Is there any practical and academic interest in predicting which orders in a limit order book will be canceled?

From a policy point of view are people interested in detecting potential spoofing within the market? or would the SEC or similar be interested in detecting anomalies with canceled limited orders? i.e. is there interest academically if somebody were to develop a model to detect which orders are likely to be canceled?

From a practitioners point of view are HFT firms interested in detecting if the sell side of an order book is overinflated with potential orders which will be canceled.

I know spoofing is illegal but does it still happen? Are there any papers out there trying to detect spoofing or canceled orders?

I am more interested in this from an academic point but its also interesting to me from a practitioners point also.


This is somewhat of an opinion based response with some factual anecdotes at the end.

Spoofing is a very difficult concept to define, identify and prove. In voice brokered markets where transactions are executed with levels of discretion, spoofing can be more readily identified by people refusing to commit to trades with insufficient reasons of cancellation. On exchanges one cannot refuse to trade if a transaction has been electronically registered, so the notion of bidding/offering a price represents a commitment at that point in time.

Whether there is a desire or not to physically transact is subjective and can easily be argued, and whether the intent of the price order is to manipulate the market is also very difficult to characterise. Neither of those necessarily come without calculated risks and cost to the dealer.

It is my own personal opinion that where a market exists that only allows firm prices to be shown and for any other participant to respond to those prices everything is fair. In practical terms the organisers of the exchange have an interest in maintaining and orderly and stable market.

Around 2011/2012 I think Short Sterling futures (3M LIBOR futures in GBP) decreased their bid/offer tick size on the exchange from 1bp to 0.5bp. Liquidity, as a result dropped (the number of participants happy to market make a reduced spread fell) and the ability to visibly impact the market by trying to transact sizeable quantities increased. Whether this was characterised by spoofers trying to manipulate Short Sterling in order to profit in other instruments, or by large dealers unable to execute similar volumes to before it is unclear (probably both). But the result was a worse market characterised in terms of anecdotal volumes and volatility. The exchange reversed the change. Whilst I cannot attest to the notion of spoofing directly it seems there must be some level of statistics, particularly user data from an exchange point of view that characterises certain type of behaviours that are either conductive to the overall functioning of a market, or are not.

  • $\begingroup$ Thanks for your answer! Do you think it is in the interest of exchange commissions to know the probability that an order will not be filled? If the sell side of the order book is filled up with sell orders (due to some uncertainty) and many orders will most likely be canceled at some point in the day (if the uncertainty goes away) would an exchange commission be interested, since it affects the supply of an asset in the market? I mean knowing that an order of a stock will be canceled with specific characteristics (time spent in the order book, time of purchase etc.) - is this interesting? $\endgroup$ – user113156 Oct 31 '18 at 20:50
  • $\begingroup$ I believe it is already being analysed. See more info here why quant.stackexchange.com/questions/40130/… . Exchanges want to maximise their profits, indirectly by generating as much liquidity as possible through consistent, active and reliable markets. Though this is not easy to guage, I imagine. $\endgroup$ – Attack68 Oct 31 '18 at 20:55
  • $\begingroup$ Recent events surrounding JPM suggest spoofing may not be as difficult to identify as this answer says $\endgroup$ – Kch Nov 9 '18 at 14:30
  • $\begingroup$ @Kch I think those events just suggest it is not impossible to catch one of the largest players who have been systematically doing it for many years. It says nothing about the success of the justice department in identifying less obvious persons from within a set of unknown size who engage, knowingly in the activity. $\endgroup$ – Attack68 Nov 9 '18 at 15:42
  • $\begingroup$ While the first part of that may be the case, it shows, one, that regulators do not take the opinion that spoofing on an exchange is mitigated by the fact that an order represents a commitment and, two, trade surveillance happens and fraud eventually does get caught. $\endgroup$ – Kch Nov 10 '18 at 18:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.