The minutes for the July 30-31 FOMC meeting might have included a whopper of a new development in U.S. monetary policy – the Fixed-Rate Reverse-Repo Facility. I see the possibility of a significant change in how the Fed conducts monetary policy and a major change in the Repo market. Perhaps I’m reading into it too much – here are my views:
Managing Money Market Interest Rates
The immediate take-away from the announcement is that Reverse-Repo operations are a tool to help drain reserves – take cash out of the system – when the Fed deems it’s time to start moving the fed funds rate higher. As long as the Fed has a large QE portfolio of securities, there’s a lot of extra cash in the system. When the time comes, the Fed needs a little help to drain all that cash. Perhaps they need an unconventional tool … when time is right. But remember, the market is not pricing a rate hike until the end of 2014. Is the Fed worried about inventing new open market tool almost 1 1/4 years in advance? Obviously, everyone knows what I’ll say next – they’re finally concerned with the unintended consequences of QE programs. But I think it’s more than that. Much more.
The Fixed Rate Reverse-Repo Facility is supposed to be like the IOER – Interest On Excess Reserves, where banks can deposit excess reserves (extra cash) at the Fed and the Fed pays them an interest rate of .25%. The IOER effectively means there’s a rate floor for banks – the lowest they can receive is .25% on their cash. Right now, the IOER rate is above market rates (fed funds at .10% and GC Repo at .03%), so it gives banks an incentive to deposit cash at the Fed and not dump it into the market. Keep that in mind for later – cash moving into the Fed.
Since the Fed announced the new facility has a “fixed-rate,” it means the rate is pre-determined. Here are are some possibilities as to what that will be:
- The fed funds target rate – If the Fixed-Rate is the fed funds target rate (when it’s not a range), it will keep GC trading slightly above fed funds each day, figuring that the market will need higher rates to attract cash away from Fed deposits.
- Bottom end of the fed funds target range – When there is a range, like there is now, setting the Repo rate at the low end of 0.0% doesn’t really help or achieve anything. The Repo rate must be set within the range. Possibly mid-range? Perhaps that would mean a rate of .125% – I think this is possible.
- Target Rate Minus X – There could be an official announcement that the Fixed-Rate is always at the Fed funds target rate minus a fixed spread, say 10 basis points. So if the fed funds target rate was .50%, the Fixed-Rate for Repo would be .40% – This is also possible.
- Auction Rate – The Fed could choose to auction collateral each day or have a series of auctions. However, if everyone gets filled (“full-allotment”), having market participants submitting different rates doesn’t make sense, unless everyone is filled at the average. Getting filled at the average is also possible. But then again, it’s not a “fixed” rate.
“Additions To The Range of Counterparties”
There’s already an expanded list of counterparties for Reverse-Repo operations, it’s not just the Primary Dealers anymore. The expanded list includes many banks, money market funds (MMFs), and the GSEs. The announcement says “additions,” which implies even more counterparties. This is big!
Suppose MMFs no longer have to go to the Street to invest their cash in the Repo market? Instead, there’s easy access to collateral through the Fed at a market rate. Well, that’s a lot of cash that’s going to be going to the Fed each day. In fact, much of that overnight cash is currently invested through tri-party Repo. Remember Bill Dudley and Ben Bernanke commenting how tri-party Repo needs to be reformed? That there’s a systemic risk if MMFs have to liquidate Repo collateral in the event of a dealer bankruptcy? The “fire-sale” issue. Remember Fed officials said if the industry doesn’t do something about reforming tri-party then they will? Well, the industry just got slammed by the Fed. The Fed might just have changed the entire Repo market in one fell swoop.
If MMFs have the option to invest cash at the Fed each day, there’s no need for them to do tri-party Repo with the Street. Perhaps the Fed won’t capture the entire tri-party Repo market, but it could be a significant enough percentage to eliminate the systemic risk from tri-party. This could be a very significant implication of the Fixed-Rate Repo Facility.
New Target Rate To Replace Fed Funds?
