We will all remember that Navinder Singh Sarao has been elected Official Flash Crash Scapegoat by the US Department of Justice.
But the case continues to bring more questions than it purports to answer – even Flash Boys’ author Michael Lewis shows a grudging admiration for the man:
The people at the CFTC who decided to come forth, five years after the fact, with this new and improved explanation for the flash crash, must have known they would be creating a controversy with themselves at the center of it. It’s actually sort of brave of them.
They’ve been ridiculed in the news media and will no doubt soon be hauled before various congressional committees. They’ll have annoyed their colleagues at the Securities and Exchange Commission, who now look like even greater fools than they did before, for not bothering to mention in their report on the crash the various nefarious activities of algorithmic traders, and instead offering up as the primary cause of the crash a stupid mistake made by a money manager in Kansas. The authors of the SEC report either consciously ignored or did not bother to acquire from the CFTC a lot of accessible, and damning, information about what was happening in the U.S. stock markets the day of the flash crash. The world will now want to know why they did this.
Then there is the biggest question of all: How can a guy working from his parents’ house in suburban England whose only actionable orders were to BUY stock market futures cause such a sensational collapse in U.S. stocks? On the day of the flash crash, Sarao never actually sold stocks. He was trying to trick the market into falling so that he could buy in more cheaply. But whom did he fool with his trick? Whose algorithms were so easily gamed that they responded to phony sell orders by creating a crash? Stupidity isn’t a crime. Still, it would be interesting to know who, at this particular poker table, on this particular day, was the fool.
It would also be interesting to know how it occurred to Sarao that his trick might work. There’s a fabulous yet-to-be-told story here, about a smart kid in the U.K. who somehow figures out that the machines that execute the stock market trades of others might be gamed — and so he games them.
Pursuant to an offer of settlement in which Jonathan Brims (“Brims”) neither admitted nor denied the rule violations upon which the penalty is based, on January 20, 2015, a Panel of the CBOT Business Conduct Committee (“BCC” or “Panel”) found that it had jurisdiction over Brims pursuant to CBOT Rules 400 and 402 as the conduct occurred while Brims was an employee of a CBOT member firm. The Panel also found that during the time period from September 2011 through December 2012, Brims, on multiple occasions, entered large orders in the 5-Year Note, 10-Year Note, Bond, and Ultra Bond futures contract markets without the intent to trade. The Panel specifically found that Brims placed a small order to sell (buy) in the futures contract market on the CME Globex electronic trading platform (“Globex”). Brims subsequently entered multiple large-lot buy (sell) orders at or near the best bid (offer) on Globex to create the appearance of an imbalance in buy/sell pressure. Once the small order began trading, Brims canceled the large orders. The Panel further found that Brims entered the large orders for the purpose of inducing other market participants to trade against the small orders resting on the other side of the order book. The Panel concluded that Brims thereby violated CBOT Rules 432.B.2. and 432.Q.
Interestingly, Sarao provided information about (some of!) his algorithms to UK regulators, including a fascinating allegation:
I have traded using a basic TT for numerous years. Due to the fact that there were some individuals in the emini S&P who quite remarkably seemed to know WHERE 100% OF MY ORDERS WERE RESTING, even if they were over 90% partially filled !!! and hence made a concentrated effort to manipulate around those orders so they would not get filled, I decided to pay Edge Financial to build a program for me that would help disguise my orders more effectively. Initially I was told that the reason these individuals knew where all my orders were was because I traded so big and was as such ‘the elephant In the room’. However, It Is worth noting that further examination showed that their special manipulative activity occurred exactly the same if I did a 20 lot order or a 200 lot order.
I asked Edge to design 3 more functions specifically to help try and hide my orders from these people. I do not know If this can be described as HFT, to me It Is just giving me the ability to have some extra functions that my base trading software (TT) does not give me and It should be noted that I only use these functions Intermittently
and sometimes not at all. It Is called Navtrader, but it could be called anything and I was the only one who helped design it, albeit my design Ideas were 100% generated from what I had already seen other traders using already in the emini SP. Please note I believe I have only had this NavTrader since the beginning of 2013 at the very earliest
I decided that the only way I could mask my orders, was to place them as the market changed price so that they may not be seen In the ‘chaos’ of a price change. So I would have my orders pending to be placed as the market went from bid to offer or offer to bid.
The 3 main functions are as follows:
JOIN : These are pending orders that will be joined anywhere requested along the order book and become active when the price changed, Remarkably, these orders were still subject to the Insider trading I describe above, even when they are as small as a 50 lot !
SNAP: These are orders that are the same as JOIN but at the market best price ao that they become traded almost Immediately. I also have a function that lets you put In a minimum quantity so that the buy/sell SNAP order only becomes active when there Is a minimum of that number of contracts on the offer/bid. This worked rather beautifully when the mass manipulator of the e-mini sp was doing his normal manipulative activity at price 1800.00 on Friday 24th January circa 12.23pm. The fake bids he had placed were being removed too quickly for me to hit. If I had put a snap for 700 with 0 as minimum volume , It would not have been filled because as soon the bid was more than 1 lot bid the 700 would have been active. With my 699 then resting the normal forms of manipulation that occur on 100% of my orders EXCLUSIVELY would then have preceded to follow. So I put a 700 lot SNAP with a minimum volume of 600, et voila I got my full 700.
ICE: The Iceberg function on the CME Isn’t adequate for me, I hardly ever use it because It puts me at the back of the queue all the time. Hence, 2,000 needs to trade to get me out of 800 lots for example. My iceberg function is placed at a price and as soon as It Is bid/offered at the price the iceberg will take all contracts at the price up to and Including the number of my order. Again, there Is a minimum volume box, so for example I can put 50 Into it and put a sell ICE of 1,000 and then at that price every time the bid Is more than 50, the ICE will take all contracts out until 1,000 Is traded. This Is a good way of catching spoofers, and et voila I can trade 1.000 lots at one price (following on from the above example).
The other orders I sometimes place during the day are slightly away from the market price and move up and down as the market moves with It This Is to catch any blips up/down In the market so that I can make a small profit as the market comes back Into line(almost Immediately). These orders are placed rarely and only when I believe the market Is excessively weak or strong. Again, this was Inspired by other traders I could see doing the exact same thing.
Well … Assiduous Readers will be sympathetic, I hope, to my readiness to believe that one guy working alone with some occasional contract help can make the big guys look like fools. But these three functions seem perfectly straightforward to me and it will be most interesting to learn whether Sarao did in fact develop and use them and if they did in fact work as well as claimed.
Hunsader has produced a chart showing the January 24 spoofing/counter-spoofing battle:
On the other hand, of course, there are some pretty damning affidavits; it seems clear that:
- Sarao was spoofing
- He was also engaged in anti-spoofing
- spoofing rules are unenforceable and should be scrapped