Spoof, Bluff, Go For It: A Defence of Spoofing

Abstract

Spoofing—placing orders on financial exchanges intending to withdraw them prior to execution—is widely legally prohibited. I argue instead on two main grounds that spoofing should be permitted and legalised. The first is that spoofing as a form of bluffing remains within the market practice of making legally binding offers—as opposed to lying or betraying trust—and primarily concerns the spoofer’s personal information. As a form of bluffing spoofing helps prevent financial speculators, in particular high-frequency algorithmic traders, from easily profiting by other market actors reliably revealing their underlying preferences through their market activity. The second is that at the systemic level permitting spoofing would benefit non-speculative actors who place orders to hedge economic risk and whose activities provide the raison d’être for financial exchanges, differentiating them from simple forums for gambling. I also address potential concerns that legalised spoofing would drive speculators out of financial markets entirely and, therefore, undermine market liquidity. This work contributes to the wider debate in business ethics regarding bluffing by illustrating the acceptability of bluffs which do not betray counterparty trust or reliance on testimony by remaining within the framework of market practices.

Introduction

In late September 2020, JPMorgan Chase agreed to pay regulators USD 920 million as part of a settlement admitting to spoofing precious metals futures and US government bonds (Shubber & Stafford, 2020). Spoofing is the practice of placing orders to trade on financial exchange markets with the intention of avoiding their completion by withdrawing them before they are accepted.¹ JP Morgan’s penalty represented the largest fine ever issued for spoofing. It prompted the chair of the Commodities Futures Trading Commission, Heath Tarbert, to state, “Spoofing is illegal, pure and simple… Attempts to manipulate our markets won’t be tolerated.” (Shubber & Stafford, 2020) My argument is that while spoofing may currently be illegal in many jurisdictions, the reasoning behind this ban is neither pure nor simple. Instead, I argue that spoofing should be permitted and legalised. Just as bluffing serves an important purpose in the game of poker, spoofing—were it to be legally permitted—should serve an important purpose in the much more serious practice of financial exchange markets.

The comparison of spoofing to bluffing highlights the important fact that spoofed orders on financial exchanges can be ‘called’ like any other bluff. All market participants, including spoofers, are obliged to complete their orders if they are accepted by counterparties prior to withdrawing them. Conversely, all market participants are free to withdraw their orders prior to acceptance. What distinguishes spoofers and their orders is that from the outset spoofers intend to withdraw their orders prior to anybody accepting them. This can be contrasted with other market participants who initially intend their orders to be completed but then change their minds and withdraw them as market conditions shift. Prima facie spoofing has the air of disreputability about it—why place an offer on an exchange if one does not want to execute a transaction? This impression is reinforced by the term often used to describe the aim of spoofing—‘manipulation’ (Angel & McCabe, 2013, p. 590; Cartea et al., 2020; Shubber & Stafford, 2020; FCA 2021, MAR1.6). The term suggests that the spoofer is exercising some sort of devious means to control the market and in doing so harming legitimate market activities. A further implication of the term ‘manipulation’ is that a non-spoofed market is somehow more legitimate insofar as its prices are untainted by strategic deception or misdirection. This article rejects these implications. Instead, I argue that the prices on such markets, particularly at the high-speed second-by-second levels spoofing works at, are the result of actors strategically seeking to maximise their returns from one another in response to the rules of any given exchange. Restricting spoofing is one potential variable to these rules but permitting spoofing would not fundamentally change the nature of the strategic interactions which underpin price determination.

This article makes the case for the permissibility of spoofing on two fronts. The first is relational—the actors involved in financial markets do not owe one another any duty to reliably signal or disclose the preferences or intentions which underly their market activities. In fact, carefully managing the information one discloses regarding one’s true plans is a key element of market practice and may result in other market participants being misled as to one’s preferences. This is illustrated through legally permitted practices such as ‘iceberging’—the practice of breaking up one’s intended large orders into multiple smaller orders to disguise one’s preferences for significant demand. As part of this argument I will distinguish spoofing from simply disseminating misinformation, a view which has recently been argued for by Hersch, (2020, p. 218). A spoofed bid or ask is importantly distinct from simply lying about an offer’s existence or misinforming others as to the state of the market, and this is for the simple reason that spoofs, like bluffs, can be ‘called’. Notwithstanding the underlying preferences of a spoofer, the offers they make on the market can be accepted by counterparties and become legally binding before they are withdrawn. In this way their actions do not differ from non-spoofed orders—which are also frequently withdrawn prior to execution—both are capable of being accepted and becoming legally binding contracts. The deception, like any bluff, comes from what can be inferred from their activities about their subjective mental states.

