This is the second post in my series of reviews of academic papers written in the last year about the market structure of the US Treasury market. As the most important fixed income market it is great to see so much discussion around the topic of market structure. This post is a review and commentary on “Enhancing Liquidity of the U.S. Treasury Market Under Stress” by Nellie Liang & Pat Parkinson
DISCLAIMER: The opinions expressed here are mine and mine alone. They are not associated with any employer of mine or organization I am associated with.
The authors present four proposals for reform of the market structure of the US Treasury market that would benefit the overall functioning of the market and some of which (#2 and #4) would enhance investor confidence.
#1 – Standing Repo Facility
Creation of a standing facility at the Fed to make cash available to dealers in the form of repo collateralized by Treasuries. This is a step in the right direction but has one fundamental flaw that contradicts their proposal #2 of central clearing, it puts the bank dealers in the middle of all transactions and reinforces their role as a chokepoint. The right answer is for the Fed to make their repo facility available directly to the end investors that need it and are the actual owners of Treasuries.
The reason that dealers pull back on market making is not that they cannot finance their position of treasuries, it is that they are faced with one way selling pressure from their clients who cannot obtain repo financing and are forced to liquidate. This one way selling pressure causes losses in the inventory that dealers hold causing them to pull back from providing market making liquidity to the market. Since bank dealers have no affirmative obligation to provide liquidity (contrary to the misconception that being a Primary Dealer has any secondary market obligations), acting in rational economic self interest bank dealers pull back.
This paper itself makes that point on p. 6 stating “the Fed announced on March 15 it would purchase Treasuries and agencies and offer repo financing, but there was little take-up on repo as dealers did not want to expand their balance sheets to provide additional intermediation.”
Since the Fed now lends money directly to corporations in the form of purchasing corporate bonds and ETFs, the concept of dealing directly with the end participant in the market is already established. With the high quality of the collateral and the proper haircut, the Fed would be insulated from the credit risk of the counterparty in the repo transaction.
For this proposal to work and have the intended impact of avoiding fire sales of treasuries due to lack of repo funding, then that funding has to be made to the actual owners, the end investors, and not be subject to bank dealers as an intermediary.
#2 – Expanded and Uniform Central Clearing
Central clearing is a natural good for the market overall and should be uniformly implemented in the market. The availability of central clearing will have the benefit of allowing the ultimate buyers and sellers of Treasuries to meet and transact freely with each other like other centrally cleared markets. The authors mention a benefit of central clearing benefiting bank affiliated dealers, and while this may be a tangential benefit to the market the main focus should be on the end investors and their confidence in the market and ability to transfer risk.
This proposal #2 would give the ability to expand the proposal #1 to all market participants because the repo could be centrally cleared eliminating the Fed having exposure to any counterparty in addition to the haircut and nature of the collateral.
#3 – Loosening Regulations on Banks
If central clearing above is done, then this becomes a non-issue because the market structure will have removed bank dealers as the pressure point in the risk transfer process.
#4 – Increasing data and Transparency
This is always a good for the market. More data collection, and more public dissemination in a more timely manner of all sorts of data will increase transparency and investor confidence.
These proposals are the right ones, but need to go the distance to have an impact. Instead of focusing on bank dealers and hoping that benefit will trickle down to the end investors, the focus should be on directly impacting the end investors.
In the opening I feel the authors make one wrong interpretation of the data. They claim that the selling was by foreign central banks, I feel the data shows it was by hedge funds and leveraged players not central banks. The data referenced in the Duffie (2020) paper on the “foreign holders sold, net of purchases, $299 billion of Treasury Bonds and Notes, far above normal.” Foreign holders here are hedge funds because they are domiciled in foreign locations, usually the Cayman Islands. This is likely not central banks who are not under the same forced liquidation pressures.