Sandra Krieger, chief risk officer at the Federal Reserve Bank of New York spoke at a conference late last week and had some things to say about repo. There were some interesting comments about fire sales and collateral.
The only press coverage we saw was in the November 16th WSJ, called “Work Still Needed to Reduce Repo Market Risks: Fed Official” by Katy Burne. We can’t link to it since it is behind the pay wall.
Burne wrote,
“…A key concern to regulators, Krieger said, is “fire sale risk,” when institutions fearing a default by their trading counterparty will elect to raise cash quickly by liquidating assets for fear of not being repaid. That can lead others to mark down the value of their assets, leading to a downward spiral of several firms unwinding at once…”
“…Krieger said the fire sale risk is “something we are thinking about internally.” She said firms have “the incentive in the face of a possible default of their dealer counterparty to front run and bail out,” adding, “that can be a problem for mark to market losses, capital erosion and prompts further de-leveraging … and before you know it you have systemic risk on your hands…”
The issue of fire sales have always been of interest. We feel the need to acknowledge the debate between wanting to rapidly liquidate securities when there is a default, in an effort to stop the bleeding quickly but with the potential for prices being driven way out of whack, versus the preference to hold onto paper in the belief that a little patience will avoid massive disruption and losses — but creating enormous uncertainty about the quality of balance sheets. When a counterparty defaults and funding liquidity disappears, it leaves little choice but to sell. Repo agreements typically allow for very quick liquidation, avoiding becoming enmeshed in lengthy litigation. Some have suggested that it might be better to slow the process down by making financing trades subject to automatic bankruptcy stays.
During the financial crisis the Fed allowed broker/dealers to access liquidity via the PDCF and other programs. This lifeline gave time for the markets to calm down and hopefully minimized fire sales. But prices still dive-bombed. The uncertainty about what the banks were holding on to did huge damage to confidence. So is the choice to sell quickly and see the impact ripple systemically or provide liquidity to allow the banks to wait it out — but create the information uncertainty that was so crippling in the crisis? What’s behind door #3, please?
Burne also wrote “…Krieger said more disclosure by participants in the lending and repo markets would be beneficial so that liquidations could be more orderly. “You could require that securities lending and repo [firms] have contingency plans for their largest counterparties, including in times of stress,” she said…” We suspect this might have been reversed or, at least, should go in both directions. In tri-party the cash flow is to the repo firms, funding their house and client positions. The unwind/rewind process in tri-party is now shorter and the cash providers are exposed to the collateral for much longer. Some of those securities they may end up owning might not be eligible for their portfolios, in which case they will be very motivated sellers indeed. Cash providers need contingency planning, too.
On October 1, 2012 Benoît Cœuré, a member of the ECB Executive Board, gave an important speech where he spoke about collateral scarcity. But embedded in the speech were some interesting comments about collateral fire sales. He said “…There are several reasons why central banks may need to adopt a broad collateral framework, particularly in a crisis, and to have it broader relative to the interbank repo collateral set. First, the role of a central bank as a backstop facility avoids a worsening crisis of confidence, i.e. it avoids a negative feedback loop-of-liquidity crisis. Second, a central bank is the only player that never has liquidity constraints. Therefore, in case of counterparty default, the central bank has ample time to liquidate collateral and await mean reversion of values. Third, central banks can impose haircuts that protect them from adverse scenarios in the collateral liquidation process, as they are considered risk-free…” It is interesting to think about how the central banks can opt to wait the market out. For the ECB, the wait might be lengthy, but that is a different matter.
A link to the speech by Benoît Cœuré is here.