As readers will recall, yesterday we highlighted the HFT-driven flash crash in Core Molding (CMT) after an algo very obviously went insane and took the stock price to $0.0001 in the span of one second. Today, courtesy of our friends at Nanex, we get a transposition of the events at 14:19 in all their visual glory, together with a succinct explanation of everything that went wrong.
Jeffrey Donovan from Nanex writes in:
Here's the jist of it:
That chart was run on my Price Sequencer. You can see, approx 14:19:07 bats start stubbing the bid..... a second later NSDQ follows it down to 0.01 and takes the NBBO along with it! Once down there the NSDQ algo goes into repeater mode, cycling the bid/ask sizes as you can see.
I also took out everything except NSDQ, PACF and BATS and plotted them as dots to see just where it all occurred:
and then again as lines:
That last one I took the lower scale to -1.0 just for a slightly better visual.
So that's it. Stub quoting takes the NBBO to zero and where the NBBO goes.......
So there you go: no NYSE Liquidity Replenishment Points to blame here, but a pure and simple example of an algorithm going haywire. We wish to thank BATS exchange (which is no longer content with busting the NBBOs in the US, it has now also decided to go after Chi-X in Europe) for allowing this vivid demonstration of a mini flash crash in action, in which there is absolutely no fat finger, no Greek parliament to blame the crash on, and is merely a function of a busted HFT algo which ends up making a mockery of the price discovery process. Our market continues to be extremely fragmented, poised for an increasing number of micro flash crashes either in one name, or multiple, and with the SEC continuing to dither and take no action to remedy the broken topology and microstructure of the market, we can be certain that we will see many more such examples in the future. Only this time everyone will know that just like in this isolated case, it will be the HFTs which have completely destroyed the credibility of an efficient market.