David Beckworth and I had an email exchange we thought worth sharing. (We did some light editing and I added some links.) David is publishing this simultaneously on his blog Macro and Other Market Musings under the title “Further Ossification of the Zero Lower Bound.”

David: I have been meaning to contact you.  I am about to start blogging again and one of the issues I want to raise is the implication of the Fed introducing its fixed rate, full allotment, reverse repo.  I understand the point that it will help with the collateral shortage problem in repo markets—though I think this point ignores the counterfactual of what would happened to repo markets had their been no QE—but I also see it as further cementing the ZLB.  It, along with the IOR, will effectively create a floor on short-term interest rates given the Fed’s preferences to shore up money markets.  While this may be a nice short-term palliative, I see it as creating problems in future recessions if Fed preferences for money markets do not change.  This particularly seems problematic in light of what you have written about creating negative interest rates with e-money.  Here is a twitter exchange with the folks at FTAlphaville on this issue. 

Does this critique make sense? If I am missing something here please let me know.

Miles: Thanks for your kind tweet today!

I confess I don’t see the issue. If the Fed has a reverse repo rate, the Fed can lower that too, when needed. The key is that when the repo rate is introduced, it should be made clear that it might potentially go negative, and have the machinery for that in place.   

In seminars on electronic money, I say that the Fed should move 4 interest rates in tandem:

1. fed funds rate
2. discount rate
3. interest rate on reserves (no longer on just excess reserves when it is negative) 
4. paper currency interest rate

This would add a 5th:

5. reverse repo rate

I also emphasize in seminars that lowering 1–3 while keeping 4 the same leads to unnatural spreads, and these unnatural spreads could, in fact, be worrisome for financial markets. On the other hand, I wouldn’t want to discourage lowering 1–3, since I think it is a big step towards lowering 1–4.  (Politically, the negative rates are the biggest deal, and the technical stuff to let the paper currency interest rate decline is more manageable politically once one has already taken the political hit from 1–3 going negative.)  If all four interest rates–or now all five–are lowered in tandem, spreads are normal, and only nominal illusion and troubles lowering nominal wages are of concern.  

In sum, if a reverse repo rate is introduced AND it is explicitly said at the time that it could go negative, that seems to me a good thing rather than a bad thing. But even if it is not explicitly said that it could go negative, I don’t see why that would be hard in the context of making 1–3 go negative, say. It is not like Japanese postal saving giving zero interest rates, where another institution is involved; all the rates are set by the Fed and it would only seem natural to lower the reverse repo rate if the others are lowered.   

Am I missing something?  

One side note: I confess that in a blog post about this, I would value highly a very detailed explanation of how the repo markets work, and what was going on with the negative interest rates in the repo market that you talked about here. What would be ideal would be an explanation understandable by a bright undergraduate with no experience in financial markets. Then maybe I can understand it, too! 

In that context, the main thing to say is that the Fed should not be using the reverse repo rate to raise interest rates, when it is trying to keep other interest rates as low as possible. If the current repo rate is negative, the Fed should not raise that rate to zero. If the strains caused by abnormal spreads tempt them to do so, that is actually an argument for lowering the paper currency interest rate instead. Note the very important point that a modest lowering of the paper currency interest rate operative for financial firms can be achieved by a modest time-varying paper currency deposit charge when banks want to deposit paper currency with the Fed, without any other institutional change. (Because of the substantial transactions costs involved in taking someone to court, legal tender issues shouldn’t arise until paper currencies are several hundred basis points below zero, and in any case are manageable. Adjustments in how vault cash applies to reserve requirements are easy to make, if needed.) It would be great if this could be done in a somewhat quiet, technical-seeming way at this point. With that in place, it would be clearer that the interest rate on reserves (IOR) could be lowered.     

David: Thanks Miles. No, I agree that introducing the IOR and the reverse repo in principle does not prevent the Fed from making rates go negative. (It may even make it easier to push rates negative since it has now has greater influence over more markets.)  My concern is that the reason—at least what I have seen—for doing IOR and now the reverse repo is to help the interbank and repo markets.  And in practice that means keeping these rates in these markets positive. For example, the reverse repo seems geared toward helping the collateral-starved repo market by reselling to it the much desired treasuries.  But this would keep the repo rate up, or least keep it from falling.  

So it is not the technical capability of the IOR and reverse repo that concerns me, but how it is being used.  If this pattern continues, it seems the Fed will be further cementing the ZLB.

Miles: I think this is the unnatural spreads problem. As long as the paper currency interest rate is kept at zero, I think there is a genuine tension between wanting to keep normal spreads and wanting to keep interest rates down. Of course, QE is making other spreads unnaturally small. So it is a matter of being concerned about certain spreads. My image is not of cementing the zero lower bound, but adding a cushion to the zero lower bound to keep spreads from getting too thin. To the extent it is felt that that cushion is necessary to keep certain spreads from getting too thin, it makes the zero lower bound tighter than it otherwise would be. But the tension is all coming from the fixed paper currency interest rate. 

I feel there can be legitimate disagreements about how much to worry about making certain spreads that are critical for certain financial institutions extra-thin and wanting to keep interest rates low. But what there should be no legitimate disagreement about (but of course, in fact there is plenty of disagreement about) is about the virtues of lowering all interest rates in tandem, including the paper currency interest rate. My UK friends Tomas Hirst and Frances Coppola have not fully come around to my view that with normal spreads, financial institutions will be fine even with negative interest rates. (See the links on their names, and my many discussions with them chronicled on my electronic money sub-blog.) But I think that is basically right, at least given a little advance warning of negative rates in order to adjust the nominal illusion bits of their business models. The strains on financial institutions from lowering all interest rates but the paper currency interest rate should not be extrapolated to a situation where all interest rates are lowered, including the paper currency interest rate. 

David: Great point. If all interest rates are lowered in tandem so that spreads are relatively stable, then negative interest rates should not be a problem for financial firms.  Their net interest margins will be stable too and financial intermediation should continue.  That is a neat way to frame this issue.