Turbulent Exit?

In a followup from their much-discussed (by me here) May memo, Pozsar and Sweeney predict “A Turbulent Exit” when the Fed begins to raise rates.  FT Alphaville and Bloomberg both appreciate the importance of the memo, but focus attention on the exchange rate dimension, and so miss the main point.  Let’s walk through the argument more slowly, so as to put the exchange rate consequences in proper perspective.

As in their May memo, P&S point to an imminent and very large rotation of existing cash pools away from prime money funds and various bank liabilities (domestic and foreign) into government-only money funds.  Here is the picture:  Exhibit 14.

There are two reasons for this rotation.  One is the shift to floating NAV for prime funds, but that is not imminent and so the resulting flow should be “slow”.

The other is the Fed’s shift to higher base interest rates, and that apparently is imminent (whether September meeting or the next one), and so the resulting flow should be “fast”.  Fast flows are likely to be disorderly.  In the words of P&S, “someone won’t get things right”.  Hence, turbulent exit.

The reason for the fast shift is that, given regulatory changes, banks really don’t want to be in the business of meeting demand for cash balances from the large institutional cash pools.  So probably they will not match the Fed’s interest rate hike on liabilities they offer to the cash pools, providing cash pools with a clear incentive to move their money elsewhere, pronto.  Government-only money funds are the likely destination because they are constant NAV.  The question is what government assets are out there for the funds to buy?

The answer, according to P&S, is the RRP facility of the Fed.  And the best way to reduce turbulence to a minimum, they suggest, is for the Fed to offer a full-allotment RRP facility at the new higher interest rate.  Suppose, for the sake of argument, that it does so.

As cash pools withdraw from bank deposits, U.S. banks shrink their balance sheets on both sides, shedding excess reserves.  As money flows into government-only money funds, that money gets invested in overnight RRP at the Fed.  On the Fed’s balance sheet, the effect is simply to shift the liability structure from reserves (held by banks) to RRP (held by money funds).

Similarly, and more slowly, as cash pools withdraw from prime money funds, these funds liquidate their holding of foreign bank liabilities (repo, CD, CP), causing foreign banks to shrink their balance sheets on both sides, also shedding excess reserves.  As the cash pool balances get shifted into government-only money funds, these reserves also get shifted into RRP balances at the Fed.

A key reason for turbulence in this process is that while some banks have excess reserves more than adequate to meet the cash pool draw down, other banks do not.  Interbank markets presumably could make up for this, as surplus banks lend to deficit banks, but these markets are a bit rusty these days from disuse.  Hence expect turbulence as they get up and running again.

That’s the main story.  (One reason that FTAlphaville and Bloomberg miss it, possibly, is that the key Exhibit 14 unfortunately shows shift of cash pools from bank liabilities to money fund liabilities, but does not show the commensurate shift of reserves, neither on the balance sheet of banks nor on the balance sheet of the Fed.  Pedagogy matters, my friends!)

The real question is whether the Fed will in fact offer a full allotment facility.  P&S suggest that they will be forced to do so, and it is the Fed’s commitment to controlling base money rates that will force them.  The RRP facility (for non-banks) is supposed to be a floor on rates, while the IOER facility (for banks) is supposed to be a ceiling.  Floors and ceilings don’t work unless you have elastic supply at those rates.  That’s what full allotment would be.

Supposing full allotment, what happens to rates?  P&S basically say that the existing rates structure will simply shift up pari passu with the base rates.   RRP currently 5 b.p. shifts to 25 b.p., while IOER currently 25 bp shifts to 50, and everything else (tri-party repo, fed funds effective, LIBOR) shifts with it.

They say nothing explicit about exchanges rates, but they do suggest that “A full allotment RRP facility could accelerate FX reserve managers’ shift out of euros for U.S. dollars”, which I suppose implicitly suggests appreciation of the dollar.  That’s the phrase that has captured attention so far.  But I wonder about that.

It seems clear to me that the Fed does not want to be central banker to the world.  Also clear, the Fed is willing (more or less necessarily) to serve as such at one remove.  Foreign banks issue dollar liabilities and make dollar loans, so they need dollar reserves, but those reserves should be accounts at domestic banks, not at the Fed.  Foreign central banks manage the balance of payments for their individual nations, and hold some dollar reserves for that, but those reserves should largely be Treasuries, not deposits at the Fed.  The Fed is prepared to be lender of last resort to both of these channels, but at prices away from the market, outside spread not inside spread.

