Tuesday, July 19, 2011

The Inadequacy of the Balance Seet Recession View

Thanks to this David Leonhardt article, the balance sheet recession view is once again getting much discussion.  This view holds that households acquired excessive amount of debt during the housing boom, the value of their assets plummeted during the crash, and now their balance sheet are in need of great repair.  Consequently, the U.S. economy is undergoing a great deleveraging cycle that is slowly restoring household balance sheets.  Some take this view to also mean that only time can heal the wounds of a balance sheet recession. 

I don't like this view for two reasons.  First, it is at best an incomplete story.  For every household debtor deleveraging there is a creditor getting more payments.  Yes, household debtors have cut back on spending, but so have creditors.  The creditors could in principle provide an increase in spending to offset the decrease in  debtors' spending.  They aren't and thus the economic recovery is stalled. In other words, the problem is as much or more about the build up of liquid assets by creditors as it is the deleveraging of debtors.  The balance sheet recession view, however, sees the debtors deleveraging as the main problem.  It completely ignores the creditors buildup of liquid assets and its implications for spending.   

When one begins to focus on the creditors' role, it becomes apparent that the underlying problem is excess money demand.  For if the creditors are not spending their newly acquired dollars there must be an unsatiated demand for money.  Even in the case where banks are the creditor, the excess  money demand problem is present.  For example, if a bank loan is paid down both loans and deposits fall. If those deposits were checkable, saving, small time, or money market accounts–assets used as money–the money supply falls too. For a given demand for money, this drop in the money supply now means there is--if not already--an excess money demand problem. The key issue, then, is to satiate creditors' demand for money and get them to start spending some of their money assets.  This insight is ignored by the balance sheet view of recessions.

The second problem I have with the balance sheet view of recessions is that it leads people to think there is nothing monetary policy can do.  Part of the issue here is the failure to see the underlying excess money demand problem.  If it were widely understood that the fundamental problem was excess money demand, then there would be more faith in using monetary policy.  

Another part of this problem, though, is a failure to look back to history for other examples of balance sheet recessions.  As Frederick Mishkin has shown, households were also significantly deleveraging during the Great Depression.  This experience would fit the standard definition of a balance sheet recession.  Below is a table from his paper that shows household balance sheets in real terms.  Note that between the 1933 and 1936 U.S. household underwent a cumulative deleveraging  in real terms of 20%. This is far more in percentage terms that has happened over the past few years. And yet between 1933 and 1936 the U.S. economy had a robust recovery.  Real GDP averaged almost 8% growth during these years. 

The balance sheet recession view cannot easily reconcile the large deleveraging by households and the rapid real economic growth that occurred between 1933 and 1936.  What can explain it is a more nuanced view that acknowledges creditors with excess money demand were confronted by FDR's original quantitative easing program.  This QE program was far better than recent ones in that FDR clearly signaled a price level target and backed it up by devaluing the gold content of the dollar and allowing unsterilized gold inflows. In otherwords, FDR signaled that he was going to allow a significant and permanent increase in the monetary base and followed through on it.  This change nominal expectations and caused creditors to start spending their money balances.  The same could be done today with something like a nominal GDP level target.

Unfortunately, I fear Edward Harrison is correct in saying Fed has burned up most of its political capital. So it is unlikely to try anything radical like nominal GDP level targeting.  That means the economy will be stuck in stall speed for the time being.

Update I: A quick follow-up point to some of the comments.  Whether consumers default or pay down debt is irrelevant to whether there is an excess money demand problem. If consumers default and hold on to their money balances there is an excess money demand problem.  If consumers pay down their debts and the banks (the creditor) mark down their assets and liabilities accordingly, there is still an excess money demand problem since there is now less money supply for a given level of money demand.(By the way, this New York Fed report shows that many consumers have been paying down their debts, not defaulting.)

Update II: Just to be clear, there are more creditors than just banks so it is a little misleading to focus solely on banks.


  1. I just want to get a big red stamp with the words "FALLACY OF COMPOSITION" and a picture of a winking Bill Woolsey. Then I want to stamp in David Leonhardt's forehead.

  2. I don't really get the point about the Fed using up political capital. Perhaps they would need to spend political capital if they needed permission to do OMOs, but they don't. Bernanke can say he ate eggs for breakfast, and therefore he's doing $200 billion of OMOs. He does not need to explain or justify his actions to anyone outside of the Fed. I guess my question is, why does the Fed need political capital when they already have obfuscation and lack of accountability in spades?

