On Wednesday, we asked if Monday’s catastrophic ETF collapse which saw over 200 funds fall by at least 10% was just a warmup for a meltdown of even greater proportions.
The problem, you’ll recall, was that in the midst of Monday’s flash-crashing mayhem, a number of ETFs traded at a remarkable discount to fair value. Essentially, market makers looked to have simply walked away (there’s your HFT "liquidity provision" in action) or else put in absurdly low bids in order to avoid getting steamrolled when the constituent stocks came off halt. The wide divergences weren’t arbed for whatever reason and the result was an epic breakdown of the ETF pricing mechanism.
As we wrote on Wednesday, this was proof positive that contrary to popular belief (which, incidentally, is itself contrary to common sense in this case), an ETF cannot be more liquid than the assets it references and when liquidity dries up in the underlying as it did on Monday, the market structure is clearly inadequate to cope.
But don’t worry, because the problem has been identified.
It’s simply a "computer glitch" at Bank of New York Mellon. Here’s WSJ:
A computer glitch is preventing hundreds of mutual and exchange-traded funds from providing investors with the values of their holdings, complicating trading in some of the most widely held investments.
The problem, stemming from a breakdown early this week at Bank of New York MellonCorp., the largest fund custodian in the world by assets, prompted emergency meetings Wednesday across the industry, people familiar with the situation said. Directors and executives at some fund sponsors scrambled to manually sort out pricing data and address any legal ramifications of material mispricings, those in which stated asset values differed from the actual figures by 1% or more.
A swath of big money managers and funds was affected, ranging from U.S. money-market mutual funds run by Goldman Sachs Group Inc., exchange-traded funds offered by Guggenheim Partners LLC and mutual funds sold by Federated Investors. Fund-research firm Morningstar Inc. said 796 funds were missing their net asset values on Wednesday.
Ok, got it. So basically, if you want to know what the NAV of your fund is, you’ve got to go stock-by-stock and calculate it the old fashioned way. And what, you might ask, does this mean for investors in these most liquid of all securities?
The effects of the breakdown are threefold: It has made ETFs more costly to trade, hindered investors’ ability to trade accurately in and out of popular investment vehicles, and forced fund companies to scurry to price securities.
Here we see the hallmarks of liquidity: i) rising trading costs, ii) an acute inability to trade in and out of the market accurately, and iii) issuers that have no idea what’s going on.
And how about HFTs, who, you’re reminded, insist that they provided liquidity during Monday’s chaos?
Several traders said they were forced to calculate their own net asset value for ETFs and that they widened the spreads, or the difference, between listed buying and selling prices to accommodate for the higher risk of trading.
"We measure our edge in terms of subpennies," one trader said. "We can’t afford to be off by a penny."
So if we had to venture a guess as to what might have happened here, it might go something like this: once the halts got started, it became impossible to calculate ETF NAV causing "liquidity providers" to widen out their bid- asks in order to protect themselves against NAV uncertainty, and that, in turn, caused anyone who had a market order in to hit the bid at absurdly low prices, taking out stops, and before you knew it, the rampant confusion simply caused Bank of New York Mellon's/ SunGard's platform to malfunction.
Of course we'll never know what really happened, but what we can say is that if the following is true, it would be some damn coincidence:
The outage wasn’t related to the market turbulence Monday that included the largest-ever intraday point decline in the Dow Jones Industrial Average, the bank said.
* * *
Incidentally, the official word is that the problem was caused by "an operation systems change performed on Saturday." Read the note below and decide for yourself.