Readers of this website know I'm very suspicious of flash crashes, those sudden market down turns where stocks, or even a few stocks or a single stock, plummet in value in a matter of second due to the sudden action of High Frequency Trading (HFTs) on "dark pools." I have advanced various hypotheses to explain this behavior, from the possibility that artificial intelligence is already here, looming, as it were, in the networks of these trading concerns, to the possibility that someone might be using them to probe market cyber-security and vulnerabilities. Well, so many people saw this one, and sent me this article, that I have to extend my high octane speculations once again. As one individual who shared this article stated in his email, it raises... "questions." Here's the story from our friends at Zero Hedge:
Ponder these statements:
Nasdaq has issued a market-wide trading halt amid what appears to be a "glitch" that sent a number of the largest Nasdaq-listed stocks to crash or spike to exactly $123.47 per share.
This move crashed the value of companies including Amazon and Apple, sparked chaos in Microsoft, while sending Zynga rocketing up more than 3000%.
On the eve of the US Independence Day holiday and in after-hours trading, The FT reports that market data show that companies such as Apple, Amazon, Microsoft, eBay and Zynga were repriced at $123.47.
The Bloomberg data terminal listed either “market wide circuit breaker halt — level 2” or “volatility trading pause” on all the stocks affected.
And then, towards the end of the article, we read this:
In a statement, Nasdaq said the glitch was related to “improper use of test data” sent out to third party data providers, and said it was working to “ensure a prompt resolution of this matter”. In cases of any clearly erroneous data, trades made are cancelled. (All emphases in the original)
Note the implicit assumption: the "safety valve" or "Market wide circuit breaker halt" gets tripped when volatility reaches a certain level. That should trigger the question that some former quants want answered, but it appears not to be "getting through": is volality itself a certain enough epistemological foundation on which to assume that algorithmicly-driven trades are not altogether divorced from human market realities? In other words, does the mere absence of such phenomena on a trading day mean that real human market conditions are being reflected? I would submit that this is not so.
But there's something else here, and it should give everyone considerable pause: these "glitches" appear to be happening with no end in sight, and therefore, how does voiding these trades contribute to the healthiness of our markets, if one has to "lose time" and a potentially whole day of trades to "reset" after a "glitch"? Ultimately, this would seem to be a very inefficient way to conduct trades. However what disturbs me is the constant appeal, on such events, to the ever-present "glitch." We've had several such "glitches" since the famous May 2010 flash crash. One might as well say that these events are due to "gremlins in the system"; such an assertion would possess just as much rational explanatory power as the ever-useful "glitch," and this is a key to the glaring problem once again: our markets are not transparent, and not reflective of genuine human market conditions. I begin to think that, perhaps, "gremlins in the system" might even be a more useful explanation, for that at least would seem to imply that perhaps some of these trading networks have "woken up."
Whether or not that high octane speculation be true, there are a couple of other problems here that greatly puzzle and disturb me. Who got to decide, after the "reset", that various stocks were repriced at $123.47 per share? On what basis was this decision taken, or did it, too, fall within some agreed-upon artificial protocol adopted in the case of "how to reset market prices in the event of a flash crash?" Who (or what) took this decision? What was the explanation?
Finally, this new "glitch" was, according to Nasdaq, related to "improper use of test data sent out to third party data providers."
Ok... what was the test data? and WHO were the third parties? Was their "improper use" of that data intentional or not? It's the general lack of any useful information here that once again states, as clearly as possible, that the market is not transparent, and that it ill-reflects any human market realities or analysis of the value of the shares affected. Did that "improper use" of "test data" constitute an attempt to deliberately probe Nasdaq's potential cyber vulnerabilities? And if one can "improperly use test data" to create a "glitch" that leads to suspension of trading, why not use it to erase all records of trades altogether? Of course, I can hear it now: "Oh we have backup systems and there's no danger of that," and etc &c usw. and k.t.l. All of this is a nice euphemistic way of saying that the markets have little to no integrity.
I don't know about you, but every intuitive bone I have says there's much more to this story, and to the whole phenomenon of flash crashes, than meets the eye, and that "they" don't want to talk about it, either because they know what's really going on, or - worse - don't really know what's going on. Why do I get the queasy feeling that this is all related to the Inslaw scandal and the theft of its PROMIS software?
See you on the flip side...