Everything You Thought You Knew About Spoofing is Wrong!
(Except that it is illegal. It is.)
The topic of spoofing, particularly in the futures markets, has come up again as of late. On January 29th, 2018, the CFTC has filed anti-spoofing enforcement actions against eight market participants. Michael Coscia is serving a three year prison sentence for spoofing from a 2013 action (2016 sentencing). The CFTC sought to regulate and punish spoofing prior to Dodd-Frank, but the 2010 law made it punishable as a crime. Spoofing is clearly illegal, but should it be?
Also the campaign against spoofing has been entering the crypto market, where I have been spending my time & effort. It is worth noting that regardless of opinion, spoofing IS illegal so don’t do it. My purpose in writing this is to show how spoofing not only should be legal but plays an important part in the trading ecosystem. Clearly a part with a poor marketing department.
The mechanics of spoofing have been covered elsewhere, but an explanation is necessary to properly discuss the subject. The goal of spoofing is to get the “market” to fill a small order on one side of the market by posting a much larger sized order on the opposite side of the market. For example, a spoofer wishes to purchase 3 lots of crude oil at 66.71 when the current market looks like:
In order to purchase the contracts, the trader enters a 66.71 bid for 3 lots and at the same time offers 350 contracts at 66.72:
The “market” sees the outsized offer, the trader is filled at 66.71, and the 350-lot spoof offer is canceled. Presumably, the sellers are now forced to cover and the market moves higher. Voila! Manipulation writ large.
This is a clear example of how spoofing works but for a prosecuting CFTC, it may be hard to determine intent which is a requirement of the law. Perhaps the original intent was to sell 350 lots at 66.72 and then, due to something on Twitter, the trader became “tentatively” bullish.
Regardless of intent it is important to note that the 350 lot offer is live. That means that if a large trader decided to purchase all 354 lots at 66.72 when the “spoof” was on offer, then the spoofer would be short 350 lots. Further, in order to post the trade at the exchange, it is necessary to have the proper margin and trading permissions to enter a 350 lot order. In other words, offering a bunch of contracts is a significant risk.
Is there a parallel outside of finance where a strategy is to indicate a false intent while incurring large risk relative to a potential reward? There is: bluffing. The purpose of bluffing in poker is to bet in such a way, often by betting a large amount, as to influence other players at the table to fold despite the bluffer having a weak hand. It should appear as though the bluffer has a great hand and is excited enough to bet large. The desired effect would be for the other players to be intimidated by the large bet even if the folding player has a better hand than the bluffer. Thus, it works exactly the same as spoofing where intimidated traders rush to avoid oncoming doom.
For both spoofing and bluffing, the idea is to “manipulate” others into action. “That’s terrible! That’s manipulation!” It certainly would feel that way if we got bluffed out of a 3 of a kind! Imagine for a moment, though, if we hosted a poker game and created a house rule that bluffing was not allowed. “No problem,” you say, “I don’t bluff.” Perhaps you do not. Perhaps I do not. Up until then, maybe we don’t pay much attention to who is betting how. You and I focus on our own cards and how we want to bet on them. Since I’m terrible at reading faces, I can’t figure out who is bluffing when. Except that now I know no one is bluffing because it is not allowed. That means that I now better pay attention to the betting because the only choice that all players have is to bet according to how they truly feel about their hand. Suddenly, my strategy of “playing it straight” is vulnerable because everyone at the table will know when I have a good hand. In fancy pants words, the information content of the betting has increased dramatically.
Returning to the futures markets, with spoofing outlawed, the information content of the order book has just increased dramatically. Suddenly, the market knows that there are (presumably) no fake orders in the order book. That puts large, institutional “real money” orders at risk. Now, that 350 lot in the order book as above is known to be a real order and the market will have greater confidence in the impact of orders and the information content of the order book. It becomes pretty clear, then, that outlawing spoofing makes it harder for real orders, especially large orders, to get filled efficiently. It is as if the river was drained and the fishermen can see where all the fish are.
