Today was chock full of derivatives-related news. One big story was that Senate Ag Committee Chair Blanche Lincoln is apparently going to propose a derivatives regulation bill that can best be described as the Dodd bill on steroids, with very few end user exemptions and a strict exchange trading requirement. You can guess my appraisal of those ideas, but I’ll defer my analysis until something official is released. I will say now, though, that it is evident that the administration is going full bore with a derivatives crack down. Obama spoke out on the issue personally today, and Treasury people have been very vocal. Ugh.
The other big story was that an un-redacted version of the Valukas report was released, revealing the names of the winning bidders for Lehman’s CME positions, and the payments to them. According to the report, the winning bidders received consideration worth about $1.2 billion dollars more than the liabilities that they assumed. Valukas argues that the Lehman estate could have a claim against the CME for forcing the liquidation of the positions, and against the winning bidders.
There are several related points that are relevant here.
The first is that Valukas’s valuation was based on settlement prices on Wednesday, 17 September, 2008. The auction was held on the morning of the 18th. The positions were large, and there is no guarantee that they could have been liquidated at those settlement prices. The report details only the values of the options in the Lehman portfolios, based on the previous day’s settle. This means that it values the futures at zero, meaning that it implicitly assumes that the settlement price hadn’t changed over night.
The SPAN margin values for the positions were large, indicating that the positions were large and risky. Particularly given the extraordinary conditions at the time, assuming that the positions could have been liquidated at the previous day’s settle is extremely problematic. Prices were extraordinarily volatile, and they could be expected to get only more volatile in the aftermath of a Lehman bankruptcy. Moreover, liquidity was drying up, major financial institutions were suffering serious liquidity strains, and could not count on tapping the capital necessary to carry big risk positions. Given the conditions in the market, it was possible that it would take some time to unwind positions assumed in the auction, and that the winning bidders would have faced considerable risk of loss during that unwind process. In these circumstances, it was inevitable that firms would assume these positions only at discounted values. Just how big a discount was warranted is impossible to say given the sparse disclosure in the Valukas report.
Indeed, even with much better knowledge of the positions, given the turmoil prevailing at the time determining the “fair” price to assume such a large position is a difficult and highly subjective exercise. (Knowledge of the CME GLOBEX order book would be somewhat helpful in estimating market depth, but still any analysis would involve considerable conjecture.) The SPAN risk numbers are probably a better benchmark than the position marks that Valukas used to derive his $1.2 billion figure. The transfer of margin monies exceeded the SPAN requirements by $100 million, but given the exceptional conditions at the time, SPAN is a problematic benchmark too. SPAN attempts to adjust to reflect changes in volatilities and correlations, but whether it did so in a realistic way on 17 September, given the virtually unprecedented conditions in the market, is open to serious question. It is quite possible that the bidders had much different estimates of the risk that they were assuming than that generated by SPAN, and bid accordingly. I would argue that the sophisticated bidders had at least as good information and risk modeling as that embedded in SPAN, and almost certainly better, for reasons I’ve set out in some academic papers over the years. SPAN also doesn’t explicitly incorporate liquidity risks (i.e., risks related to the time it would have taken to move positions, and the price impact of doing so), but in the conditions at the time liquidity risks were arguably huge, and bidders would have taken that into account. I would therefore not place too much weight on second guesses based on SPAN.
Suffice it to say that Valukas may be right that there is a colorable claim, but the case is far from proven, and it will be extremely difficult and contentious to prove it.
The second point is that although the replacement risk associated with defaulted positions has received a good deal of attention in OTC markets, it is not absent in cleared markets. The idea is that in OTC markets, when somebody defaults, its counterparties have to find new firms to take the place of the defaulter. In stressed market conditions associated with the failure of a major dealer or trader (e.g., LTCM), it is likely that these replacement trades will occur at prices that deviate substantially from the prices prevailing before the default; those stepping in for the defaulter are only willing to trade at far more favorable prices than the pre-default prices.
Well, that’s exactly what happened here. Goldman, DRW, etc., replaced Lehman as the counterparty on these trades, and were only willing to do so at prices that differed substantially from the pre-auction prices. In the event, Lehman’s collateral was sufficient to cover the replacement cost (this can happen in the OTC market too), but it is not outside the realm of possibility that the collateral of a big defaulter would not be sufficient, and that other clearing members would be called on to make up the difference.
