In the wake of yesterday's market bloodshed, the NASDAQ announced that they'd be canceling trades more than 60% off of prevailing last sale before things went haywire. Now, I addressed this in last night's post, and others are writing about it this morning.
First of all, back to a solid explanation of what happened, courtesy of, believe it or not, a Georgetown Finance professor, James Angel:
“The LRP model doesn’t work,” Angel said. “The idea that when the market is going crazy you can slow down trading in one market and not others means that sell orders were churning through the books at every other market. NYSE dropped out of the running.”
So when the NYSE tries to slow things down, Reg NMS requires orders to be rerouted to other market centers. A commenter on my previous post explained it thusly:
" I believe the core problem here was a real order imbalance with lots of volume (which took us down the original 350), then NYSE halts these stocks, the market orders get rereouted (REG NMS and all) to ECN's where there is much less liquidity, and what liquidity is normally there is mostly provided by the nefarious HF strats who were rightly scaling back risk. So the 10K shares that might have been a downtick on NYSE blows through the BATS book completely."
Market orders can do a lot of damage to thin order books. Folks - if you're using market orders, you had better be aware of the risks. Interestingly, the NYSE, who has been losing market share to other market centers as a result of the ongoing market fragmentation, is positioning their positive spin on the crash, basically blaming the other market centers for failing to stem the cascade.
Paul Kedrosky asks aloud: "why are we wiping out all the errant trades by runaway algorithms and market battle bots?"
David Merkel points out, emphasis mine: "NASDAQ should not have canceled the trades. It ruins the incentives of market actors during a panic. Set your programs so that they don’t so stupid things. Don’t give them the idea that if they do something really stupid, there will be a do-over."
And the Law of Unintended Consequences rears its ugly head again. Merkel's point is simple and accurate: if buyers who step in later see their trades canceled, it removes all incentive for them to step in - and then you don't get the bounce back that we saw! Think about how much havoc it causes a trader who astutely bought cheap stock, then sold it out at a profit. He's now short! Or, he spent the entire day wondering if his order would be canceled, in a state of limbo. What's the alternative - that traders should just assume that the orders will get canceled, and NOT buy stock? Guess what - if no one buys, the stock stays cheap! SOMEONE has to buy, and that someone shouldn't be penalized in favor of remedying the ignorance of the seller who screwed up.
I commented on both Kedrosky's and Merkel's blogs that the goal wasn't to protect the battle bots, it was to protect the retail investors who screwed up because they used the wrong order type (although that doesn't make it right - as I mentioned in yesterday's post too). Again, we see the "bailout" theme - bailing out those who screwed up and preventing them from bearing the consequences of their bad decisions. In case it's not clear, of course I don't think that algos gone wild should be bailed out either via canceled trades.
If anyone wants to defend the decision to cancel the trades, I'm all ears - but your argument needs to be better than "HEY ITS NOT FAIR THE COMPUTERS RIPPED ME OFF AND MY STOP ORDER GOT EXECUTED AT A PENNY WTF OMG *$XYS !^^!@&!*#" If you don't understand that this can happen with a stop order, don't use stop orders.