Welcome back! I believe that 2014 promises to be an interesting year for the Repo market. In general, it will be an extension of events that occurred last year. Between ultra-low rates, new regulation, tapering, and the Fixed-Rate Reverse-Repo program, there a lot to talk about, so let’s take a look back at 2013:
General Collateral In 2013
Based on the averages, general collateral (GC) appeared to trade about the same as fed funds in 2013. Whereas GC averaged 10.4 basis points for the year, fed funds* averaged 10.8 basis points, a difference of only 0.4 basis points. Though the averages seem about the same, the graph [below] tells a different story. It means that about as many days that GC traded above fed funds, it also traded below. There was no clear trend in GC this year, nor for Agencys which averaged at 11.2 basis points and Agency MBS which averaged at 12.6 basis points for the year.
The size of the tri-party market declined by $236 billion** in 2013 and that was mostly due to a large decrease in Agency MBS funding. That means one of two things, or perhaps a little of both. The decline could be due to the market selloff, tapering, and repositioning out of MBS. Or, perhaps, the market is starting to listen to the Fed about tri-party reform and cutting back on tri-party financing willingly. An interesting point here is that the decline in Agency MBS funding is slightly offset by an increase in structured repo (+$10 billion) and equity repo (+$27 billion). Both are good signs that the Repo market is trending back to its pre-crisis funding liquidity.
Fixed-Rate Reverse-Repo (FRRP)
What Sparked The Creation of The Program?
The official “reason d’être” of the FRRP program was to help manage interest rates better. I suspect the real reason was to keep GC from flatlining at 0%, like I had been warning prior to the announcement of the program.
Back in June, when overnight rates were averaging 5.5 basis points, I estimated that QE2 moved GC rates down 2.7 basis points a month and QE4 moved GC rates down 2.6 basis points each month. With continued QE purchases, General Collateral could have been at dead zero by November.
That created the potential for the QE purchases to result in a major distortion in the Repo and funding markets, and at some point, something had to give. That change came courtesy of the Fix-Rate Reverse-Repo program.
In August, the Fed dropped a bombshell on the market when they partially announced a new Fixed-Rate Full-Allotment Reverse-Repo Facility (FRRP) via the July FOMC minutes. The Fed later clarified the details of the new program. The fixed-rate started at .01% – one basis point, it was increased to a high as .05%, the highest rate authorized by the FOMC, and it’s now back down to .03%. Each reverse-repo counterparty was originally allowed to invest up to $500 million in cash at the Fed, that counterparty limit was recently increased to $3 billion. There are 139 different reverse-repo counterparties, which means the FRRP can now put $417 billion in securities into the market on any given day.
January Mini-Collateral Shortage
The “January mini-collateral shortage” or “January soft funding” hit again last year. It first appeared in 2005, though there were hints of the pattern back in 2003 and 2004. There are several explanations for the soft funding at the beginning of the year. A recurring explanation is that dealers cut back their Repo matched-books and customer financing before year-end to reduce balance sheet. As customer cash comes back into the market after year-end and the Street has not yet build back its trading positions, there is excess cash in the Repo market. More on this soon.
Details of the Treasury’s Floating Rate Note (FRN) were announced with the quarterly refunding on May 1. The floating rate index had came down to either an overnight GC Repo rate or the auction rate of the weekly 13-week bill auction. The Treasury chose the weekly T-Bill auction rate which I believe was a mistake. A daily overnight Repo rate is a better floating rate and represents where U.S. Treasurys are funded each day. Treasury Bills trade in their own world, based on the supply and demand for bills, which is not the same for other Treasurys. In addition, a floating rate based off a spread to a weekly auction rate is complicated.
April Collateral Shortage
Almost every year, GC rates decline from the middle of April to around the middle of May and it has come to be known as the “April Collateral Shortage” or “Seasonal April Collateral Shortage.” It occurs when a large amount of Treasury bills mature right after the April 15 tax date and those maturities create a shortage of collateral in the Repo market. There isn’t a collateral shortage every April, but it’s a fairly regular occurrence.
Last year’s April Collateral Shortage lasted longer than normal. After a small pause for April month-end, the “collateral shortage” remained in full force through May, mostly because the Treasury happened to be paying down a large amount of debt ($80 billion) during the month while the Fed was buying Treasurys in the QE program.
