General Collateral (GC)
General Collateral rates have been quiet for most of the summer, excluding of course, quarter-end. Since the beginning of July, GC averaged 9.3 basis points, which is two basis points higher than its 7.3 basis point average since the beginning of the year. For most of July and August, rates were trading on top of fed funds.
By the end of this week, I expect rates to spike higher due to a combination of month-end and new issuance. All total, between August 28 and September 4, $68 billion* in net new Treasurys will be issued into the market. Over the next few months, I still contend that Repo rates are on their way upward, with the combination of QE winding down and Treasury issuance still running about $50 billion a month. QE is now at $25 billion a month and moving down to zero within three FOMC meetings. Then, with no Fed program to soak up Treasury issuance, Repo rates will, ultimately, trend higher.
There’s a good short-base in this market. The 2 Year, 3 Year, 5 Year and 10 Year have all traded below -2.25% over the past two months. Remember though, Repo rates are a function of supply and demand for the specific securities in the Repo market. When demand increases (more shorts) rates move lower. No doubt, there are directional shorts (expecting a sell-off in the overall market) and hedging shorts (hedging their long positions in other securities). In addition, the current 10 Year Note is now a single issue with a good short-base and September is arriving very soon. Historically, there’s a lot of market activity in September and a quarter-end at the end of it, both factors which should push a single issue 10 Year Note deep into negative territory.
Once again, there’s no surprise there are more fails when there’s more short-selling. When short-selling demand for a Treasury is greater than the available supply and all of the shorts can’t get covered, the issue fails. Fails in the Treasury market remain relatively high and I expect that to continue. Fails-To-Deliver are running at $128.9 billion in regular Treasurys and $7.1 billion in TIPS. I understand why there are so many Repo specials failing, but the high fails number in TIPS is somewhat of a mystery.
June And July FOMC Minutes
IOER (Interest On Excess Reserves) And Federal Funds
Back in June, the FOMC hinted they would stick with the federal funds rate as the policy target rate and that the IOER “should play a central role” in tightening. Then, they followed up in July stating that “it would be appropriate to retain the federal funds rate as the key policy rate” and “the IOER rate would be the primary tool used to move the federal funds rate into its target range.” It’s pretty clear now – U.S. monetary will remain based on the federal funds rate and the IOER will be the primary tool for draining reserves.
FRRP (Fixed-Rate Reverse-Repo Facility)
Where does that leave the FRRP facility? Well, back in June the FOMC stated the FRRP “could play a supporting role” in draining reserves, but in July they indicated a sunset for the facility, “that the ON RRP facility should be only as large as needed for effective monetary policy implementation and should be phased out when it is no longer needed for that purpose” Perhaps, reports about banks using the FRRP for window dressing is starting to bother the Fed.
But here’s even bigger news and a fundamental change in Fed policy – they’re going to keeping the fed funds target rate as a rate band and not just a specific target rate. In the July minutes, “they supported continuing to target a range of 25 basis points.” The rate band will be such that “the IOER rate would be set at the top of the target range for the federal funds rate, and the ON RRP rate would be set at the bottom of the federal funds target range.” Interesting – technically, right now the fed funds target range is .25% to 0.0% and the FRRP is set a .05%, or 5 basis points above the lower band.
Before December 2008, the last time the Fed used a federal funds target range was June 1989 when the range was set between 9 1/2% and 9 5/8% Perhaps the Fed is really just getting back to its roots; for much of the 1980s, the fed funds target alternated between a target rate and a target range. Keep in mind, in the 1980s, Fed had a harder job fine tuning rates. The Repo market was less developed and nominal rates were much higher, which made miscalculations in reserves more apparent. If overnight rates moved 10% off of the rate target, it meant a swing of 90 basis points in 1989 verses 1 to 2 basis points today.
The Changing Repo Market
New regulation is driving changes in the Repo market. Because of the Leverage Ratios, Basel III, Dodd-Frank, and Tri-Paty Reform, many banks have reduced their Repo books. Since the beginning of the year, Goldman Sachs cut $56 billion of assets from their balance sheet which included $42 billion in Repo. Barclays cut their Repo book down $25 billion, BofA down $11.4 billion, and Citi cut theirs $8 billion. The reasons? For Goldman, it was to increase their Leverage Ratio to 4.5% from 4.3% and get closer to the 5% minimum that regulators mandated.
According to Jeff Kidwell at AVM after conversations with clients, “the supply of O/N collateral, particularly US Treasuries and Agency MBS at the broker/dealers, has dried up.”** If four large banks have cut $86 billion from their Repo books this year, that doesn’t even include about 40 other banks intermediating in the Repo market. It doesn’t end there, banks are also moving out of Treasury Repo and into higher spread collateral, which includes: corporates, equities, private label MBS.
During the year, banks have told hedge-fund clients that regulations have forced them to set aside more capital and that’s affecting the profits of their prime-brokerage business.*** Banks are even urging customers not to leave cash in their accounts and they’re charging clients for cash balances and imposing monthly fees.
It seems pretty clear, the Repo market is beginning to experience dislocation. With fewer banks making-markets, shorts are becoming harder to cover, specials trade at lower rates with more rate volatility, there are rate spikes on quarter-end, and a significant increase in the amount of fails.