From this point of view the important thing is not the Fed’s Foreign Repo Pool, which P&S note took an upward jump after the ECB cut deposit rates below zero in June 2014 (Exhibit 19), but rather the unlimited swap lines with the major central banks, which I have discussed in a previous post.




5 comments on “Turbulent Exit?

  1. Walker Todd on said:

    Excellent analysis by Mr. Mehrling. In fact, it is the best succinct explanation of what is likely to happen as rates shift upward (and of what is happening now at the current rate level) that I have read.
    Someone higher up the food chain should pay attention both to what it means and what the implications for future developments are when Mr. Mehrling writes (accurately, in my view) that the Fed is willing to act as central banker to the world “at one remove.” And that “one remove” is the foreign exchange swap lines channel, as he notes. There are institutional structure implications of a gradual shift of the Fed’s balance sheet toward the swap lines and away from domestic claims as the principal assets that the Fed holds. — Walker Todd, Chagrin Falls, Ohio

  2. Your blog posts are always engaging and concise, which I appreciate. It seems like the Fed is largely in agreement with your analysis (or at least their economists are). Much of what you cover is analyzed in this paper: http://www.federalreserve.gov/econresdata/feds/2015/files/2015010pap.pdf

    See the excerpt below for the paper’s analysis of the effect of a full-allotment RRP facility on the Fed’s balance sheet:
    “Importantly, increased ON RRP take-up does not expand the size of the Federal Reserve’s balance sheet or the volume of private short-term funding required to finance that balance sheet. Instead, such an increase shifts the composition of the Federal Reserve’s liabilities from reserves held by banks to RRPs that can be held by a wider range of institutions. Hence, the crowding out of private financing that results from greater use of ON RRP would largely represent a reduction in private lending by money-market investors to banks that are financing reserves, offset by an increase in private lending by those investors directly to the Federal Reserve. Given the IOER and ON RRP rates that are set by the Federal Reserve, the allocation of Federal Reserve liabilities between reserves and RRPs is ultimately determined in markets.”

    I’d love to hear your thoughts on that paper in particular given that it’s an analysis done by Fed economists and for the Board of Governors.

  3. Perry –

    Could you say a few more words on the opinion that the “existing rates structure will simply shift up pari passu with the base rates” regarding the longer term?

    I’m thinking of points JW Mason makes at his blog, excerpts and link below:

    The following changes shows the yields of Treasury bonds of various maturities, and the capital loss for each bond from a one-point rise in yield over the next year. (All values are in percentage points.)

    Maturity Yield as of July 2015 Value Change from 1-Point Rise
    30 year 3.07 -17.1
    20 year 2.77 -13.9
    10 year 2.32 -8.4
    5 year 1.63 -4.6
    1 year 0.30 -0.0

    So if the 30-year rate rises by one point over the next year, someone who just bought a 30-year bond will suffer a 17 percent capital loss.

    When long rates are low, even a modest increase implies very large capital losses for holders of long bonds. Fear of these losses can set a floor on long rates well above prevailing short rates. This, and not the zero lower bound per se, is the “liquidity trap” described in The General Theory.


    And thanks for the incisive commentary; it is a delight to learn from you.

    • Perry Mehrling on said:

      Mason’s post is about the term structure of interest rates. Pozsar and Sweeney (and so also me) are talking about the structure of money rates only–repo, Fed Funds, LIBOR, Tbill. Even if the structure of money rates all moves up by 25 bp, it does not imply that longer rates do as well.

  4. tom michl on said:

    I’m not sure Pozsar and Sweeney are right about the full allotment part. The Fed seems concerned about the unintended consequence that a full allotment facility would create instability in the event of a run on wholesale markets–a flight to quality. They may split the difference and go for some kind of moving or dynamic cap on the ON RRP?

    I’d be interested in your thoughts on the microstructure once the Fed has reduced its balance sheet and reserves return to a more normal level. They do not seem to have any appetite for keeping the ON RRP facility alive. But that is probably a blog for another day.