Debt Ceiling Crisis, Part Deux
In the bond market, the October 2013 Debt Ceiling Crisis was a mirror image of the July/August 2011 crisis. Once again, it’s all about the shortest Treasurys and GC Repo markets. The really crisis is about which Treasurys or Repo trades could miss a payment, and not whether the entire U.S. bond market would default. Luckily, the crisis ended before a technical default was tested and rates have since moved back to their normal trading ranges.
Outside of the 10 year note, it was a pretty quiet year for Specials in the Repo market last year:
Average Specials Rates for The Year:
2 year = +4.2 bps
3 year = -1.5 bps
5 year = -3.9 bps
7 year = +3.7 bps
30 year = +6.3 bps
10 Year Note
The 10 Year Note was a little different, averaging at -24.5 basis points for the year. It really heated up on specific occasions for just one reason – the auction cycle. First, 2013 ushered in a new wave of short-selling in the 10 year sector. Second, the 10 Year Note is on a very specific auction cycle. On February, May, August, and November 15 a new 10 Year Note is issued. It remains a single issue for one month then it’s reopened. After another month, it’s reopened again making it a triple issue. Now, triple issue Treasurys rarely trade Special, if they ever do. Double issues can trade special under extreme circumstances. However, single issues love to trade special.
And that’s where the auction cycle comes in. An older 10 Year Note which has been reopened twice can hold many shorts and not trade special. But once the new, single issue 10 Year Note settles, there are large shorts rolling into the “current” issue. Shorts in a triple issue moving into a single issue. That’s when the 10 Year Note became extremely special – about three times in 2013. The short-base of the 10 year sector was so deep that the shorts overwhelmed the single issue 10 Year Notes. Could this happen again in 2014? Perhaps. It will if there’s another major selloff in the long-end of the market.
For a good part of 2013, the threat of the European Financial Transaction Tax loomed over the European Repo market. Market participants expected the European Repo market to be devastated: to fall by 99% and possibly it “would disappear overnight.” Later in the year, a group of European Union lawyers concluded that the tax is illegal under the EU treaty, but the backers of the tax continue to move forward, so it’s not over yet. I believe the Repo market will be ultimately exempt from the tax.
The Fed seems interested in setting haircut floors for the Repo market based on regulatory haircuts. The Fed’s Tarullo stated that it’s “an indirect way to of extending bank capital requirements to the shadow banking system.” Look for more Fed official discussion on this in 2014.
Liquidity Coverage Ratio (LCR)
The Fed wants the strict LCR rules required by Basel III to be 80% implemented by Jan 1, 2015. That’s less than one year away! It’s just one more reason why in 2014 large banks will continue to shed assets and businesses which require a large scale use of regulatory capital.
Predictions For 2014
- Large Bank Repo Matched-Books Eliminated – No more “market making” in Repo, securities finance, and/or stock loan for customers. The matched-book businesses will disappear at large banks, spreads will widen to reflect the additional regulatory costs, and new players will fill the vacated space.
- New Trading Counterparties – There will be significantly more trading with central counterparties (CCPs), quasi-government entities, central banks, and exchanges. These entities are mostly exempt from regulatory capital charges and taxes.
- Trading Repo Direct – New Repo business will evolve between end-buyers (e.g. money funds) and end-sellers (e.g. investment funds) – effectively eliminating many “middlemen” in the Repo market.
- Regulatory Business Migration – Wherever there’s a regulation, there’s a loophole or an exemption. The pendulum is swinging back to favor securities businesses at large U.S. broker-dealers.
- Rate Band – I believe the Fed may eliminate the use of federal funds and a policy toll in favor of a rate band between the IOER and FRRP. The IOER pays banks .25% on excess reserves which sets the upper limit. The FRRP is set a 3 basis points and serves as the lower band. More discussion on this in 2014.
* I use the fed funds opening rate because it’s the most used in the market as a morning benchmark rate. In addition, the timing is close to the 10:00 am GC average. The fed funds effective is not good because of late afternoon volume swings.
** Fitch Ratings; “Repos: Market Decline Amid Policy Risk”; December 20, 2013.