Thursday, May 15, 2014

Interest on Reserves in 2008 and now

Interest on Reserves in 2008 - Highly Contractionary

I was reading this paper from Peter Ireland about Interest on Reserves (IOR).  Reading that paper, and thinking through the effects we should expect from IOR, I am thinking that the Fed Funds Rate (FFR) is still the overwhelming factor with regard to monetary policy.  If I have his argument correct, given the current level of reserves, a rising FFR along with a rising IOR rate should basically have the same effect as a rising FFR in an environment where there weren't excess reserves.  If there weren't excess reserves, I imagine that IOR could move around below FFR without much effect on the money supply.  So, it appears that they can use IOR to manage the reduction in the Fed balance sheet.  But, it seems to me that the level of FFR relative to natural interest rates would work basically the same way that it has without IOR.

This leaves the question, however, of whether the implementation of IOR in 2008 was that contractionary to begin with.  Generally, IOR is considered to be a floor for FFR because if IOR was higher than FFR, banks could borrow at the low FFR and hold it as reserves, pocketing the difference in rates.  They would bid the FFR up to a level near the IOR rate, in that case.

In October 2008, the Fed implemented IOR, and over the course of a month, ratcheted up the IOR rate until it was equal to the FFR target.  But, throughout this period, the effective federal funds rate was volatile, and tended to run below the FFR target.  So, from November 5, when the Fed pegged the IOR rate to the FFR, until December 16, when the FFR (along with IOR) was finally reduced to 0.25%, the effective FFR was running 0.5% to 0.75% below the IOR rate.  During this time, the Fed had accumulated a substantial amount of non-traditional assets, but it was not purchasing treasuries, and in fact would not start adding to securities held outright until March 2009.

There were so many things going on at the time with the Fed balance sheet, and I am no expert on the micro structure of trades between the Fed and commercial banks.  But, something in November and December 2008 was pushing the effective FFR well below the IOR rate while excess reserves ballooned.  It looks to me like the Fed should have pushed the FFR to 0% in October with no IOR, since the market risk free rate was in free fall, and the economy was desperate for cash.  Instead, the Fed sucked all the panicked cash out of the economy with IOR above the market risk free rate, to the tune of more than half a trillion dollars.

IOR and Excess Reserves at the End of QE

With regard to the amount of excess reserves currently outstanding, I wonder if there is a point where rising short term rates with stable IOR rates would actually be expansionary.  Further, if reserves aren't the only constraint on bank lending, then I think short term rates could rise even while excess reserves remain in the system.  With FFR, IOR, and the level of securities on the Fed balance sheet, there are several moving parts to consider now, but I think, regardless of Fed policy in the very short term, natural short term interest rates might behave fairly typically as the economy continues to recover.  It looks like the Fed is planning on raising IOR and FFR together. But, normally a rising FFR rate resulting from the Fed selling securities would be contractionary.  In the current environment, if the Fed raised FFR but held IOR constant, some of the sterile excess reserves at the banks would be injected into the economy, so that a rising FFR resulting from some Fed selling could still be expansionary.  If this is the case, then the liquidity effect of Fed open market operations would have a very different shape in the context of high excess reserves.

I'm not a banking expert, but it seems like if capital constraints on the banks led to an increase in short term interest rates while reserves remained high, deposit interest rates would remain low, incentivizing depositors to purchase securitized assets from the banks or loan funds outside the banks until the interest rate, new bank assets, and reserve levels reached a new equilibrium.  Peek and Rosengren at the Boston Fed show that for capital constrained banks, loans can expand in response to monetary tightening, or at least that capital constrained banks will not markedly change their credit activity in response to monetary policy.  Is it possible that velocity would be self-correcting as excess reserves decline?

It seems unlikely for the expansionary counter-effect to be greater than the original contractionary effect of reduced reserves.  But, if I imagine an extreme world with $15 trillion in excess reserves with only $1 trillion in treasuries left in private hands, it seems clear to me that a rise in interest rates due to Fed OMO could happen with a large amount of reserves still in place, leading to a strong increase in credit and velocity.  So, at some quantity of excess reserves, a reduction in the Fed's asset base must be expansionary.

I had been wondering if rates would rise slowly as we leave the zero lower bound, but now I am starting to wonder if this context, market rates could rise, which would require the Fed to either overshoot the FFR or raise IOR rates along with FFR in order to counter these odd expansionary effects.  If that is the case, there shouldn't be a drag on the rise in short term rates - at worst they would rise at a pace typical of past recoveries.

Please let me know in the comments if I'm completely out of my element here, but be gentle with me.

No comments:

Post a Comment