The second reason is systemic—that financial markets would not be ‘harmed’ in any normatively meaningful sense by allowing spoofing. Instead it would plausibly help shift the balance of market power towards actors who, often because of economic interests in the goods underlying their trades, use financial markets to hedge risk. Spoofing would make it harder for certain types of high-frequency algorithmic traders (HFATs) to make profits. HFATs who rely on changes in order flow to infer the preferences of other market participants would find such signals harder to interpret were they to be mixed in with spoofed signals. However, an inability of HFATs to anticipate market movements by adjusting prices quickest in response to perceived shifts in supply and demand is not necessarily detrimental to markets as a whole (Pine, 2018). The uncertainty faced by these HFATs would be to the benefit of the large economic traders they seek to profit from. This article argues that relaxing anti-spoofing regulations would move the balance of market power away from anticipatory trading by HFATs who rely on rapidly reacting to perceived supply and demand shifts. This change would be to the benefit of market actors with underlying economic interests in orders. Defenders of the anti-spoofing status quo could argue that without such protections HFATs might withdraw liquidity from financial markets, thus, harming economic actors. However, this is unlikely to occur in any significant macro sense as HFATs have many other ways to profit from trading in financial markets, even if the possibility of spoofing renders some of their strategies more uncertain and less profitable.

The remainder of this article proceeds as follows. I outline a basic example of spoofing and explain how it aims to profit by countering a type of anticipatory trading strategy which rapidly responds to apparent shifts in market demand. I classify two basic actors in the markets in which spoofing occurs—economic hedgers and financial speculators—and note how speculators, in particular HFATs, aim to profit by anticipating price movements caused by—and at the expense of—hedgers. I then go on to explain that the context and medium of spoofing matters—it takes the form of recognised market activity, making legally binding offers. In doing so it invites trust from counterparties only insofar as spoofers are committed to completing said orders should they be accepted. When counterparties try to infer intentions behind such orders they are trying to discern the agent’s own proprietary information—their purchasing preferences—which such counterparties are not entitled to. I explain how this justifies spoofing as a form of defensive deception to obfuscate such preferences, illustrating this with the analogy of bluffing in the game of poker. Building on the foregoing I end by sketching an outline of spoofing-enabled markets. I argue that enabling spoofing would work to the benefit of hedgers who provide the raison d’être for such financial exchange markets. I address concerns that such a change would damage market efficiency or drive HFATs out of business. I then conclude with a summary of the argument and note the wider applications of this analysis beyond spoofing to bluffing in the wider business ethics literature.

Spoofing, Information Asymmetry, and Markets

Spoofing is the practice of placing orders (either bids or asks) on a financial exchange market without the intention of having the orders fulfilled, in particular by cancelling them prior to execution (7 U.S.C.A. §6c(a)(5)(C)).² The most common usage for spoofing in past identified cases has been to cause prices to change in response to spoofed orders and to profit from the resulting shifts. The means by which spoofers have sought to profit from this activity have varied (Dalko et al., 2020). A common strategy is known as ‘layering’ (Mark, 2019) and characterises a number of the more famous spoofing cases such as that involving Michael Coscia (FCA 2013) and the infamous ‘Hound of Hounslow’ trader Navinder Singh Sarao (CFTC 2015; Vaughan, 2020). Layering consists of placing a series of orders going the same way, usually at some distance from the prevailing best price and often quite sizeable. These ‘layered’ orders are intended to go unfulfilled and to instead present an impression of demand on that side of the order book. This series of orders is accompanied by smaller orders placed on the other side of the order book. These small orders are intended to be executed. The strategy can be profitable because the ‘layered’ orders are significant enough to create a shift in the perceived balance of supply and demand, which causes an overall shift in the prevailing best price in favour of the small orders intended for execution. A toy example might be helpful.

Table 1 below represents the market prior to any layering activity. The ‘contracts’ columns represent the volume of individual contracts that constitute the buy and sell orders on the market. The rows represent the prices at which these contracts are being offered to be bought or sold at.

