The Federal Reserve Bank of New York made some key changes to the Overnight Reverse Repo Program (RRP) last week following the conclusion of the Federal Open Market Committee meeting last Wednesday, October 29. As part of an "operational readiness exercise," the New York Fed will modify the offering rate for the remainder of the year. For operations conducted Oct. 30-31, the offering rate is 5 basis points; for operations conducted Nov. 3-14, the offering rate is 3 bps; for operations conducted Nov. 17-28, the offering rate is 7 bps; for operations conducted Dec. 1-12, the offering rate is 10 bps; for operations conducted Dec. 15 and after, the offering rate is 5 bps. The NY Fed explained, "The Federal Reserve continues to enhance operational readiness and increase its understanding of the impact of RRPs through technical exercises. In further support of its objectives, the FOMC instructed the Desk to examine how term RRP operations might work as a supplementary tool to help control the federal funds rate, particularly when there are significant and transitory shifts in money market activity. In support of this goal, and to reduce potential volatility in money market rates, the FOMC instructed the Desk to conduct term RRP operations that cross year-end. The operations will mature on or about January 2, 2015." So what does this mean for the money market industry? Strategists Andrew Hollenhorst from Citi Research, Alex Roever from J.P. Morgan, and Joe Abate from Barclays discussed the implications in recent commentary.
Wrote Hollenhorst in his monthly, "Short End Notes," the move surprised the market. "The announcement caused some to speculate that the Fed may be more willing to entertain tighter IOER-RRP spreads than the announced 25bp spread.... We disagree with this interpretation. The Fed is likely most interested in the extent to which other money market rates including fed effective and T-bill rates move higher with the RRP rate, and less interested in the tighter IOER-RRP spread. Since the Fed plans to use RRP as the backstop to prevent fed effective from falling out of the bottom of its target range, it makes sense that it would want to test RRP at higher rates that put more meaningful upward pressure on fed effective. After all, at its recent level of 9bp fed effective is only slightly higher than its lows in 2011. As such we would resist the urge to read into temporarily higher RRP rates a change in long term policy plans."
He adds, "All else equal higher repo rates in November and December should lend support to money market rates. With market repo rates recently trading in a volatile range around 10bp, it is unclear that a move in the RRP rate up to 7bp would cause a significant move higher in T-bill yields or fed effective. At a 10bp RRP rate, however, we would expect fed effective to move at least a few basis points higher. T-bill rates may be relatively less affected as demand for cash securities may keep bills dislocated from broader money market rates. While some had been looking for an increase in the $300 billion aggregate cap on the size of the RRP facility, we think this is very unlikely now that plans for the remainder for 2014 have been announced. However, the Fed will be supplying up to $300 billion in collateral through term repo operations over year-end which should relieve some of the downward pressure on rates and may potentially keep the $300 billion cap on the overnight RRP facility from binding over year-end."
Roever said that while the NY Fed's statement doesn't offer a specific explanation for the shifts, "we suspect the exercise is intended to paint the demand curve for ON RRP, illuminating aggregate demand for ON RRP at differing offer levels." He elaborated, "We suspect the Fed is primarily looking at how non-peak demand is influenced: will usage fall at 3bp and will it rise at 7bp or 10bp? Will the higher yields increase demand to the point that the $300 program cap comes into play and Dutch-auction process results in a clearing yield below the offering rate? Or -- if the program cap doesn't come into play -- will the higher offering rates apply upward pressure on Fed funds, ON GC repo rates or perhaps even very short T-bill yields? For policy makers evaluating ON RRP as a draining tool, this yield sensitivity analysis is essential, although it would probably be more informative if the Fed would test a higher offering rate -- like 15bp or 20bp -- so that it could better gauge the yield lifting power of the ON RRP. Unfortunately, the FOMC doesn't seem ready to take this step and perhaps is concerned about sending an inadvertent tightening signal to the market."
Roever also commented on how the FOMC instructed the Desk to conduct term RRP operations that cross year-end. "While the Operational Statement promises more details in early December, this short statement is telling in a few ways. First -- and perhaps most importantly -- the Fed acknowledges the "significant and transitory shifts" in the money markets and links these "shifts" to the Fed's ability to raise Fed funds. For decades, money markets have played a key role in monetary policy, primarily by participating in repos that helped the Fed guide the Fed funds effective rate near the targeted policy rate. Its ON RRP testing over the past year is affirmation that the Fed sees the money markets playing a similar role in the future, even if that role is now lessened by the presence of IOER. Even so, structural changes in the money markets, most brought about by various financial stability driven regulations, may create unanticipated challenges for monetary policy implementation and the Fed wouldn't mind a bigger tool kit."
He writes, "For now, it's also unclear what the Fed's longer run intentions will be with respect to term RRPs. Possibly they could be offered on a regular basis (perhaps in 14- or 28-day increments) or seasonally (just over quarter-ends, for example). From a market perspective, the difference is that regular offerings would provide a more persistent lift to short-term interest rates by creating more competition for scarce money market assets -- 28-day reverse repo with the Fed is a close substitute for a 4-week T-bill or a term treasury repo with a dealer. In contrast, seasonal offerings would address surges in quarter-end demand, helping to keep short-term rates from plunging at quarter-end and taking pressure off of custodian banks that see cash balances surge at quarter-ends if money fund managers and other market participants can't find enough securities to invest in."
Roever concludes, "At this point, not knowing the details of how the Term RRPs will be offered, or more about the Fed's longer term intentions, we hesitate to interpret this announcement as an expansion of the Fed's use of the money markets to drain reserves. For now, the term RRPs look to be temporary and at best should help to absorb potential spikes in ON RRP demand at year-end. If the term RRPs do become a regular tool like ON RRP, we think that that they could help lift short-term Treasury yields as the Fed attempts to normalize monetary policy. For now, the $300bn cap on the ON RRP has exceeded market demand every day but one. But structural change is coming to the money markets over the next few months, and we suspect the demand for short-term government securities and for Fed RRP will expand."
Barclays' Abate also shared his thoughts on the modifications. "We expect the Fed's term RRP to bias overnight rates higher. The term program will absorb a substantial portion of the cash that would ordinarily pile up in overnight repo. Indeed, while we expect overnight repo to average 12bp in December, the rate could reach as high as 15bp in early December. Separately, the spread between the overnight market repo rate and the term RRP rate will measure the strength of money fund (and other cash long investor) demand to "invest early and forget it" -- that is, their year-end premium. Looking beyond year-end, the outlook for repo rates will depend on two factors: the parameters of the RRP program and the supply of bills."
Abate concludes, "The bigger uncertainty in the outlook is what the Fed decides to do with the parameters on the RRP program. Testing is scheduled to end on January 31, 2015. The FOMC may decide to change the rate on the overnight transactions -- like it plans to do in November and December. We suspect that the increase in the December RRP average rate (to 7.5bp) is driven by the Fed's desire to test its ability to floor interest rates in a period of extreme stress such as year-end. Since there is no such stress in January (or Q1, for that matter), there probably is not a strong case for changing the overnight rate from 5bp.... We think the Fed's testing of a term RRP is prudent given that balance sheet pressures are likely to become more than just a quarter-end phenomenon in the future as leverage ratios and daily average reporting become more binding, increasing the likelihood that market rates will trade below the RRP floor."
