Sunday, September 29, 2013

Is QE contractionary?

I ran across a fascinating blog post by Peter Stella at Vox-Eu on exit strategies and QE. 

Peter points out that only banks can hold reserves, while anyone can hold short term Treasuries. And you can easily use Treasuries for collateral.  That means that short term Treasuries are in some sense more liquid than reserves, and that by buying huge amounts of Treasuries and issuing reserves, the Fed may be actually contracting. 

In Peter's words:
Large Scale Asset Purchases (LSAPs) have inadvertently caused a significant change in the composition of assets available in the open market.
  • The stock of marketable, highly liquid, AA+ collateral fell by trillions (disappearing into the Fed’s portfolio, i.e. System Open Market Account).
  • The stock of assets available only for interbank trade (bank reserve deposits at the Fed) rose by trillions.
..Treasuries and Fed deposits are equally safe. But they differ significantly in their marketability. Anyone can trade Treasury securities; only banks can exchange Fed deposits. ... 
  • Banking and money creation has not worked for at least two decades in the way that most people learned in school.
The old system was rather simple in the textbooks. The basic assumptions were (i) all credit was provided by banks; (ii) all bank credit (assets) were funded by the issuance, or creation, of depository liabilities (money) subject to a reserve requirement; and (iii) central banks controlled credit/money/inflation by rationing bank reserves. A stable 'money multiplier' was hypothesised to allow central banks to accurately predict the eventual impact of changes in bank reserves on money and credit. 
The problem with the old theory of monetary operations is that none of the three assumptions has been true for at least a generation. 
Most credit in the US is created by nonbanks; virtually all bank lending is funded by the creation of liabilities that are not subject to reserve requirements,3 and central banks do not ration reserves. In fact they take great pains to provide banks with the amount of reserves they desire. Central banks influence credit not by rationing the quantity of reserves but by altering the interest rate that banks must pay to obtain the quantity of reserves they desire.
  • Today, credit creation in general and money creation in particular are no longer tied to the stock of reserves (i.e. the stock of banks’ deposits at the Fed).
Today, bank deposits at the Fed have only one real role – to facilitate management of the payments system. They are used to settle transactions among banks. Thus:
  • The old notion that the quantity of bank reserves constrains lending in a fiat money world is completely erroneous.
  • Traditional monetary policy has virtually nothing to do with money.4
....Plainly the stock of reserves is no longer connected to credit or meaningful measures of “money” via the old-notion of a reserve-ratio-based money multiplier. 
I don't buy it all, and I think some of the magic properties of treasuries as collateral and money are a bit overstated. But I'm collecting interesting stories by which it might be the case that current monetary policy has the opposite of the intended sign or other unexpected effects.  Peter certainly offers an interesting example.

He also points out that simply raising interest paid on vast reserves may have different effects than conventional policy which rations reserves. At a minimum he corrects my frequent assertion that reserves and Treasuries are perfect substitutes. No, Treasuries might be more "liquid''!

(Thanks to Thorvald Moe for pointing me to this interesting post.)



  1. It is interesting as a theoretical proposition, but don't the empirics contradict it? Every time QE is announced the stock market surges and remains high until talk of "tapering." That would suggest QE is expansionary in reality, even if it is theoretically ambiguous.

    1. Unfortunately, it seems empirical results as measure of success of models do not much interest many economists. After all, you can always resort to "controlling for all factors" argument.

    2. What you state is very true. But I wonder is it only an artifact of pure herd mentality? If there is no real substance behind it we could have another crash and be left with simply a lot more debt.

  2. John,

    Two points. First, most of the collateral shortage Stella alludes to was caused by one, the crisis destroying private label safe assets and two, the subsequent spike in demand for safe assets. The Fed's contribution is minor. As you know, it now holds roughly the same share of marketable treasuries as it did before the crisis. So at worst, its contractionary effect is small.