Table 1 Example market prior to layering activity

Now imagine that the layering party places a large number of sell orders at 1.001 and 1.002, respectively. This signals to the market that this is becoming a ‘buyer’s market’—with more market participants interested in selling than in buying. This means that sellers will have to accept lower prices in order to have their orders fulfilled, resulting in a market which may resemble something like Table 2.

Table 2 Example market after layering activity

The resulting price movement has established a new prevailing best price for selling—at 0.999 instead of 1.000. The party engaged in layering can now execute buy orders to purchase at 0.999 and cancel their sell orders at 1.001 and 1.002 which precipitated the price movement. The outcome is a cheaper purchase price for the layering party. In order to realise their profit they might execute the same activities in reverse, layering large buy orders in order to execute a sell order for an improved price.

As illustrated by the above example the impact of spoofing derives from the simple economic fact that prevailing market conditions—the prices people are willing to buy and sell at—are responsive to perceptions of supply and demand. Observing shifts in buy and sell orders provides information market participants use to inform their own behaviour and willingness to trade. This gives rise to one famous aspect of markets—the importance of information asymmetry (Akerlof, 1970; Rothschild & Stiglitz, 1976; Spence, 1973). An example of problematic information asymmetry is insider trading in public securities. In financial exchange markets a variant on insider trading is illegal front-running. This occurs when brokers, who execute buy and sell orders on behalf of clients, use knowledge of their clients’ market orders to trade on their own behalf. The brokers personally benefit by knowing in advance the effect on the market price their clients’ order will have. For example if their client wishes to place a large buy order the broker may purchase on their own account in advance, anticipating that their client’s buy order will cause prices to rise in response to the increased demand.³

HFATs work to legally exploit information asymmetry through anticipatory trading strategies. Anticipatory trading means trading strategies which take positions in advance of, intending to profit from, subsequent expected movements in prices.⁴ While anticipatory trading may be carried out by non-high-frequency traders (Fishe et al., 2019) the advantage of HFATs comes both from the predictive abilities of their algorithms and the speed with which they can react to new information such as new buyers or sellers entering the markets.⁵All actors in the market are concerned with what future prices will be and try to incorporate that information into their market activity. HFATs aim to do so faster and more accurately than their competitors by recognising patterns in market orders. In doing so they provide a service, they ensure that new information—the type their algorithms incorporate into buying and selling decisions—is rapidly, almost instantaneously, incorporated into prevailing prices in response to their activities.

If this was all there was to financial exchange markets they would resemble nothing more than a complex form of gambling. Actors seek to incorporate any and all kinds of information into their strategies, including the activities of their opposition, to inform their own betting strategies on the direction and magnitude of future price changes to maximise returns. However, there is an important distinction between these markets and gambling. This is that financial markets serve an important economic role (Hersch 2020, pp. 214–218). They are not solely a means of financial speculation from otherwise disinterested actors who would—if they had the requisite algorithms and knowledge—bet on horses as soon as they bet on oil futures. This economic role can be understood if we broadly divide market participants into two types: hedgers and speculators.⁶ The former are those market participants with an economic interest in the underlying assets being traded. For example consider a company which extracts and sells oil. The oil company might participate on an exchange buying or selling oil futures not to speculate on future prices per se, but to hedge and reduce their risk exposure.⁷ Instead of being subject to future market fluctuations in the price of oil the company can, through the use of futures contracts, effectively lock in a price for the oil they intend to sell in the future. They benefit from this arrangement by increased certainty over their future cashflows, which enables them to plan their business affairs accordingly. The risk of price decreases, or the upside of price increases, is then instead taken on by a counterparty in the market. The second category of actors are speculators, whose market activity is driven by the aim of profiting from the spread between purchasers and sellers over time. In simple terms they aim to buy low and sell high. Therefore, if they think that oil is underpriced at USD 50 for a future barrel delivered at a certain date they will buy that contract with the aim of selling it when the price appreciates. Conversely they will lose money if the date for delivery approaches and they need to sell the contract at a value below USD 50.⁸ The term speculation often has negative connotations but that is not the intention here. Speculators help provide liquidity to the market by being willing to take either side of trades so long as the price is ‘right’. There is a price to this liquidity as speculators will only take on trades they identify as profitable.

