Exodus: a few thoughts

If market consensus is correct, today's FOMC minutes should start clarifying the committee's views on its exit strategy and the shape of monetary policy moving forwards.  As discerning students of the financial plumbing are no doubt already aware, a solid roadmap to the new policy regime was  published in February under the auspices of the Peterson Institute.

Although the PI is obviously not an official organ of policy, that Brian Sack put his name to the above-linked piece gives it an imprimatur of legitimacy that could only be exceeded by an Eccles building dateline.   Simply put, here are some of the issues:

* The current size of bank reserves in the system make it problematic to drain them

* The current size of the Treasury and MBS holdings at the Federal Reserve make it problematic to sell them, especially when the issue of market impact is taken into account

*  Non-bank participants in the Federal funds market (i.e., the GSEs) have been heavy lenders into that market, creating a wedge between the effective funds rate and the ostensible target

* Banks are unable to arbitrage the difference (i.e., borrowing in the funds market and lending back to the Fed via IOER) because of the regulatory and financial (FDIC levy) of doing so

* Therefore, Fed funds can no longer serve as the policy target rate

* The Fed has already tested a reverse repo facility (in which it lends securities to the market and takes in cash, paying a fixed interest rate.)   This is likely to be the new policy rate, conducted on a full-allotment basis (unlike the current tests.)  The PI paper suggests that the RRP rate would be set at the same level as IOER

* Because the GSEs and money market funds can participate in the RRP program, the RRP/IOER rate would effectively floor most short term interest rates

* The Fed will also offer a term deposit facility to banking institutions, offering a higher interest for longer-than-overnight money.  The most recent test was a 7 day tender at 0.26%

So essentially, RRP becomes both the policy rate and floor, and Fed funds likely withers away due to lack of transactions.  There are a few questions we need to ask ourselves:

Can any rates trade below the RRP rate?   Sure.  Pension funds, bond managers, hedge funds, and dentists are all denied access to the RRP facility, and thus will buy securities like T Bills under the target rate.   This already happens- the current 3 month T Bill offers a princely 2.5 bp yield.

What about bank commercial paper?  Well, that's an interesting question.   CP has a duration, unlike overnight money, so if, say, 90 day paper discounts an expected move lower in the RRP rate, then it can trade below the spot RRP rate.   We say this in the early Noughties vis-a-vis the Fed funds target rate:

Of course, if the CP market exactly discounts an expected move lower in the RRP rate, money market funds (the target purchasers) would be completely indifferent between the RRP facility and the CP from a rate perspective while exhibiting a 100% preference for the RRP from a credit and liquidity perspective.   The conclusion, therefore, is that bank CP should always require a yield premium to the expected path of the policy rate based on these credit and liquidity factors.

What does that say about LIBOR?  The chart above shows the CP rate for top-tier, A1/P1 credits.  Sadly, the current LIBOR panel contains some names that are less than pristine, including 3 French banks, 3 Japanese banks, a nationalized UK bank, etc.  LIBOR is currently setting 7 bps above the top-tier CP rate denoted above.   However, even that is low by the broad standards of history.   As you can see, in times of normalcy LIBOR tended to trade between 10 and 20 bps over CP.   The LIE-BOR phenomenon is clearly visible on the chart; readers can judge for themselves whether the current low spread is symptomatic more of the absolute low level of yields or the authorities' greater interest in levying fines than fixing the system properly.

To be sure, spot LIBOR can conceivably trade below the spot RRP rate, for the same reasons that CP can.   However, LIBOR will obviously be impacted by the same credit and liquidity risk premia that affect CP, only more so (hence the traditional spread between LIBOR and CP.)  As a rough target, Macro Man assumes that 90 day CP will trade something like 10 bps over the RRP rate whilst the latter is on hold.  This is more or less where it traded in the first half of 2004, when the Fed was slowly gearing up for the previous rate hike cycle.   Being charitable, we'll assume that the current LIBOR/CP spread is the new normal, and tack on another 7 bps.   That would put LIBOR 17 bps above the RRP rate, or at 0.42% if and when the RRP-as-policy is implemented at a rate of 0.25%.   (There is in fact a school of thought that it will be rolled out at a lower rate to give markets time to adjust.)

Of course, Macro Man would submit that there will be a tremendous asymmetry to these calculations, insofar as there are really no plausible circumstances where bank CP trades below the expected RRP path, and plenty where the liquidity preference for the RRP over CP widens quite a bit.

Comments and feedback from informed readers on this section are particularly welcome.   Also, if anyone has a time series of bank CP rates of the dollar LIBOR panel that they'd be willing to share, Macro Man would be particularly grateful.   He's denied access in the Bloomberg account he's currently using.

What about OIS?  Another good question.   If the Fed funds market withers, it kind of leaves the OIS market up the creek sans paddle, because there won't be a mechanism to trade the RRP rate interbank.  The PI paper above acknowledges a potential problem but suggests that the RRP transition is made early enough to allow existing contracts to be replaced by ones referencing RRP.  All well and good, but what if you have a a 30 year swap on the books that uses OIS discounting, and you don't plan on taking it off for ten years?  Will there be a mandated switch?  There doesn't appear to be an obvious and easy answer here, and a potential threat from the law of unintended consequences looms large.

What about the Fed's securities portfolio?   When will they start selling?   There are two hopes that the Fed ever sells anything currently on its books:  Bob Hope and no hope.