The ETF Loophole (Almost) Everyone Missed

September 21, 2011

Amidst a barrage of ETF-related warnings issued this year by regulators, no one seems to have focused on settlement risk.

[This blog first appeared on our sister website,]


For the webinar we ran last week on the subject of regulation, I put together a slide listing the ETF-related warnings that have been issued by an alphabet soup of international supervisory bodies since 2009. In repeated public statements, regulators have reiterated their concerns over ETFs’ structure, their collateral and counterparty risks, possible contagion effects from ETFs to the broader financial market, short selling, leverage, and fallible liquidity. Ironically, though, given what’s just happened at UBS, not one warning mentioned the issue of clearing and settlement as a potential concern.

Although we still don’t know the full details of the fraud at the Swiss bank, and my colleague Kumaara Velan reminded us yesterday of the danger of jumping to conclusions on the basis of insufficient evidence, multiple reports now point to loopholes in the way ETF trades are reported, cleared and settled in Europe as the key to unlocking the latest rogue trading scandal. We tried to fill in some of the details ourselves a couple of days ago.

For many observers, it’s still hard to understand how UBS could have traded in ETFs using settlement dates that extended weeks into the future—as they apparently did—without the requisite post-trade checks occurring.

“When I found out banks were not confirming forward ETF [trades] until settlement date, I was pretty surprised,” Conrad Voldstad, chief executive of ISDA, the trade association for the world's over-the-counter (“OTC”) derivatives market, said yesterday, according to Reuters.

Many criticisms of European ETF market transparency have focused on the fact that a large proportion of trades is conducted OTC. However, the simple fact that transactions may be conducted away from a public exchange doesn't relieve the counterparties involved from the obligation to record, administer and settle them properly, Voldstad is implying.

“Trillions of dollars trade every day in the OTC market…these trades are verified by back and middle office personnel, generally within 24 hours. For the operations department [at UBS] not to have called or e-mailed the fictitious counterparties to verify these multi-billion dollar trades does not make sense. Given the magnitude of the losses on the real trades, the fake trades must have had a multi-billion dollar gain, creating a large counterparty credit risk which should have been elevated to the credit risk department for vetting, margin calls or other action," one industry insider commented yesterday on a Wall Street Journal article, making exactly the same point as Voldstad.

Details of how exactly the UBS fraud escaped the attention of the bank’s supervisors, and why trades were apparently not documented properly, will emerge in due course. However, according to many press reports over recent days, several involving apparent leaks from people inside UBS, there’s been a widespread practice in Europe of failing to confirm immediately those ETF trades that are conducted on a bilateral basis. Two days ago asked two European investment banks with large “delta-one” desks—Deutsche Bank and Credit Suisse—whether they send out confirmations as a matter of course following trades in ETFs. Neither bank has yet responded. I should add that there's no suggestion that either institution was involved in the UBS scandal.

On the basis of the available evidence, it appears that the UBS rogue trader combined two key bits of knowledge: awareness of the fairly liberal trade settlement rules in London (where there’s little sanction for late settlement, a compulsory “buy-in” of unmatched trades only occurs 30 days after the intended settlement date, and so by itself the forward settlement of an ETF transaction might not have raised suspicions); and the knowledge that many counterparties wouldn’t automatically request trade confirmations.

Taken together, these two loopholes may have enabled the creation of fake transactions in UBS’s systems. Even if this was the immediate cause of the fraud, the bank’s risk controllers seem to have missed other warning signs. High gross trading positions, even if the trader reported his position as hedged, plus what were presumably significant cash outflows in margin as the result of losing futures positions, might together have been expected to flag that something was wrong. Perhaps this is how the fraud was eventually spotted.

But it’s now clear that there’s a specific ETF element to the story too.



Find your next ETF