Furthermore, consider the mechanics of what makes spoofing work. The spoofer puts in the 350 lot offer but it is not the “market” that is filling the 66.71 bid. Actual individuals or algorithms have to decide and then act to sell crude at 66.71 based on the available information: see the 350 lot offer at 66.72, decide (in this case wrongly) that there is selling pressure, and then hit the 66.71 bid. It is these traders that are seeking to front run what they perceive as a real order. Which is the manipulation: the spoofing or the front-running? Before you answer “both!”, understand that pragmatically, once you outlaw spoofing, market participants know that the 350 lot would be real. Essentially it boils down to the uncertainty of the situation puts both the spoofer and the front-runner in an uncertain situation — but the real orders benefit from the camouflaging and the churning that results. In other words, it becomes more advantageous to front-run big orders without spoofers because you know the orders are real. Spoofing functioned to keep the front-runners more honest.
Why is spoofing successful? It is successful not because of the spoofer, but because of market participants that are looking to prey on large orders. There are, essentially, two kinds of traders that try to take advantage of large orders. The first are front-runners anticipating an order imbalance to come as the large order sweeps the market one way or the other. The second are those entering orders and choose to take a more passive stance given the presence of a large order. Consider the crude oil market above and a market maker who just sold some futures in a different month. Rather than simply lifting the offer, the trader joins the bid. It may even be a market maker in a related market (say crude oil products like RBOB or crude options) that change their market based on the 350 lot in crude oil. It would seem then, that market makers and front-runner have the most to gain by having an order book with high information content.
No one, of course, calls their business “front-running”. They use descriptions that sound mathematical like “reading the order book”. Of course, there are market makers and if my hypotheses are correct, then there ought to be an effort by market makers (and “order book readers”) to move against spoofing. It turns out that there is: it is exactly these participants who have coaxed the CFTC into being so aggressive about prosecuting spoofing. These organizations have been assisting the CFTC in their efforts. For example, in a case against 3Red in 2016, Citadel supplied an expert witness, who noticed Citadel’s “ES performance deteriorate” (ES is the S&P 500 mini futures contract). Source The analyst is Richard May, who went on “We always want to trade on orders we put in the market, but we cannot do so profitably when the order book does not reflect genuine interest.” In other words, they noted that spoofing deteriorates the information content of the order book obfuscating true intent and making it hard for them to trade (I’d argue it makes it harder for them to front-run real orders). It seems odd to me that the CFTC or anyone else cares whether 3Red or Citadel trades better (although the evidence seems fairly clear in that regard) or makes money. Citadel, CGTA, and other complainants (a group with HTG was also noted in the WSJ) would like real money accounts to make their trading intentions known so that Citadel and HTG can maximize their profits front-running them.
Allow me to note that this is not classic front-running where one is literally holding an order from a customer and then trades ahead. This, like so many things in finance these days, is probabilistic front-running. Spoofing makes the probabilities line up so much better.
Based on the information from Citadel, HTG, and others as testified, the spoofing law is designed to allow front-running by firms that have trouble trading without knowing which orders are real and which are not. Michael Coscia does not come across as a sympathetic character but he ought not to be serving prison time. It should not be worth noting, but probably is anyway, that spoofing is not my trading style (it replicates selling an extremely short term option and I tend toward asymmetric payoffs the other way).
The legal document that includes reference to the affidavit from Citadel and CGTA can be found here.
As I close this note, I want to make clear that I am not opposed to regulation. As distasteful as spoofing or bluffing may be (or may not be), regulation should work toward markets that provide greater efficiency for what may be termed “real money” investors/traders. The games of short term traders and market makers should be allowed if and only if they serve the goals of the system overall in a fashion that is fair. Fair means here that the orders entered are live and at risk in the same way as any other plain limit order. By those criteria, spoofing does serve the interest of the market because it camouflages orders of, as Citadel puts it, “genuine interest”. Therefore, while spoofing is currently illegal, I hope that this helps you see spoofing as, actually, an essential part of a functioning market.
