know why Canada’s banks seem to have done well (touch wood). Maybe we just got lucky 😉
They have taken losses, but not bad enough to need bailing out. We had a mini-ABCP liquidity and solvency crisis, where it wasn’t clear whether the banks were on the hook, but it got sorted out. Our housing prices have only recently started falling, having peaked in mid-2008. Here’s our version of Case-Shiller: http://www.housepriceindex.ca/Default.aspx
But it doesn’t seem to be as simple as “Canadian banks are more tightly-regulated”.
1. We never had restrictions on interstate banking, so Canadian banks spread their assets and liabilities across Canada. (So it doesn’t matter if a local housing market goes bust).
2. We don’t have Glass-Steagal. The investment banks joined the retail banks some years ago.
3. We don’t have mortgage interest deductibility from taxes. So paying down your mortgage is a tax-free investment. So most people want to pay down their mortgages.
4. (Except in Alberta), mortgages are fully recourse. You can’t just walk away from a negative equity home and hand the keys to the bank; the bank will come after you for the difference.
I wouldn’t describe those differences as “Canada is more regulated”.
But we do have higher capital requirements. And mortgages over 80% must be insured (mostly by the government-owned CMHC).
The biggest and most important difference seems to be the *style* of regulation — “principles-based” in Canada, vs “rules-based” in the US. And a difference in banking “culture”. I and my commenters discuss it here:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/04/canadian-vs-us-bank-regulation.html
But I really don’t know.
5. May 2009 at 09:18
Bill W:
>Superficially, it looks like this paper, like so many others
>in this vein, is assuming that an increase demand for
>liquidity generates an excess demand for base money and hence
>monetary disequilibrium.
>We have a cure for that problem.
What are we talking about: Printing money, penalizing reserves or quantitative easing? Does it matter which? (I’ll concede that Diamond and Rajan don’t really consider any of these. What they do show is that the Bagehot rule doesn’t solve the problem and can in fact exacerbate it, which you seem to agree with.)
I think Scott mentioned and disagreed with this before, but don’t some macroeconomists view printing money as fiscal policy? (e.g. seems to fit De Long/Krugman’s narrow view on monetary policy) I’m still somewhat suspicious about whether penalizing reserves alone will be enough to deliver the required liquidity. I’ll have to re-read some of Scott’s posts and the threads before I comment further…
Scott: Looks like you have a De Long blog post (or two) to respond to http://delong.typepad.com/sdj/2009/05/i-will-never-understand-chicago-today.html
5. May 2009 at 11:37
I don’t know whether penalizing reserves are enough or not. I favor open market operations of whatever amount needed. All the T-bills and then on up the yield and risk curve as necessary.
DeLong and Krugman can call it what they like. I think it is obviously monetary policy. The notion that targetting the interbank loan rate using open market operations in T-bills is he limit of monetary policy is just absurd.
At some point, I would go with a currency suspension, but no need to worry about that as long as there are plenty of high quality securities for open market operations.
I don’t have having the central bank make loans to banks against bad assets. I don’t favor using central bank loans to try to jumpstart securitization.
Anyway, bagehot’s rule, as I understand it, is to lend freely at penalty interest rates. We already learned in the Great Depression that using historical standards isn’t too sensible.
Scott (and I) favor quantitative easing. Paying interest on reserves just means that more of it is needed. Stopping the interest on reserves and going to a penalty rate just means that less is needed.
5. May 2009 at 16:10
Bill said:
Why does the inability of a firm that plans to refinance but cannot cause other firms that had planned to refinance to also be unable to refinance? If all firms that had planned to refinance cannot, why does this adversely impact the flow of income, expenditure, and output?
Bill, you assume that expenditure is funded by income. However, we have an economy where expenditure is also funded heavily by perceptions of wealth via asset prices. Home prices (as well as other assets) are driven by the cash flows they generate but also by the ease and terms with which buyers can obtain credit to buy those assets. As asset prices climb further away from the underlying cash flow at some point a lender will not be able to refinance. That reduces the availability of mortgages (in the house price example) and therefore lowers home prices and the perception of wealth and reduces AD. Certainly enough QE could be performed to offset the AD affects of declining home prices but then you really are propping up a Ponzi scheme by allowing those who bought the most overvalued assets to delay (or avoid) bankruptcy as their income increases via QE. You may avoid some short-term pain but at the cost of keeping all the wrong people wealthy and making asset purchase decisions.