This price of liquidity helps explain one ironic feature of markets—that participants tend to make their transactions more expensive by trying to execute them. If our oil company wishes to sell a large number of contracts to deliver oil at a fixed price in the future then its fellow market participants and speculators will register its interest as increased supply-side pressure. If the oil company’s request forms a significant portion of the market in that specific contract it may be enough to cause the price to move against them, reducing the price at which others are willing to buy the contracts from them. This is part of how speculators maintain their ability to buy cheap and sell dear; they use the information that a company wishes to act in a certain way to adjust their prices against the company’s orders. This is also the basis of the illegal broker front-running described above. In both cases, actors are looking to adjust their buying and selling behaviour to accommodate relevant new information entering the market—that a participant wants to place a significant buy or sell order. The difference derives from the source of the information. In the illegal front-running case, the broker is exploiting their client’s information to turn a personal profit. They react to the new information before it even enters the market in the form of an order, to the detriment of both their client and their fellow market participants. In the legal case of anticipatory trading, the speculator must react to the information after it enters the market, in particular the HFATs aim to do so faster than their competitors. In the words of one former trader describing HFAT activity:

A front-runner profits by gleaning the intentions of legitimate market participants and jumping in front of their orders, thereby causing the original traders to buy and sell at a less favourable price. (Arnold, 2015)

Arnold here is using the term ‘front-running’ in a loose sense not to refer to illegal breaches of fiduciary duty by brokers, but to describe what will be referred to in this article as anticipatory trading. The ‘gleaning’ Arnold mentions does not derive from illegitimate sources of information like insider trading or illegal front-running. It is the orders themselves which betray the intentions of market participants and which enables HFATs to react to the new information so as to profit from said market participants’ revealed preferences.

What Market Participants do not owe to one Another

The methodology adopted in this section is influenced by the practice-dependent method developed within political theory (James, 2005; Jubb, 2016; Miller, 2002; Sangiovanni, 2008, 2016).⁹ The term practice here means, “…any form of activity specified by a system of rules which defines offices, roles, moves, penalties, defenses, and so on, and which gives the activity its structure” (Rawls, 1955, p. 20), in this case the practice of financial exchange markets.¹⁰ The practice-dependent method works by interpreting, making sense of, and critically assessing practices to develop normative rules to govern them (Sangiovanni, 2008, pp. 142–144). This position can be contrasted with practice-independent approaches which instead specify abstract higher order principles and subsequently applies them to specific situations, for example, deontological or consequentialist ethics. The practice-dependent method particularly focuses on the relationships created through practices. Features of these specific interactions provide the substantive material to ground and justify their normative regulation. In the words of Jubb, “Practice-dependence is in this sense a justificatory and epistemic strategy, a way of finding principles to govern a practice and drawing attention to the features that make them more appealing than their competitors as rules for that practice” (Jubb, 2016, p. 84). In applying this methodology, a researcher must, therefore, combine interpretative and critical roles, understanding an existing practice on its own terms while still maintaining sufficient analytical distance to identify ways in which it should be reformed or altered, as I propose here.¹¹

This approach is apposite given the question at hand—assessing spoofing within the practice of financial exchange markets sketched in the previous section. The relationship between market participants is readily specifiable in parsimonious terms through their competitive interactions on exchanges in making, withdrawing, or accepting orders to trade. This contrasts with the application of practice dependence within political theory where it is often used to derive accounts of either domestic (James, 2005; Miller, 2002; Ronzoni, 2012) or international (Ronzoni, 2009; Sangiovanni, 2008, 2016) justice. Arguably practice dependence is most appropriate in contexts such as those described in this article—where one can clearly delimit the scope and purpose of the practice in question. In this section I provide an account of the relationship between market participants engaged in the practice of financial exchange markets. I illustrate this account with analogous competitive situations such as poker and auctions to highlight why the nature of the relationship between market actors entails the permissibility of spoofing as a form of defensive deception.