Now, there’s an interesting theoretical issue that comes up. The Fed currently sets the fed funds target rate, but it doesn’t set a target rate for Repo. Many other central banks use a Repo target rate. Should the Fed be using a Fixed-Rate for Repo when the spread between fed funds and Repo can move from +5 basis points to -5 basis points regularly. In other words, how do you set a Fixed-Rate for Repo when Repo rates move around fed funds constantly? Which then poses the question, why target a rate (the fed funds rate) that the Fed can only indirectly influence in the first place? If the Fed implements a Fixed-Rate Repo, they might as well just target general collateral Repo rates instead of fed funds. The fed funds rate is clearly outdated, certainly the general collateral Repo rate is a more important rate, it’s more liquid, and it has many more market participants. Perhaps the Fed should use the creation of the Fixed-Rate Repo to scrap the use of the fed funds rate. It will be much easier for the Fed to set a Repo rate if they target Repo rates in the first place.
When the Fed say “full-allotment,” it means everyone who submits a request for collateral gets filled. It means everyone is allowed to play the game and no one gets left out. What’s important here is that there’s unlimited collateral available at a certain level, so it puts a floor on GC rates. There are seasonal benefits to a floor in Repo rates, it will smooth out the April collateral shortage and the January mini-collateral shortage. It just might just end quarter-end distortions, when rates can drop below 0%, and of course, it will end the unintended consequences of QE programs – eventual 0.0% Repo rates.
Remember though, when the Fed puts a floor on Repo rates it’s only good while the facility is active during the day. If the Fed’s Fixed-Rate Repo Facility expires at 10:00 am, then it’s back to market supply and demand after that.
Timing of The Facility
The next question is “when” will the Facility be available? It could possibly be continuous – available all day long – but that might create a lot of additional operational work at the Fed. Just for old times sake, I’d like to see them bring back “Fed Time” and enter the market between 11:30 am to 11:35 am each day. However, the entire Repo market operates on an earlier schedule these days, so 11:30 am is too late in the morning. Certainly opening the Facility at one specific time in the day is preferable and traditional. The timing should be the first thing in morning, when MMFs and cash investors are the most active, some time between 8:00 am 9:00 am.
Types of Collateral
Obviously, the Fed is limited by what’s in their SOMA portfolio. For smaller operations, it’s easy for the Fed to pick and choose which Treasurys they use, but say the size is in the hundreds of billions? Would the Fed allocate the collateral to each cash investor? If so, is it all U.S. Treasurys? What if some investors want <2 year paper? What happens when a cash investor is not allowed to hold bond collateral? Remember, they’re not just dealing with Primary Dealers anymore, there are “retail” customers involved. Also, does the Fed also offer Agency and Agency MBS paper? Would different collateral types have slightly higher rates? Of course, they’re in the Fed’s SOMA portfolio so they might as well use them.
Criticism And Other Observations
Let’s say, for arguments sake, the size of he Fed’s SOMA portfolio stood at $400 billion, around the size before the first QE program. Would there be enough available collateral to satisfy all the collateral demand? Suppose the Fed captures a large percentage of the tri-party market, then would a $400 billion portfolio be large enough to satisfy SOMA securities lending for specials and the Fixed-Rate Repo Facility? If the Fixed-Rate Repo Facility is permanent, then there is also a perpetually larger Fed SOMA portfolio.
Now, everyone knows the Fed is a pretty good counterparty. In fact, risk-free securities with a risk-free counterparty is extremely good for capital charges, Basel 3 Risk Weighted Assets, and exemption from the European FTT tax. Not only from a risk perspective but also from a regulatory perspective, it’s preferable to do business with the Fed rather than other market participants. So is it good for the Fed to “compete” with regular market participants? If to better regulate the market, the Fed becomes a market participant, is that a good thing?
Overall, I like the idea of a Fixed-Rate Repo Facility. It puts a floor on overnight Repo rates, potentially solving the QE induced collateral shortage. It potentially solves the tri-party “fire sale” issue, and quite possibly, the Fed will use the new facility to overhaul the outdated fed funds target rate in favor of a modern Repo rate. What’s not to like?