5. May 2009 at 18:59
Bill-
Suppose a business has a project that will only earn enough money to pay off all of its debt in 10 years, but borrows only for 5 years. It plans to refinance after 5 years.
Let’s say that the business is investing in fruit trees that take ten years to grow. Only in the tenth year after it sells the fruit will it be able to pay back the loans. Let’s say there are two possible lenders bidding for the deal. One lender wishes to send their kid to college in five years. The second lender wishes to retire in ten years. It would be a very dumb move for the lender who needs the money back in five years to invest in the fruit business. There is no way that the fruit business will be able to make the loan payment, so the lender faces significant roll-over risk in addition to the classic entrepreneurial risk. The lender who needs the money back in five years, is always better off lending out to finance a business that provides returns in five years. The lender that needs the money back in 10 years will do the deal.
Note that nothing stops the lender and borrower from negotiating incentives and callbacks into the deal. For example, the lender may only provide half the loan at first, and provide the second half when the business appears to be working. Or the lender might have provisions to recall the loan if there is some adverse business event. There could also be potential for refinancing halfway through the loan. The only guideline is that the lender should not expect its money back for ten years. As long as the lender is willing to forgo use of the money for ten years, there is no maturity transformation risk.
Note, also that this is a relatively mild case of maturity mismatching. The lender who wants his money back in five years has a fairly good chance of reselling the loan and getting cash. But if everyone is mismatching, the roll-over risk dramatically increases.
Your contagion theory is that one bank fails and then people won’t lend to other banks. (One nonfinancial firm doesn’t pay off its debt and then people won’t lend to other firms.) This only makes sense if you assume poeple foolishly thought that banks couldn’t fail. Or, of course, that the assume that all banks have the same asset portfolios.
Roll-over risk is an entirely different animal than entrepreneurial risk.
Imagine a maturity matched world. Lenders put their money in a 15-year CD. The bank issues 15 year mortgages. Let’s say the bank screwed up its risk evaluations. An abnormal amount of loans default. The CD’s only return .95 cents on the dollar. What happens? Well probably nobody lends at that bank anymore. The bank will probably end up shutting down. The lenders lose 10% of their money. People become less willing to lend to banks, but this simply results in fewer new projects using round about methods of production. This unwillingness to lend does not have any impact on the default risks of existing projects. People will spend their money in other ways, so there is no change to aggregate demand. There is no contagion.
Now imagine a maturity mismatched world. Lenders put their money in 3-month CD’s. The bank then lends out for 30 years. Every three months some lenders roll-over, and some redeem for cash. The bank must always find new 3-month lenders to pay off those redeeming. Now imagine there is just the rumor that the bank made bad loans. The rumor spooks lenders, and not enough new lenders are available to pay off the 3-month CD’s that need redemption. The bank defaults on a few CD’s. This spooks lenders even more. Now no one will buy 3-month CD’s. The existing holders all ask for redemption. Now the bank cannot pay anyone. People who have money in other banks get spooked from hearing the same rumor. They all rush to withdraw. Demand for 3-month CD’s dries up entirely. Banks fail on their redemptions left and right.
The banks are now forced to liquidate to sell of their 15-year mortgages. There are three problems with this. First, everyone else is doing this at the same time, suppressing the prices. Second, all the would be buyers cannot redeem their CD’s, and so do not have the cash to buy the 30-year mortgages. Third, during the normal times the bank had taken advantage of the maturity mismatch to offer lower interest rates on long term loans. Since the 3-month CD’s only cost 2% interest, it could offer home loans for very low interest, maybe 3 or 4%. But when it liquidates, it must find a buyer who is actually willing to hold the loan to maturity. That buyer will demand a much higher interest rate than 4%. They will demand maybe 9 or 10%, or even higher. Thus the price of the asset falls dramatically. So even if the bank manages to liquidate, it can only return fifty cents on the dollar to the original CD holders.
To recap, in the maturity matched scenario, real mistakes in risk estimation and poor entreneurship led to losses, but no contagion. This is a functioning banking system. It is taking risks, but it is sound.
In the maturity mismatched scenario, the mere rumors of losses can cause the entire banking system to collapse. The maturity mismatched scenario is inherently unstable. The game theory incentive is to always cash out before everyone else does. Thus the Nash equilibrium is the bank run.