There has been extensive commentary within business ethics on the general acceptability or not of bluffing or deception in market transactions, particularly negotiations (Allhoff, 2003; Carr, 1968; Carson, 1993; Strudler, 1995, 2005). What I argue here is that market participants do not owe it to one another to reveal their true preferences through their market activities. The case of illegal front-running illustrates that knowing what others intend is a recipe for profiting from them. Because other market participants are trying to discern, and profit from, the intentions behind your market activities your options are either to be taken advantage of or to obfuscate. If you choose to try and prevent your intentions from being deduced others will be deceived or misled as they nevertheless try and draw inferences from your activities. One example of this type of obfuscatory practice is known as ‘iceberging’. Icebering occurs when a market participant tries to disguise the information given away by their market activity by breaking down a large desired order into many smaller orders. This is by no means the only way in which hedgers might try to disguise the preferences behind their orders, all kinds of more complex strategies might be implemented. Arnold himself reveals that, “To limit the effect of front-running on my firm, I spent millions of dollars developing a proprietary order-entry system to disguise and conceal strategies from external algorithms.” (Arnold, 2015) Tactics such as iceberging and Arnold’s order-entry system help disguise the preferences of hedging market participants from speculators, and particularly from HFATs, who would otherwise take advantage of that information to profit from these hedgers. Just like in poker the important thing is to keep your cards close to your chest. Of course, just like in poker your counterparties are incentivised to do their best to pierce your informational defences and look for ‘tells’ or other inadvertent signals which betray your genuine state of mind in order to profit from it. What we have here is an arms race—hedgers who need to place large orders for economic purposes want to avoid giving this information away through their order activity to HFATs and other speculators who are seeking to profit from both hedgers’ activities as well as one another’s trading.¹²

In a recent article defending the legitimacy of spoofing Cooper et al. analogise iceberging and spoofing to suggest that although iceberging is deceptive, “…no one (not even high-frequency traders) considers iceberging to be misconduct, much less immoral. Taking large positions stealthily in order to conceal one’s true intentions is part of being a good trader.” (Cooper et al., 2016, p.15). In response to Cooper et al. Hersch decries spoofing as a type of offensive deception—as opposed to defensively protecting information the spoofer is under no obligation to divulge—and argues that spoofers are ‘offensively’ distorting information in the ‘public domain’ (Hersch 2020, pp. 213–214). Hersch goes on to raise as a possibility that if spoofing is unethical then perhaps so is iceberging (Hersch 2020, p. 214). In what follows I contribute to this ongoing debate by correcting Hersch and demonstrating that spoofing is best thought of as a legitimate defensive form of deception in that it only pertains to information which the spoofer is entitled to guard—their own preferences. In addition, the medium of the deception, placing a market order which in itself does not purport to invite trust beyond the fact that the order exists, further mitigates any question of the deception being problematically unethical.

The context of a given practice matters when considering the legitimacy of deceptive techniques. Part of the power of the poker bluffing business analogy made famous by Carr’s Harvard Business Review article (Carr, 1968) derives from the fact that poker would lack much of its purpose or strategy if bluffing were not permissible (Cooper et al., 2016, p. 2). Since Carr, there has been significant critique of the business as poker analogy (Solomon, 1992; Varelius, 2006; Radoilska, 2008; von Kriegstein, 2019; Sinnicks, 2021a). In particular, Koehn points out that the strength of arguments from analogy can only be as strong as the analogic relationship itself (Koehn, 1997). Therefore, any salient difference between a game of poker and the much more important realm of business can be used to pry such arguments apart. My argument here is not intended to rest on the validity of the analogy with the game of poker, its claims should stand even if one rejects financial exchange markets as being appropriately analogous to poker. However, I contend that the analogy can nevertheless be helpful here to aid understanding. Poker is the most common experience most people have of bluffing and there are a number of helpful similarities between how the practice functions in both the game of poker and in financial exchange markets in the form of spoofing.

Both poker and financial exchange markets are strategic competitions where profits can be extracted through parties discerning one another’s true intentions and strategies. Bluffs are designed to misrepresent the bluffer’s internal mental state and are conveyed through the generally accepted moves of the practice—offering to enter into transactions which involve putting money at risk. In the case of both poker bluffs and spoofed offers these can be ‘called’ by counterparties in which case the bluffer risks suffering a loss. The point here is that in both poker and financial exchange markets nobody can, or should, expect parties to willingly betray their internal preferences through their conduct. This is because in both contexts savvy opposition will take advantage of, and profit from, such naïveté. This is different from saying ‘anything goes’. There is an important distinction between a certain bidding strategy designed to confuse my opposition and, for example, straightforwardly lying to them. For example, in a poker game it is one thing to bluff someone into folding their hand, but it is another to lie and tell them a loved one has been in an accident to convince them to rush from the table and abandon their hand.