    Second, this analysis is a static accounting exercise. It does not consider the counterfactual or the dynamic effect of LSAPs. On the former point, what would have happened to demand for treasuries had the Fed not done QE2 and QE3? I contend that the economy would have been worse off and the demand for safe assets would have increased even more, further exacerbating the existing collateral shortage. On the latter point, I would note that to the extent the QEs have spurred asset price growth and moderate economic gains, they have also supported increased financial intermediation which should lead to more private safe asset creation. The only sustainable way to a full recovery is increased private safe asset creation, not more public safe asset creation. QE, arguably, has done just that on the margin.

    I wrote about this point earlier here:

  3. Its true that reserves are mainly held by banks while most of the economy transacts in broader money components such as commercial bank deposits. Thats why targeting interest on reserves is innefective and the fed should target the interest rate or quantity of broader money in order to attain its monetary targets more effectively.

  4. While short term treasuries may be "more liquid" than bank reserves with the Fed, it seems to me that treasuries carry an asset risk that reserves don't. If short term interest rates change, the portfolio value of existing treasuries must be affected. Thus, while treasuries may have some transaction value that is better than reserves, reserves have better asset value traits than treasuries, esp. close to the zero-interest boundary where interest rates can only rise.

  5. Treasuries more liquid than reserves? I just don't see how this can be. If true, this sounds like a regulatory problem. What prevents someone who needs a bunch of liquidity from contracting with a member bank to hold 100% of their massive deposit in reserves at the Fed? Help me understand why such an arrangement wouldn't dominate treasures. In fact, with excess reserves earning 25 bps a bank could afford to pay more than the 9 bps that treasuries yield! What institutional detail am I missing?

    1. It could never dominate treasuries because contracting is a huge pain operationally and from back office point of view, whereas treasuries you can just buy on the open market in 5 minutes

  6. In case you missed this:

    Federal Reserve Bank of New York, Staff Report, Repo and Securities Lending, Revised
    February 2013

    "Given the essential role of these markets to the functioning and efficiency of the financial system, it is important to better understand and monitor repo and sec lending.
    • The key question addressed in this paper is, what are the data requirements for monitoring repo and sec lending markets so as to inform policymakers and researchers about firm-level and systemic risk?
    • One conclusion emerging from the paper is the need to better understand the institutional arrangements in these markets.
    • To that end, we find that existing data sources are incomplete. More comprehensive data collection would both deepen our understanding of the repo and sec lending markets and facilitate monitoring firm-level and systemic risk in these markets.
    • Specifically, we argue that, at a minimum,six shared characteristics of repo and sec lending trades would need to be collected at the firm level: 1) principal amount, 2) interest rate (or lending fee for certain securities loan transactions), 3) collateral type, 4) haircut, 5) tenor, and 6) counterparty.
    • In addition, we believe there would be value in collecting data at the firm level on the instruments in which securities lending cash collateral is invested. The reinvestment of cash collateral as practiced by securities lending agents potentially introduces a source of risk in addition to the “run” risk that also exists in repo markets. "

    Let's read this part again, "the existing data sources are incomplete". I take them at their word that they don't know the risk.

    See also:

    FT article of April 23, 2013, The Misuse of Collateral Can Help Create Systemic Risk by Satyajit Das

    These and a few other articles on the topic are linked in my non scholarly rant here:
    The Fed, collateral, and repo: more systemic risk

    which also has a neat, but unsettling graph of the Japanese repo failure rate from the FT. No pun intended.

  7. Isn't the fed buying MBS´s rather than Treasuries in this last QE round, with much the same affect as when it bought treasuries?

  8. Reasons I am against QE:

    1. It is hyper-inflationary.
    2. It is inert, does nothing.
    3. It does not help the broader economy, but does cause bubbles, EMT notwithstanding.
    4. The Fed ends up with a big balance sheet. Why that is harmful, I do not know, but I know it is bad.
    5. QE causes"unquantifiable financial risks."
    6. QE is actually contractionary (although I guess that means we can fight inflation through QE in the future, pay down the national debt, and make sure banks have all the reserves they want, and make a slight profit on their reserves. No bank failures!).

