Monday, November 1, 2010

How Would QE II Help?

Mark Thoma is doubtful QEII will pack much economic punch:
It seems to me that everyone fighting today over whether QEII will work are worried about whether the Fed can affect real rates, but are forgetting about the second step in the process. Once real rates rates fall, firms and households then have to be induced to borrow more, then consume or invest (I'm including the response to expected inflation in this). Even if we manage to change real rates, and I have never quarreled with the Fed's ability to do this (though the extent depends upon their ability to affect expectations), why do people think it will bring about a strong consumption and investment response in the current environment?...
Here is why I think QE will pack an economic punch, if done correctly. The expectation of permanently higher prices will cause cash-flushed firms, households, and other entities to start spending more today. Right now there is an excess money demand problem that could be stemmed by meaningfully changing the inflation outlook.  Those folks and entities hoarding money would on the margin face an greater incentive to start spending given an significant increase in inflation expectations. (Yes, many households with weakened balance sheets are deleveraging and saving more.  However, the rise in saving by these troubled households--by paying off debt, cutting back on spending, or buying other assets--should lead to more money  for other non-troubled households unless the money is being hoarded somewhere else.  Maybe the non-troubled households choose to sit on their money, maybe the creditors to whom the troubled households send their  money are sitting on the money, or maybe it's the creditors' creditors that are sitting on the money. The details are not important, what is important is that somewhere in the economy there is an excess demand for money right now that is not being met by the Fed.) 

Now assume the Fed does indeed address this excess money demand problem with QEII.  Given sticky wages and prices, this pickup in spending (i.e. drop in money demand) will translate into real economic gains.  This will encourage banks to start lending more as they see better credit risk going forward while the improved economic outlook encourages firms and households to start borrowing more too.  On top of that, the higher expected inflation will drop the real interest rate and encourage more interest-sensitive spending. Next, we could consider how the pick up in asset prices might have a wealth effect on consumption.  The pickup in asset prices could also improve troubled households balance sheets and thereby enhancing their access to credit.  Finally, further depreciation of the dollar may spur exports. Bottom line is that there are multiple channels through which QE could work.


  1. Again, as I mentioned in my last comment (on the other thread), there is a lot wrong with your theory regarding QE.

    Your theory is based around the idea that corps and households are hoarding cash, but that's simply not true. You're ignoring the liability side of the ledger. Non-financial corporations, have in fact, never carried more debt than they do today.

    HH's are also carrying near record amounts of debt. The reason why there is demand for cash is directly due to this. Everyone wants to be liquid in case asset prices fall further. The problem is, the debts need to be dealt with and QE does nothing to address this as Japan, UK and QE1 in the USA showed that interest rates and refinancing effect do not occur.

    In fact, if anything, QE appears to promote more recklessness - "kick the can" policy.

  2. And the corp debt levels entirely ignore the fact that our banks are still largely insolvent....

  3. MarkS:

    You are overlooking a key point: for every debtor there must be a creditor. Thus, for every indebted firm and household there is creditor. Sure, the debtors quit spending as much, but the creditors should be offsetting this since they are receiving more money. Thus, the only reason there can be a problem is that the creditors are increasing their money demand. And QEII can get the creditors to start spending their money.

  4. MarkS,

    A lot of people miss the point that for ever debtor there must be a creditor. Here is a recent post where the implications of this for excess money demand are fleshed out more:

  5. The best way for expectations to make QE to work is for households and firms to sell existing bonds that they are holding and buy goods and services.

    These could be short term bonds or long term bonds.

    Whether this is a decrease in the quantity of bonds demanded due to lower nominal yields on long term bonds (the Fed buying the long bonds) or else a decrease in the demand for long and short bonds because of expectations of higher inflation, so that nominal rates rise less than expected inflation,(which would be a decrease in the quantity demanded due to lower real rates) or a decrease in the demand for long and short bonds due to expectations of higher future real incomes (motivating more consumption) or higher real profits at existing quantities of capital goods (motivativing more investment) doesn't matter too much.

    Nominal and real interest rates can go either way.

    Lower nominal rates, lower real rates, already indebted people borrow even more, and they buy more goods and services, is only one possible pathway for QE to work, and probably not the best.

  6. David- One thing I never quite get in the "for every debtor there is a creditor" formula as used in mainstream economics, is that it seems to ignore debt default.

    As I understand it, right now the increase in net saving in the US is primarily being driven by debt default on mortgages, CRE loans and credit cards. I fail to see how that makes creditors better off!

    In any credit crunch there is a similar story, asset prices fall and debtors begin having trouble making payments. The assets held by creditors are now worth less, in fact they may switch positions from net creditors to net debtors.

    I assume, of course, that you know this. But, though the whole debtor-creditor thing is an accounting identity, it doesn't seem to explain economic behavior the way economists typically present it.

  7. OGT: Isn't that one of the arguments put forward for QE II : that if the policy succeeds in reflating nominal income, people can service their debts and so end defaults.

    Of course, it should be obvious that this is an extraordinarily ineffectual way to go about it. Much more effective to do it by fiscal transfers to income constrained households, ultimately monetized.