  3. James,

    Bernanke is the chairman of the FOMC; he is not the supreme dictator of monetary policy. The Federal Reserve is a complex institution: its policies cannot be reduced to the whims of one man. In any case, the Federal Reserve's much vaunted "independence" is a matter of degree. Moreover, that independence is like your right to privacy -- it will disappear whenever it becomes too much of an inconvenience to the powers that be.

  4. For a non-economist like me, the most striking thing about the excess demand for money story is that it is not the default explanation for these kind of events. It's just a supply and demand analysis, but with a few kinks because money doesn't have a market of its own.

    Perhaps it's too simple to be impressive.

  5. "What can explain it is a more nuanced view that acknowledges creditors with excess money demand were confronted by FDR's original quantitative easing program. This QE program was far better than recent ones in that FDR clearly signaled a price level target and backed it up by devaluing the gold content of the dollar and allowing unsterilized gold inflows. In otherwords, FDR signaled that he was going to allow a significant and permanent increase in the monetary base and followed through on it. This change nominal expectations and caused creditors to start spending their money balances. The same could be done today with something like a nominal GDP level target."

    So do you and all other economists just sit around and think about how to make the rich (probably including yourself) richer? Try working on fixing the balance sheet of the lower and middle class. Of course, you probably associate wage increases with price inflation.

    "It completely ignores the creditors buildup of liquid assets and its implications for spending."

    Well maybe if they weren't so rich from excessive real earnings and real earnings growth there would NOT be a buildup of these "liquid assets".

    "Unfortunately, I fear Edward Harrison is correct in saying Fed has burned up most of its political capital."

    Good! It is far, far, far beyond time to abolish the fed forever!!!!!

    "So it is unlikely to try anything radical like nominal GDP level targeting. That means the economy will be stuck in stall speed for the time being."

    Nonsense!!! You should focus on debt levels instead of this stupid NGDP targeting stuff vs. price inflation targeting stuff.

  6. David, I agree that the Leonhardt article and the AP Survey article, both of which you link to, are nonsense. I also agree that your proposal WOULD WORK. But I’m still not happy with boosting an economy via a relatively small proportion of households: the cash rich or any other small group (i.e. I rather agree with the comments by Anon just above).

    Also there is the question as to whether it is best to boost demand by encouraging the cash rich to spend by threatening them with inflation (your proposal, as I understand it), or whether it is better to pump more cash into ALL households. I prefer the latter, amongst other reasons, because if the hoarders only spend when threatned by inflation, presumably they’ll start saving again when inflation subsides, and that means Keynes’s paradox of thrift (what you call excess demand for cash) rears its ugly head again.

  7. BBUUUTTT ..... isn't a lot of the deleveraging due to insolvency or foreclosure which doen't result in cash piling up in the hands of creditors?

    And in the great depression isn't part of the secret that make work programs helped with the deleveraging by keeping up the flow of wage income.

    I happen to agree with you (and Paul Krugman for that matter) that an increase in inflationary expectations would help - but I also think (like for instance Steve Waldmann from interfluidity) that there is a class of people who use marginal income mainly for collecting financial assets and another class of people (somewhat poorer) who are more likely to use marginal income for consuming goods and services. Expanding the money supply by printing it and giving it to poor people is the best solution of all.

  8. I think I should also point out, that the class of creditors who like to accumulate wealth, could of course start making real productive investments with their money instead of spending it on luxuries or lending it again. But then they would need an international financial system that ensured appropriate (to trade flows) real exchange rates or consumers willing to increase their spending! Giving money to poor would help these people to find good investments as well.

  9. Anonymous,

    An excess demand for money makes money scarcer. The efforts one must exert to earn a given quantity of money is greater than before. In other words, the "price" of money increases relative to all other goods. This increases the real burden of nominal debts; debts which were manageable before the excess demand for money become overwhelming. The poor state of household balance sheets is a largely consequence of the excess demand for money.

    It may be that even in monetary equilibrium, there remains too much debt, and that is a problem for the market to resolve. However, excessive debt isn't going to cause a major recession by itself -- an excess demand for money is needed for that.

  10. Ralph,

    It isn't about saving vs. consumption. The problem is an increase in money demand without a proportional increase in money supply.