The only way a maturity mismatching can survive is if government props it up. In the first half of the 19th century State governments routinely suspended redemption. In the second half, the National Banking Act essentially banned all maturity matching banks. The Federal Reserve System had maturity mismatching at its core. The shadow banking sector and the money market funds grew big because they were “too big to fail” ( and indeed the money market funds have been bailed out).
This ignores that if there is a bankrupcy, the debtors are paying less. For every borrower there is a lender.
It’s not the bankruptcies that cause the fall in aggregate demand. It’s the collapsing credit bubble. Let me try explaining with a different tactic.
Let’s say that everyone in the economy has their savings equally split between cash and equities. Let’s also say that due to some sort of market manipulation scheme or temporary imbalance between buyers and sellers, the price of stocks has been abnormally high for the past ten years. Everyone has grown accustomed to this new high level and has believed the high level to be permanent. They have changed their spending habits accordingly. One day, the scheme collapses, and stock prices collapse, dropping in half. Note the underlying dividends have not changed. Business is fine. It’s just the stock price (the last sale price) has changed.
Everyone logs into their online investment account, and is shocked that their paper net worth has fallen by 25%. Everyone has a preferred paper wealth to expenditures ratio. They start cutting spending to restore this ratio. Demand for cash increases. Aggregate demand falls, the recession starts.
Do you buy this story – that a collapsing bubble, if big enough, can cause a fall in aggregate demand? If yes, I will follow up and explain how this connects to maturity transformation.
Thruth-
Maturity mismatching without a lender of last resort will always collapse. It’s like a Ponzi scheme.
Maturity mismatching with “wink and nod” backing of the state tends to be cyclical. The system grows big as banks think they will be bailed out or have access to the lender of last resort. But when some or many are actually allowed to fail, the system contracts sharply. This is the business cycle.
Maturity mismatching with ironclad guarantees ( FDIC ) is essentially equivalent to the government printing money to subsidize lending. Perhaps printing money to subsidize lending is a good idea. But if so, it could be done directly, consistently, and honestly.
Scott-
Regarding the bank run, my understanding is that the key problem was solvency, not liquidity.
The actual losses from the subprime loans were in the low hundreds of billions. But the actual total drops in prices of securities concerned were in the trillions. The bulk of the damage to the banking sector came from the bursting maturity transformation bubble, not from the defaults directly. ( I still consider this a solvency problem. The use of the term “liquidity crisis” is Orwellian. The assets are perfectly liquid at the market price).
6. May 2009 at 05:51
I will skim over some of the banking MM stuff, which I am not qualified to address.
Alex, Very funny!
David, The key line is “when the economy goes sought.” Which I interpret as when (and if!) the Fed allows NGDP growth expectations to fall sharply. It need not ever happen.
Bill, I agree that with 4% short rates the interest on reserve problem vanishes.
Thruth, I don’t think DeLong is referring to me, I don’t think that quote is from me either. I view monetary policy as when the Fed injects money by buying assets at market prices. Some purchases may be risky, but that has always been true, even T-bonds have price risk. There might, ex post, be some capital gains or losses (either are equally likely if bought at market prices) but that hardly disqualifies it as monetary policy. Of course in the end monetary and fiscal policies can never be completely separated, as the “fiscal view” points out.
Mark, I don’t think Bill’s 3% NGDP rule would bail out Ponzi schemes, they’d have to rise and fall on their own merits. We should focus on sound macro policy without thinking about how it impacts Ponzi schemes, and I think that is what Bill tries to do.
Devin, Those additional losses beyond the subprime losses, weren’t some of those in other mortgages, which increasingly went bad as NGDP fell? That was my impression. I think if NGDP had kept growing at 5%, then those other losses wouldn’t have ballooned into the trillions.
6. May 2009 at 16:46
Mark:
Here is how that works. Stock prices fall. Net worth falls. Households want to rebuild their net worth. They save more. There is less consumption spending out of income. The funds that are saved are used to purchase financial assets. This reduces interest rates. Firms invest more. The sell financial assets to fund purchases of capital goods. There is a reduction in spending on consumer goods and an increase in spending on capital goods. Aggregate demand is not effected.
