Bloomberg posted a piece yesterday on a plan by the FICC to “guarantee the most-liquid assets used as collateral between dealers and investors in the tri-party system.” Although not a public announcement, the article relied upon “people familiar with the discussions, who asked not to be identified because the talks were private.” Does this mean that the Fed has finally found a solution to the issue of ‘fire sales’ and the systemic risk they can cause in the financial markets?
On the subject of tri-party repo, the Fed has pursued two clear goals. First, through its tri-party reform task force it addressed the historical weaknesses in the operational aspects of the market through the implementation of a three way trade confirmation process, a 24 hour trading day and a dramatic reduction in the daylight exposure of the tri-party clearing banks. Although the Fed was at times frustrated with the pace of these reforms, the full implementation of this initiative is in sight. The second goal was the more difficult issue of how to ensure that a default by one significant dealer did not precipitate a market-wide liquidation of assets as the sale of the defaulting dealer’s positions led to a downward revaluation of other dealers’ positions, aka, a fire-sale.
In solving for a fire-sale dilemma, as always the devil is in the details. As outlined in Bloomberg, the proposed FICC facility would cover US treasuries and mortgage backed securities backed by FNMA and FHLMC. We assume that although not mentioned in the article, US agency securities would also be included along with GNMA MBS. This would cover approximately 85% of the current $1.6 trillion daily tri-party market. No bad.
In theory, the FICC would “guarantee” these transactions but what does that mean? The current market relies upon BNY Mellon and JP Morgan Chase to act as the tri-party agent between seller and buyer. The FICC won approval in January from the SEC to allow registered investment firms to become members of its securities division, but it is hard to envision that a structural shift of a large portion of the market away from the two traditional clearers is likely, especially if BNYM and JPM are then left with the non-liquid portion of the market. Further, what exactly would the FICC be guaranteeing? No one can be expected to guarantee a client against the possibility of market price movements. What the FICC could put in place is an orderly liquidation procedure, but is this a substantial improvement over the current market structure? And at what cost? No one can be expected to provide a service to the market for free.
The issue of fire-sales is certainly an area that the Fed should address. The FICC plan is a viable solution for 85% of the market, but is the 85% that does not need a solution. The loss of liquidity that led to fire-sales following the demise of Bear Sterns and Lehman Brothers involved primarily non-liquid assets. The market for US treasuries and agency MBS securities (at least pools) is deep and liquid. Faced with a loss of funding a dealer can quickly reduce his positions in an orderly manner. In addition, as long as he remains a member of GSCC he can fund those securities in the inter dealer market. In the event of default the market can absorb substantial inventory without a significant movement in prices. It also has to be pointed out that a default of a major dealer generally leads to a flight to quality which increases the value of high quality assets. While it may be reasonable to question whether a money market fund will have the resources or incentive to find the optimum price for a security, this concern is probably somewhat overblown given the long established procedures in this area. But all told, the “solution” to fire-sale risk in this 85% category sounds a bit like the story of the Emperor Has No Clothes.
The segment of the market that cries out for a solution is the 15% ($240 billion of non-liquid collateral) that is not addressed by the FICC plan. The article only mentions that “industry organizations are working on guidelines, or best practices, that would include ensuring investors, are equipped to trade or liquidate the assets in an orderly manner after a dealer default.” By definition non-liquid collateral (corporate, asset backed securities and private label MBS) is substantially more difficult to price and liquidate in a distressed market than the liquid collateral included under this plan. These assets also generally fall in price when the market is in turmoil and market participants “flee to quality.”
There is no easy solution to the risks associated with funding this market segment. Even the creation of a liquidity buffer and efforts to lengthen the maturity of funding trades for non-liquid assets may not be enough to ensure that a fire-sale does not occur. In addition, there is often a chicken and egg aspect to this dilemma. As often as not, the event that impacts a dealer’s liquidity is related to a revaluation of non-liquid assets in the broader market place. A loss of faith in the asset quality of Alt A MBS for instance could precipitate the first margin call along with a flight to quality. This is an event which cannot be addressed in the repo market alone.
While we applaud the market’s and the Fed’s efforts to comply with Dodd-Frank and remove systemic risk from the repo market, we feel that this is really only a first step. It is the reform of the smaller but substantially more risky non-liquid repo market that will prove the more important and problematic area of reform.
The Bloomberg article, “Repo Fire-Sale Plan Said Within Reach After Fed Sounds Alarm,” can be found here.