    The "paradox of thrift" is a misnomer: it has little to do with saving and everything to do with money demand. In fact, it is possible for total savings to be falling while the so-called paradox of thrift occurs.

    Most of the time, people do not save by holding larger money balances, but by increasing spending on capital goods. Most of the time, saving does not reduce aggregate demand.

    However, when interest rates approach zero, strange things begin to occur. Because the Fed fixes the interest rate on base money near zero, it effectively places a floor on all interest rates. If the equilibrium interest rate is negative, then there will be a frustrated demand for safe and liquid assets at the lower nominal bound. This frustrated demand will spill over into money, and potentially cause a major recession.

    There are three good solutions I can think of.

    (1) Adjust the supply of money so that its equilibrium interest rate is always near zero. If the equilibrium interest rate falls below zero, then it follows the money supply is too small.

    (2) The Fed can start doing interest rates on reserve balances relative to rates on other safe and liquid assets. For example, it can always offer less interest on its accounts than what is offered on near substitutes (like T-bills). If the interest rates on near substitutes approaches zero, then the Fed will begin charging negative interest rates. This will abolish the lower nominal bound, and would require significant institutional reform.

    (3) My favourite: abolish the Fed and let free banking take over.

  11. What a terrific blog--and so painful to read.

    We have met the enemy, and he is us. Or Richard Fisher of the Dallas Fed, a living menace to the US recovery.

  12. It would be easy to call this a matter of semantics but I view semantics as more important to thought than most realize. My problem is with the phrase excess demand for money. If anything, creditors have an excess demand for higher return investments or higher future consumption. It is that these demands cannot be fulfilled by themselves that lead to what can be equally cslled, an excess supply of money in their hands resulting in a deficient supply of money for the economy as a whole. It is not money that is demanded but invesstments with better expected returns and greater future consumption. A deficient supply of worthwhile investment is closer to the truth. More money can boost returns and incomes to provide these, but it is these that are sought, not money, money is just the means to provide them.

    While it could be misleading to say the recession was caused by individuals rebuilding their balance sheets, it is not to say that doing so make the decline larger and recovery slower for an equivalent level of monetary intervention and that this necessitates a greater level of intervention. To ignore the rebuilding of balance sheets, a critical factor in such recessions and recoveries is a mistake. The solution is not to attempt to prevent or dissuade rebuilding balance sheets but to support sufficient growth to allow them to do so. Monetary policy could work, but if it is too much for anyone to contemplate actually doing it can it? Since it has to rely on stimulating creditors rather than debtors, it has a more difficult task to do and if we can't manage to do that then fiscal monetary policy in the form of a free money drop may be more effective.

  13. Lord:
    "if we can't manage to do that then fiscal monetary policy in the form of a free money drop may be more effective."

    Everyone laughs at me, but I say we we drop grocery bags of money onto street corners in lower-middle income neighborhoods all across America. They would spend the money, stimulating growth.

    The inflationary impact would be minor, and manageable.

    There is a matter of moral hazard--people getting money for no work or risk to capital.

    My second option is a national lottery that pays out more than it takes in. And it pays out in $50 to $1,000 wins, not winner-take-all.

    So one guy gamble $50, and win $150, while the next guy gambles $50 and loses it. The extra $50 enters the economy, and the precious moral hazard threat is sidestepped.

    What is sad is that I think the Beneficent National Lottery would actually work--for a while, and that's all we need.

  14. David,
    A couple of questions and observations: As you noted, both creditors and debtors seem to be deleveraging simultaneously. The question in my mind, is which side is deleveraging more? If creditors are understating liabilities, (I think they are), then it would stand to reason that the effect of excess demand on the creditor side is where the problem lies. When does an asset become a liability? When it doesn't get repaid, and when there is an expectation that it will not be repaid at face value? The suspension of mark to market rules have allowed financial institutions to overstate assets while understating liabilities. No? This was supposed to allow institutions to meet reserve requirement and still free up capital for investment. As I noted in a previous comment thread, there appears to be an interest rate paradox that has occurred. Perhaps this is a direct result of the fact that banks and financial institutions realize that their exposures are understated and the risk premium on lending new capital is not enough to offset the current set of understated liabilities.

  15. Nanute:

    What do you think it would take to solve this paradox?

