The big story this week is quarter-end. Not because securities became extraordinarily scarce or overnight rates were extraordinarily volatile, but because the market planned to use the Fed’s Fixed-Rate Reverse-Repo facility (FRRP) for a large amount of window dressing. Those plans were foiled by the FOMC two weeks ago.
Since the facility’s creation once year ago, banks were using the FRRP to have the Fed as a 0% risk-weighting counterparty on their books for quarter-end in order to reduce Basel III regulatory capital charges. On June 30 quarter-end, a record $339 billion flowed into the FRRP. This quarter-end, it was estimated that banks were going to place between $400 billion to $500 billion at the FRRP. However, after the Fed announced the new $300 billion limit on the program, banks were scrambling to find other assets to fulfill their window dressing needs. With a limit on the facility, hundreds of billions of dollars began looking for a regulatory risk-free home and U.S. Treasury bills yields dipped into the negatives.
Today, GC averaged at -.015% (-1.5 basis points) and federal funds at .05%. A total of $407 billion in cash was submitted to the FRRP, but only $300 billion was accepted. This is the first time I’ve ever seen GC average negative on quarter-end.
The Fed published an analysis* about the puzzling increase in U.S. Treasury fails which occurred in June. All total, there were $1.2 trillion fails for the month, including $627 billion in on-the-run issues, like the current 2-Year, 5-Year, and 10-Year Notes, but, most interesting, $470 billion of those fails were in off-the-run issues, those defined as issued more than 6 months prior.
Fails in the on-the-run issues, like the 2-Year, 5-Year, 10-Year Notes are all pretty standard – the kind of gossip that’s always flying around the Repo market. The main reason for fails is that there’s a short-base in the market and there’s not enough supply for all of the shorts to get covered; thus resulting in fails. For the 10-Year Note, it was all deemed a part of the auction cycle. I’ve described this phenomena many times before: When a new 10-Year Note is first issued, it’s only a single issue. The previous 10-Year Note is a triple-issue, having been re-opened twice already. As the potentially massive amount of shorts that can be contained in a triple-issue roll into a single-issue, it often creates a shortage. The 5-Year Note failed from June 20 to 26, and then again on June 30. June 30 was clearly related to quarter-end and probably a deep short-base in the 5-Year Note explained the June 20 to 26 period. However, the 2-Year Note is interesting. There was one single day when $71 billion 2-Year Notes failed and then cleared up next day. The fail was not related to the auction cycle, instead, it was a one day event when a significant amount of supply disappeared from the market. Was it a single large trade changing hands or did a large customer choose not to loan their position that day?
The explanation of the off-the-run fails is even more interesting. Perhaps, it’s a sign of a fundamental change occurring in the Repo market. Overall, the off-the-run fails were small in size, frequent, and widespread across many Treasury issues. That’s a bad sign. There was no flight-to-quality, frequent operational errors, or a dealer squeeze which helped create the fails. It basically means there was less intermediation by banks; less market-making in the Repo market.
If there’s less intermediation in the Repo market, it means there is less supply of securities getting from end-owner to end-borrower to cover shorts; especially on month-end and quarter-end. Basically, when there’s a supply shortage, there are fewer dealer banks willing to get supply from the end-user portfolios to cover the shorts in the market. It is a result of banks managing their balance sheets more actively; they’re trying to reduce their regulatory capital changes. Here’s an example: suppose it’s month-end and there’s a security which becomes very special in the early afternoon. One bank knows a customer who owns security and they can borrow the security from that customer and then loan it to the entity who’s short. However, if regulatory capital charges are being calculated that day, there’s a significant marginal cost to the trade. It’s the cost of capital for the bank for 30 days versus the revenue on an overnight Repo trade. Yes, even if the bank makes the 300 basis points (3%) overnight on the fail charge, the trade is still not economical. Therefore, the trade doesn’t get done and the result is a supply shortage in the market.
In the past, fails were generally the result of dealers purposely holding supply out of the market. Now, it appears, the fails are increasing due to less dealer intermediation (market-making) in the Repo market. And all of this will kick-in even more as more of the new regulations are fully implemented.
Liquidity Coverage Ratio (LCR)
The Liquidity Coverage Ratio was approved by the Fed and FDIC on September 3rd. It requires banks to hold enough high-grade securities (like Treasurys) that can be sold during a crisis and provide funding for the bank for 30 days. It’s estimated that banks have already added $200 billion in high-quality assets to their balance sheets and are expected to hold as much as $2 trillion once the new rule is fully in effect. In the Repo market, it’s expected the LCR will push Treasury Repo rates lower by creating a permanent premium for Treasury collateral in the market. Of course, that’s assuming the securities are not actively loaned into the market. If a bank can loan the securities into the Repo market, there won’t be as much of a market distortion.
Net Stable Funding Ratio (NSFR)
This month, a group of banks sent a letter to the Basel Committee about the NSFR, saying they believe it will make short-selling stocks 5 times more expensive for Basel III banks. They concluded it will “significantly increase transaction costs across equity markets.”
New Regulation Timeline
- The LCR implementation was postponed until July 1, 2015 for many banks and January 1, 2016 for the largest banks.
- Leverage Ratios – The Basel III Leverage Ratio is to be fully implemented by 2018; the Dodd-Frank Leverage Ratio by January 1, 2015 for U.S. banks, and foreign banks have until January 1, 2016.
- NSFR is not a final rule yet, it’s only a draft. If it is implemented, it’s not binding until January 1, 2018
* Liberty Street Economics; September 19, 2014; “What Explains the June Spike in Treasury Settlement Fails”