How could this go wrong? Lower interest rates cause an increase in the demand to hold money. The nominal quanity of money does not rise to meet the demand. To clear the increae in money demand, the price level needs to fall, increasing the real demand for money to meet the supply.
The increase in real money balances increases wealth (counteracting the fall in stock prices.) This added wealth motivates both an cinrease in consumption and an increaes in investment.
In other words, the deflation of prices and nominal incomes results in a recovery of expenditure financed out of income.
I never would suggest that the only think that determines expenditure is the flow of income.
When stock prices fall, those who sold the stock have money. What do they do with it?
6. May 2009 at 16:58
Finbarr:
As usual, in your story about the how problems with maturity mismatch, there was “there is not enough cash?’ Why not?
What do those who are no longer purchasing the short term debt that is financing the longer term debt do with the money?
Implicit in all of these stories is an increase in the demand for money. People don’t want to purchase short term debt, so they hold money.
The solution to all of these “problems” is to simply increase the quantity enough to meet the demand at the target value of nominal income.
If that is done, nominal income is not effected. What might happen, is that there is a change in the allocation of resources. For example, people might consume more because they can only receive a sufficiently high interest rate for them to postpone consumption if they tie up their funds for too long. The interest rate they are offered to fund short projects is too low for them, and they purchase consumer goods and services.
Or perhaps projects that provide a return more quickly are undertaken. Of course, the only reason a longer one would be undertaken is that it is more productive.
So, an unwillingness to match maturies results in more consumption and less productive investment.
If there was maturity mismatch, and then it changes so that there is none (or at least less) then there is a shift in the allocation of resources towards more consumption and less productive investment before. Investors earn less.
I don’t have any doubt that this process could increase the demand for money. And if the supply of money fails to adjust, there will be a shortage of money and nominal income would need to fall.
The solution is to increaes the quantity of money to match the demand.
Your claims that maturity mismatches (creation of liquidity) can only exist with government intervention are inconsistent with the historical record.
The additional yield motivates people to do it all the time.
7. May 2009 at 05:13
Scott: didn’t mean to imply that De Long’s post was directly addressed at you or that he was quoting you. But clearly he’s playing on your lawn with a can of gas, matches and some spray paint.
I see Buiter now has an interesting post on negative rates.
7. May 2009 at 18:10
Thruth, DeLong has a model with (I think) nominal interest rates and real GDP. My concern is with the supply and demand for base money, and nominal GDP. If we are at less than full employment, like right now, higher NGDP will probably raise RGDP. So if we can make that assumption then we can talk in terms of “Y”. But what to do with nominal interest rates? Or real interest rates for that matter. I assume that any expansionary policy that is expected to be effective will raise rates, at least most rates. But even that assumption depends on all sorts of issues:
1. Is the policy credible (in terms of raising expected NGDP growth?
2. Is it a one-time increase, or an increase in the money supply growth rate?
3. Are you buying T-bills, or a wide range of assets?
4. Is the policy expected to increase RGDP growth, or just inflation?
All this will influence whether nominal rates or real rates go up. It will influence whether short rates or long rates go up. It will influence whether risky rates or safe rates go up.
Whatever happens, I suppose there must be some way to explain it in terms of IS-LM. But we already know that the more easily identified a monetary shock, the less IS-LM seems an appropriate model for explaining that shock. The best example is the most identified monetary shock in U.S. history, the devaluation of 1933. According to (the simple version of) IS-LM, nothing of interest happened in 1933. According to the Woodford forward–looking new Keynesian model, 1933 was one of the largest expansionary shocks in U.S. history–probably the largest. That’s what Keynesians should be talking about (in my view, of course I’m not a Keynesian.)
I did read the Buiter piece earlier. At least Mankiw said his was a joke. Buiter seems to think his idea is a serious proposal. The only idea that is somewhat serious is electronic money, but that’s still decades away as a replacement for cash. And until you replace cash, these ideas won’t go anywhere. I’m not trying to criticize Buiter, it’s good to see people talking about negative interest rates. But it frustrates me that so much ink is spilled over ideas that won’t help in this crisis, and the big names are ignoring negative interest on reserves (except one big name–Robert Hall.) It’s this Keynesian idea that money policy works through short run rates, not the supply and demand for base money.
9. May 2009 at 15:06
As usual, in your story about the how problems with maturity mismatch, there was “there is not enough cash?’ Why not? What do those who are no longer purchasing the short term debt that is financing the longer term debt do with the money?