This should be a cautionary tale for those hot to force a substantial extension of the use of clearing. Many of the products that are not cleared now, but which would be cleared under mandates, would be less liquid, more difficult to value, and pose risks more difficult to evaluate than currently cleared products like those that Lehman held at CME (and similarly, at LCH.Clearnet’s SwapClear). As a result, any replacement cost problem would likely be more severe than that experienced at CME in 2008, and the risk that collateral would be insufficient would be greater.
Put differently, just because clearing worked OK (not great) for the futures and futures options at issue at CME with Lehman, does not mean that things would work just as well with an expanded set of cleared products. Indeed, things would almost inevitably be worse. The set of cleared products is not exogenously determined: it is endogenous. The products that are easier to clear are more likely to be cleared now. Forcing, exogenously by government fiat, clearing of more products which almost by definition are more difficult to clear (and perhaps substantially so) means that the process of handling the next Lehman will be messier than the process of handling the last one at CME.
And think of the spillover effects. Consider a realistic scenario in which during the next crisis, difficulties associated with positions that are cleared only because of a mandate imperil a big clearinghouse, like the CME’s or ICE’s. This could impose losses not just on those products, but on the other products cleared, including those that would have been cleared without the mandate. (For instance, if replacement cost losses on these newly cleared products were so large that the CME clearinghouse’s financial condition were threatened, those trading ordinary futures contracts could suffer default losses too.)
Again, clearing does not make risk disappear. The experience of CME with the Lehman situation shows that clearly. Forcing risks that are harder to manage into a clearinghouse does not necessarily improve the safety of the financial system. It could do the exact opposite, by threatening the safety of clearinghouses that could handle the risks they currently manage, but not the additional, more problematic risks forced upon them.
The third issue relates to the auction process. CME originally solicited bids from 6 firms, and then in a second round from only 5 (dropping Morgan Stanley because its earlier bids had been uncompetitive.) Felix Salmon considers this “scandalous”:
Not only was the whole thing utterly unprecedented, it was also far from transparent: only six firms were invited to bid. If the CME was actively trying to get the lowest possible bids it’s hard to think how they could have done better than this, holding a hurried auction in the midst of utter market chaos, with no minimum bids and seemingly not a care in the world about whether Lehman was getting a fair price for its assets. It looks very much as though the CME wanted to hand Lehman spoils to its largest clients, since Lehman itself was clearly not going to be a client going forwards and was in no position to object.
Regarding “unprecedented”–well, circumstances were pretty unprecedented. Regarding “hurried . . . in the midst of utter chaos”–is Salmon proposing a langourous process, waiting for the market to return to “normal”? Like, when, exactly? And what if CME had waited, and Lehman had gone bankrupt, what then? Would it have been any easier to deal with these defaulted positions? As if.
Regarding the number of bidders. Consider the CME’s position. It didn’t want to jump from the frying pan into the fire, and transfer the positions to another firm or firms that didn’t have the capital to handle the risk thereby assumed. It wanted to move the positions into firms that it felt could handle them. (Not that, in retrospect, Goldman and Morgan Stanley were in much of a condition at the time to handle the risk.) Under the circumstances, limiting the number of bidders makes some sense.
And consider the irony of Salmon’s concern for whether Lehman was getting a “fair price.” (Just exactly what were fair prices “in the midst of utter market chaos,” Felix?) The whole reason that many–including Salmon–advocate a move to clearing is to protect other financial entities from the knock-on effects of a default by a particular entity. That’s exactly what CME was trying to do: to protect the other clearing members from the knock-on effects of a default. Yes, that has distributive effects: to the extent that the clearinghouse reduces the losses its members suffer, the defaulter’s other creditors lose, dollar for dollar. But that’s part of the clearing package, as I’ve written repeatedly. By advocating clearing, you are providing advantages to one set of creditors, at the expense of others. That may be advisable, if the advantaged creditors are more systemically important than the disadvantaged ones. But if you advocate that one group of folks should get first crack at the seats in the lifeboats, in the event you really can’t cry over the people left at the rail while the ship goes down. That’s a predictable consequence of the choice.
One would hope that this window onto clearing as it is, as opposed to clearing in the fantasy land that is DC, would be a revelation to Timmy!, and Gensler, and Dodd, and Frank, and now Blanche Lincoln. But, alas, one would almost surely hope in vain.