The European Repo Council, part of ICMA, came out with an interesting paper today, “Haircuts and initial margins in the repo market”. It was authored by Richard Comotto, Senior Visiting Fellow at the ICMA Centre, University of Reading. It is definitely worth a read.
A main theme of the paper asks why, on the same securities, do less creditworthy counterparties seems to suffer higher haircuts than better quality counterparties? Comotto’s point is if the haircut is intended to absorb losses in the case of a counterparty default that leads to liquidation, then shouldn’t the haircut be independent of the counterparty, depending in whole on the expected volatility of the collateral? It’s a fair point but not the way traders, credit and risk manager’s intuition take them.
The author bemoans the lack of good data on haircuts and the confusion between US and European markets. Comotto notes that the regulators are looking closely at repo market practices (especially the tendency for haircuts to be pro-cyclical) and may not have enough good information to make informed decisions. Comotto writes, “It is a matter of concern that the regulatory debate is taking place in the absence of a clear understanding of the precise function of collateral haircuts/initial margins but also without sufficient empirical data to confirm the scale of the role they may or may not have played in the recent crisis.”
Comotto tries to argue that, in Europe at least, since haircuts (on higher quality paper) did not go up very much during 2007-09 then higher haircuts were not a driver for the overall market de-leveraging. We think this is where the argument suffers from over-generalization. The repo markets can be be divided into three broad segments. The first is tri-party. In this market haircuts are hard wired by agreement. As a result they don’t get changed often and when a cash lender gets uncomfortable with the market, they will simply withdraw. The second segment is dealer-to-dealer trading. This has traditionally been done without any haircuts, although like (virtually) all repos, trades are margined. These are also sticky from a haircut point of view – traders are reluctant to change practice, lest they be on the receiving end the next time. Related to this segment are trades which clear through CCPs (LCH.Clearnet, for example). These too are inter-dealer, but initial margin is determined by LCH and can change during the course of the trade (as MF Global found out). Finally, there are bilateral trades done with clients. Here haircuts are negotiated all the time and adjust to market conditions. These segments should also be divided by geographical region and collateral quality too. You simply can’t generalize by looking at just one of those boxes.
The on-off nature of tri-party (and sometimes inter-dealer trades) leads to potentially very disruptive de-leveraging, much like a bank run. For client bilateral trades, the changes in haircuts during a crisis are muted for some types of securities, but large for others. In the latter case, de-leveraging is the result. Comotto notes that in Europe the vast majority of collateral are sovereign issuers and there weren’t material increases in haircuts (in that part of the market) in 2007-09. That is correct. But the de-leveraging in 2007-09 wasn’t really in high quality assets. Pointing to that segment as a reason to debunk the haircuts and de-leveraging linkage doesn’t ring true.
The article downplays the concept of correlation and wrong-way risk than permeates repo. We wrote about “jump to correlation” yesterday (a link is here). Comotto writes, “…Haircuts/initial margins could take account of the market impact of the sale or purchase of collateral securities following a default, in other words, counterparty credit risk feeds into market liquidity risk…” but then follows up saying, “…However, haircuts/initial margins should only vary with counterparty credit risk…where the counterparty and its holding of collateral securities are atypically large and/or market liquidity has been sufficiently drained by a critical loss of confidence to make the market unusually sensitive to the impact of the sale or purchase of collateral securities”. But in a crisis, everything becomes correlated and the risk is the aggregate book of business. Any risk manager who only uses a microscope to examine risk isn’t doing their job.
Traders one bitten are twice shy. It shouldn’t be surprising that traders (and the credit officers and risk managers behind the curtain) will think back to the extraordinary market illiquidity and sudden market correlation that came with the Lehman collapse. All but the best quality collateral was difficult to buy or sell and haircuts often inadequate. One wonders how the European markets are currently dealing with the paper issued by sovereigns in distress, even when the counterparties are high quality?
Comotto does bring up a very interesting point on how haircuts are used to generate a return. By re-hypothecating collateral at a smaller haircut (vs. the haircut used when taking the paper in), a dealer can generate liquidity. In this day and age when every dollar or euro generated is one less that has to be borrowed internally at high rates, there is a lot of incentive to manage haircuts aggressively.
All in all, this paper is worth a close read. A link to the paper is here.