    The many faces of QE: causing bubbles and contracting the economy simultaneously, and setting off hyper-inflations, while inert. You gotta love it.

    I like that the Fed can pay down the national debt through QE...and we fight inflation at the same time. Now that is voodoo-nomics, especially if it works.

  9. Comment sent by email, author said it's ok to post:

    Although I’ve seen lots commentary (rarely evidence) on the usual suspects for QE to cause more harm than good, one thing I’ve always thought was clear is that the Fed’s super strong responses in the last two recessions could CAUSE jobless recoveries, as they are massively subsidizing corporate credit. If wages can’t adjust down quickly but capital costs do, this could easily encourage companies to engage in capital intensive investment to the detriment of labor. I think of home depot replacing check-out workers with check-out computers. There is also long history of heavy tech spending in recessions, as firms replace workers with new technology. Some might say this is what happened in the great depression, with firms widely adopting electricity, mass production, etc., replacing labor with machines. Anecdotally, I heard lots of grumbles about this as heavy US manufacturers like Caterpiller issued bonds whose proceeds were purportedly used to cover the cost of building large manufacturing facilities abroad. Anyways, hadn’t seen much done on this but it seems a plausible way that the fed could be shooting itself in the foot.

    JC: There is something really interesting in here. Keynesians like investment because they like "spending." But what if investment all goes to capital/labor substitution? OK, you might say if it's on the margin, but what if we're distorting the margin to inefficiently substitute capital for labor? Low interest rates is just one distortion. Machines don't want health care. Efforts to spur investment in a recession could also be counterproductive. There is a whole new post here. Heck there is a great academic article here.

    1. The "automation causes unemployment" argument is deeply flawed, in this or most contexts.

      If Home Depot hires fewer people to effect cost savings, we can assume those savings are passed along to consumers (in a competitive market), who now spend more elsewhere, thus increasing hiring elsewhere.

      As Milt Friedman once said, we could have people dig trenches with spoons and not shovels, but is that really a plan for increased employment?

      I realize John Cochrane protected himself with the words "inefficient substitute of capital for labor." But "inefficient" would imply that the capital spent did not result in savings compared to the labor avoided. In that case, Home Depot would not use the automated check outs.

      Automation always increases living standards and real output (relatively). It does not hurt employment (in the macroeconomic sense).

      I never thought I would hear people criticizing the low cost of capital for business investment and expansion. In fact, I would guess low cost capital is a boon for employment...I think that is covered in Econ 1 or somewhere....

    2. "The "automation causes un-employment" argument is deeply flawed, in this or most contexts."

      I think you are overlooking the fact that nobody here has made the argument that the detriment to labor as a result of automation is permanent. Your reply seems to suggest that those Home Depot check out employees are going to run across the street and get a new job without any additional training, etc. The comment, as I understood it, refers to "during a recession". Hopefully, that's not a permanent condition.

      It may interest you that Posner and Becker had an interesting exchange on this a short time back. Here's Posner:

      “If technological advance is very rapid, causing in turn a large and very rapid drop in demand in a large labor market, the economy may not be able to absorb the sudden surplus of labor in a short period of time. The result will be soaring unemployment that will retard normal market processes by reducing incomes and in turn production and therefore in the demand for workers.”

      “But there is nothing inevitable about the virtuous process whereby automation and related negative effects on particular jobs merely shift workers to other jobs that are equally or more desirable. Workers may be highly compensated for possessing human capital that is specialized to a labor market that is shrinking. We are seeing this in law today. Because of automation, outsourcing, and more efficient management practices, the demand for lawyers is down, forcing many lawyers either to drop to lower rungs in the profession’s ladder or to leave the profession entirely for work in types of job which their human capital specialized to the practice of law has less, or maybe no, value to employers.
      There is nothing new about this phenomenon, and in the long run labor markets adjust…”

      Nota bene. “In the long run”.

      Maybe that was in Econ 2?

  10. My own analysis of Stella's article:

    I also don't buy it. There's a lot of sense in it, but I think the implied conclusion is wrong.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.