They hoard the money. This increases the demand for money, causes deflation, etc. The MT story is important, because it tells us why demand shocks happen. Once we understand why demand shocks happen, perhaps we can find some preventative medicine.
The solution to all of these “problems” is to simply increase the quantity enough to meet the demand at the target value of nominal income.
I agree.
If that is done, nominal income is not effected. What might happen, is that there is a change in the allocation of resources. For example, people might consume more because they can only receive a sufficiently high interest rate for them to postpone consumption if they tie up their funds for too long. The interest rate they are offered to fund short projects is too low for them, and they purchase consumer goods and services.
Sure, I agree.
So, an unwillingness to match maturies results in more consumption and less productive investment.
Your sneaking in a normative judgment with your use of the term “productive”. An unwillingness to mismatch securities results in consumption and investment decisions reflecting people’s actual desire to forgo consumption. Should the government strong arm people into investing more? I don’t think so. But even if it did, it would be a lot more stable to just print the money at a fixed rate every year and invest it directly. Same result, but no business cycle.
Your claims that maturity mismatches (creation of liquidity) can only exist with government intervention are inconsistent with the historical record. The additional yield motivates people to do it all the time.
Well, certainly it can exist for a short time in the completely private sector. So can Ponzi schemes. But widespread, economy-wide adoption of maturity mismatching has always required government support. In the English model, there were regular suspensions of redemption and bailouts by the Bank of England. In the American model, most of the “wild cat” banks were supported by the state governments and had suspensions of redemption during runs. Then the National Banking Act made maturity mismatching a nation wide practice, and essentially banned non-maturity mismatching banks.
What are your examples of economy wide maturity mismatching existing without government support?
We seem to have agreed about a bunch, but still disagree about a few things. In order to see if our disagreements are meaningful in practice, let me state a few potential policy proposals that are implications of my arguments about theory. I’d interested in seeing how close or far apart we actually are.
1) The government should explicitly renounce all “too big to fail” or “lender of last resort facilities”. The FDIC should be abolished. Banks should stand or fall on their own. Privatize the profits, privatize the losses.
2) Since the entire system is based on maturity mismatching, and it would collapse in a second without government guarantees, there would have to be a very careful transition phase. Dollars should be printed, and FDIC insured accounts and money market accounts would be replaced by 100% accounts holding actual dollar bills. A transition agency would give people guidance over which banks were engaging in safe practices and which one’s dangerous practices. The most grievous forms of maturity mismatching would be banned during the transition.
3) I personally would desire a 0% rate of monetary expansion. That’s because all dilution is a tax on existing money holders. I do not want to be taxed, so I do not want monetary dilution. But, if the government does insist on seinorage, for its own sake it should keep the dilution rate under 3%, or else people start to seek out alternative stores of value ( gold, etc). And whatever the rate is, it should be consistent. The new money should just be printed and spent directly, rather than injected via a Bagehot scheme where one actual dollar becomes ten Ponzi dollars.
4) If for some reason a monetary demand shock happens, the government should print money to meet the demand. It should do so by making an accounting of all accounts backed by the full faith and credit of the government ( bank accounts, treasuries, and maybe money market funds ) and then simply multiplying by the desired expansionary factor. That allows the Fed to reinflate the exact desired amount needed, whether it be 5% or 60%, without killing demand for dollars or reallocating real wealth.
10. May 2009 at 05:24
Devin, Regarding point 4 “if for some reason a money demand shock happens” These occur all the time. Do you mean a “very big money demand shock?”
10. May 2009 at 19:01
Scott-
Devin, Regarding point 4 “if for some reason a money demand shock happens” These occur all the time. Do you mean a “very big money demand shock?”
Yes. And also, by money, I don’t mean “base money”. In the system I propose there would be no such thing as base money. I mean all fiat paper that is backed by USG and is guaranteed to trade on par with a dollar. I actually think that under my proposed system big shocks ( 10% or more in a year) would be very rare, hopefully non-existent.
18. May 2009 at 14:48
[…] Turns Out That Life Is More Difficult Than Anticipated Filed under: Catallactics, Curmudgeonliness, Economics, Sports — Kevin Feasel @ 7:37 pm – For all of those who were laughing at those silly Anglo-Saxon banks and how they are all failing…err…check out the German ones. […]