Tuesday, January 15, 2013

More on Floor Systems

The discussion of floor systems, between Steve Randy Waldman and Paul Krugman has expanded. In line with my last post, Waldman gets it, more or less, with a little confusion. Krugman is confused.

What adds to the confusion, in part, is Krugman's example:
...what happens if the US government issues a trillion-dollar coin to pay its bills?
I discussed that in this post. Krugman thinks it makes a difference whether the interest rate on reserves is close to zero, as it is now (at 0.25%), or say 5%. If there is a significant quantity of excess reserves, actually there is no difference.

Suppose, as in Krugman's example, the Treasury issues currency to pay its bills. Where the currency is deposited - at private financial institutions or the Fed - or even if the currency issued by the Treasury circulates, doesn't make a difference. What matters is the consolidated balance sheet of the Fed and the Treasury, and so there is no difference between what happens when the Treasury issues currency to finance its deficit, and what happens if the Treasury issues debt, of any maturity, and the Fed purchases the debt with reserves. It's neutral - a liquidity trap - whether the interest rate on reserves is 0%, 0.25%, or 5%. For example, if the Treasury issues currency, that will not ultimately affect the quantity of currency in circulation, with the extra outside money ultimately increasing the quantity of reserves one-for-one.

Where Waldman is confused, in part, is here:
Further, a floor system is very attractive to central bankers. It maximizes policy flexibility...
Actually, a floor system gives a central bank no more or less flexibility than does a channel system. For the Fed, the only difference is that they control the overnight policy rate in a different way.

12 comments:

  1. I have a somewhat different take. I wrote this on Waldman's blog:


    I don't understand what Krugman is saying in his new piece. Let's say they can't redeem the coin because congress won't authorize the borrowing. Then then market interest rates can't just rise off the floor. Why would anyone *ever* borrow at a high interbank rate when they can just settle by actually transferring low IOR reserves? Or think about it this way: if a bank has too little excess reserves to make a payment, every other bank will be scrambling to *lend* them the money at a slight premium above IOR since those banks will otherwise be earning IOR on those reserves. The interbank rate simply *cannot* rise significantly above IOR with anything remotely resembling the current quantity of reserves. The basic framing of his argument is just wrong.

    The fact is, having a big coin on balance sheet could be a real problem. Lets say inflation starts to rise and the size of the coin exceeds the sum of public currency demand and required reserves by some amount x. This means that the Fed is powerless to raise the target-rate-to-IOR spread above whatever level is consistent with banks wanting to hold a quantity of excess reserves in the amount of x. If x is a big number (a few hundred billion) then the target rate (the interbank market rate) will be stuck at IOR.

    So if they want to raise the interbank rate they *have* to raise IOR. The problem with *that* is that the Fed's assets (the coin) don't earn interest, so they *can't* pay IOR without being in the red. I don't know the Fed's budgeting rules, but I'd be surprised if they are authorized to operate in insolvency. Don't they have to make up deficient equity from the treasury? And doesn't that need to be approved by congress? Sounds to me like they would be stuck, unable to hike rates without congressional approval. And in case of inflation that could lead to hyperinflation for which Obama could easily get the blame (he made the coin). So congress would have every incentive *not* to fund the Fed. Which means a full blown hyperinflation.

    Even a small probability of exiting the liquidity trap in the next couple of years resulting in hyperinflation would immediately raise inflation expectations thus possibly setting off the self-fulfilling hyperinflationary chain of events. The only way the white house would ultimately be able to limit the damage would be via some kind of MMT-style fiscal austerity inflation control. I guess it's all possible but seems like a suboptimal path.

    ReplyDelete
  2. Dr. Williamson,

    This is essentially a re-post of my comment from Waldman's blog.

    As with SRW, I am on board with you.

    But I think this logic can be taken one step further, which is to say, even *prior* to IOR (i.e., the channel system), we were in “liquidity trap” conditions, as it is being discussed in the context of this debate. This is because the CB cannot do proactive OMOs (i.e., increasing or decreasing the monetary base independent of defending the target federal funds rate) without driving the federal funds rate to 0 or the ceiling rate. To illustrate what I am trying to say, prior to IOR, if the Fed were to do QE, the FFR would have plummeted to 0, given all the excess reserves, bringing us to classic Krugman liquidity trap conditions. And likewise, we couldn't have minted and spent the coin prior to IOR without driving the FFR to 0, given all the excess reserves it would have created. The only way to support a positive FFR above 0 after these types of policies is to pay IOR or drain those excess reserves back out of the system. Either way, we are in "liquidity trap" conditions as you and Waldman point out. So, in other words, it doesn’t appear you can ever escape liquidity trap conditions.

    Is this also what you meant when you wrote on Waldman's blog that "Actually, I think it makes essentially no difference to how policy work"?

    ReplyDelete
    Replies
    1. And to clarify, I realize some people may include the lack of effectiveness of monetary policy due to the zero lower bound when defining liquidity trap conditions, but I am putting that aside. I am defining it context of this current debate, where we are just considering whether printing money vs. also selling bonds has a differential impact on inflation (assuming the Fed still targets the interest rate where they want it).

      Delete
  3. I think the confusion stems from the fact that before the IROR a liquidity trap was only possible at a zero interest rate. The IROR has changed that, but Krugman still thinks in pre-IROR terms.

    ReplyDelete
    Replies
    1. I get the feeling in this discussion that the term 'liquidity trap' is being abused a little bit.

      We seem to be using it as a term for a situation in which standard open-market operations by the Fed have no impact, because bank reserves and short-term debt are interchangeable. That definition has some meaning in a regime in which buying and selling short-term debt is the primary mechanism by which the Fed conducts monetary policy. If we are now in a regime where IROR is their primary tool then that is not a very useful definition of a liquidity trap -- rather we should be defining the term to mean a situation in which the Fed's preferred IROR would be negative.

      Delete
    2. And rereading Krugman's posts on the subject, he may be using 'liquidity trap' in that way already -- i.e. a situation where standard Fed policy is impotent, rather than the narrower meaning of a situation in which bank reserves and short-term debt are equivalent.

      Delete
    3. You can define your terms any way you want. It's useful to say "it's a liquidity trap," as people can then key into what's going on, and understand it. I don't much like your definition. Why do I care about what the Fed prefers? Why do I attach some significance to the fact that people at the Fed might prefer something that is infeasible?

      Delete
  4. the treasury cannot issue currency. the fed cannot issue currrency unless it is to buy assets, such as debt. the debt the treasury issues must be for cash from someone. doesn't this matter?

    ReplyDelete
    Replies
    1. "...the treasury cannot issue currency"

      That's what I thought, until recently. That's what #mintthecoin is all about. There's a loophole that allows the Treasury to issue coins directly to the public, with apparently no restrictions on denominations. So, the Treasury can in fact issue currency, but likely they would never do it, so the example is just hypothetical, but nevertheless instructive (in spite of Krugman's confusion).

      Delete
    2. On this issue, what do you think of the recent headbutt between Krugman and John Steward? :)

      http://www.huffingtonpost.com/2013/01/15/jon-stewart-smacks-paul-krugman_n_2478680.html

      Delete
    3. Well, it's a bit much for Krugman to be giving Jon Stewart a hard time, when he doesn't understand all the subtleties himself. To call Stewart lazy and unprofessional seems unfair.

      Delete
  5. Where are the details of law about who can actually print dollar bill currency? I understood from the Fed website that Treasury prints the dollars.

    Who has the legal authority to decide to increase the number of printed dollars?

    ReplyDelete