  16. I'm confused - how are the creditors supposed to "spend" the cash they are receiving from loan repayments if there's very little demand for loans by potential debtors? We are already near the zero bound of interest rates, not sure what more banks can do. Of course, this also ignores potential "lurking liability" on bank balance sheets that could cause them to desire to deleverage at the same time due to capital adequacy concerns.

  17. And a follow up to my previous comment (confusion over how creditors can spend cash from repayments in a zero-interest rate environment)...another factor you have missed is that a large proportion of the deleveraging is occurring through default and foreclosure, meaning capital is getting hit at the same time as debt is "paid down."

  18. Anonymous:

    You make a fair point that banks themselves may be deleveraging given the "lurking liability" problem. I think that was Nanute's point above too.

    But on your first question, the answer is that if the Fed did something as radical as what FDR did in 1933 (say adopt NGDP level targeting) it could in principle change expectations that would kick start the economy and increase the demand for loans.

    The idea is to increase expectations of higher nominal spending, which in an economy of excess capacity would also mean higher real growth. That in turn would increase the demand for investment and consumer spending. A lot of that spending would initially come from existing money balances of creditors. But eventually the pickup in economic activity would cause the demand for loans to increase and banks would accommodate.

    Read the FDR link above. FDR was able to turn around a far worse economy for almost four years, but ultimately was thwarted in 1937 by the tightening of monetary and fiscal policy. Also, see Gautti Eggertson's paper, "Great Expectations"

  19. Anonymous:

    Actually, most of the deleveraging is not occurring through default. See this New York Fed report:

    Either way, though, money demand remains elevated. Either (1) the consumers defaults, hang on to their money, and do not spend or (2) the consumers pay down their debts, the banks (the creditor) mark down their assets and liabilities accordingly and thus lowering money supply for a given level of money demand.

  20. David:
    Are you conceding my point? lol. I think you've answered the question; via an FDR style price level target. The problem with this solution is political in nature. Things are going to have to get much worse before this approach will gain any traction on the fiscal side (congressional) of the equation. The Fed has the power to expand the money supply, but even here, political and philosophical issues make this an unlikely option. It is no doubt, most probably the 1st best choice.
    If I had the power to fashion a solution, it would be a combination of consumer debt forgiveness coupled with direct money transfers to consumers; particularly at the middle and lower end of the economic ladder. Now, I can hear the critics arguing about the moral hazard of debt forgiveness. I would answer that if we did it for finance, why not for borrowers? The government guarantees student loan and most mortgage debt through the GSE's. Why couldn't the government write off some of this debt? QEI transferred a significant amount of debt from the private sector to the public sector and no one seemed to have a problem with it.

    I would also stop paying interest on excess reserve deposits. In this environment, it is probably a good idea to penalize (tax), reserves. I understand that paying interest on excess reserves is being utilized as a means to repair balance sheets. Nevertheless, if you want to induce lenders to lend, you need to create an incentive to do so. Either erode the value of money throught the inflaiton rate mechanisim, or tax it above the level of "savings."
    Thanks for the consideration.

  21. Nanute:

    Yep, you are sharpening my thinking. A payroll tax holiday funded by the Fed until some nominal level target is hit seems reasonable to me. So I am with you on the direct money transfers to the public. The longer this morass goes on, the more sympathetic I am becoming to some kind of debt forgiveness program too. These, though, seem like politically infeasible. And so does a NGDP level target for that matter. What to do...

  22. Warren Mosler (leading light of the Modern Monetary Theory movement) has advocated a Fed funded payroll tax reduction for a long time. While this benefits a wide section of the population it does not benefit EVERYONE, e.g. the unemployed or pensioners. So ideally unemployment benefits need to rise in proportion, and in countries with a state pension, that can be raised.

    Re the amount of deleveraged money which is simply extinguished rather than ending up as spendable cash in someone’s pocket, this Credit Suisse paper is of relevance. See in particular the charts on the first four pages or so.

  23. David:
    Thank you. Notice, I did qualify my solution with the caveat, "if I had the power." (Insert Cowardly Lion dialogue here. If I were the king of the forest....)

    So many good options to stop the bleeding, and no one with the political will like FDR to step up to the plate. What to do? Sit back and watch the train wreck. It will seem as though it is in slow motion. Unfortunately, the pain caucus wants its pound of flesh, but I'm afraid they will bite off more than they can chew.

    I've learned a lot since coming here, and look forward to continuing the experience. Again, thanks for the compliment.