In this respect my argument is more limited than much of the business ethics literature applying the poker bluffing analogy to cases of verbal business negotiations (Carr, 1968; Carson, 1993; Strudler, 1995). These wider arguments have to rely on what Shiffrin terms ‘justified epistemic suspended contexts’ (Shiffrin, 2014, pp. 16–19), whereby facts about the situation reasonably and justifiably deprive listeners of the epistemic entitlement to presume that the speaker will tell the truth without deception. Shiffrin cites joking, fiction, plays, and devil’s advocacy as examples of such contexts (Shiffrin, 2014, p. 16). The difficulty for advocates of business bluffing then becomes explaining why business negotiations should qualify as such a context. For example, because negotiators understand certain statements to lack a warrant of truth (Carson, 1993) or because participants endorse bluffing (Allhoff, 2003). Critics reasonably point to differences between poker and business to explain why the analogy with poker can provide no support for this argument. For example they cite the types of goods at stake (Sinnicks, 2021a), the ability to change and develop the rules (Koehn, 1997), or the fact that bluffing is not necessarily endorsed by all business participants (Varelius, 2006; Hersch 2020).¹³ That said, my argument is narrower than the claim that bluffing is a part of business negotiations. In particular my argument does not rely on any kind of suspension of normal rules of truthtelling. As both Cooper et al., (2016, p. 9) and Hersch (2020, pp. 210–211) agree offers on financial exchanges are not truth claims which can be warranted. Therefore, unlike instances of bluffing in verbal negotiations, there is no need to argue for a suspended context.

The form of communication matters—an offer to enter into a contract only invites trust that if a counterparty accepts it then a legally binding transaction is mandated. While trust is widely considered to be a species of reliance (Goldberg, 2020, p. 98) it is important to distinguish trust from mere reliance (Baier, 1986, p. 234; Pettit 1995, p. 205; O’Neill, 2002, p. 15). Market participants trust because they believe that counterparties are committed to completing accepted offers and rely on this commitment through their market activities (Hawley, 2014, p. 10). This is distinct from relying on the fact of offers to reliably infer information about the offeror’s internal market preferences. We might come to rely on Kant’s regular walking habits to keep time ourselves, but this reliance falls short of normatively relevant trust (Goldberg, 2020, p. 101). The oft-cited example in the trust literature is that of the ‘confidence trickster’ who relies on their target acting in certain ways in order to profit from them (Holton, 1994, p. 65; Hawley, 2014, p. 12; Goldberg, 2020, p. 99; Mcleod 2020). It can be perfectly legitimate to subvert such reliance without endangering one’s trustworthiness, especially when such reliance is to your detriment. In situations such as scams, poker, and markets where the ability to read intentions behind actions is key to competitive advantage the only options are to act ‘sincerely’ revealing one’s true intentions, confusing attempts to read true intentions such as iceberging or spoofing, or simply not acting at all. Protecting one’s internal preferences from counterparties is logically linked to confusing one’s opposition who are trying to read your intentions. It is crucial for this argument that misled counterparties rely only on their interpretative abilities rather than being in a normatively important relationship of trust, as illustrated by the following example of auctions.

Imagine that I am in a competitive auction for two items and there is one other bidder. Also imagine that I have an important piece of personal information which I conceal from the other bidder, that I am far more interested in Lot 2 than Lot 1. My preferences are my own business, which the other bidder has no right to know. Say I reasonably believe that a certain bidding strategy on Lot 1—say coming in high and hard—will convince the other bidder that I have a serious interest in Lot 1. As a consequence, if I adopt this strategy the other bidder will seek to strongly outbid me on Lot 1, which if I let them win allows me to devote all of my cash reserves to acquiring Lot 2. This strategy relies on allowing the other bidder to be misled as to my personal preferences by my bidding activity. Nevertheless, it is unreasonable to argue that my bidding strategy creates a normative commitment on which the other bidder can reasonably place their trust (Carson, 1993). For example, imagine instead a scenario where I explicitly tell the other bidder that Lot 2 has intense personal value for me and, therefore, ask that they let me have it cheaply, whereas actually I suspect it has great resale value and it holds no personal interest for me whatsoever. Both strategies might lead me to acquire Lot 2 cheaper than I would have done by confounding the other bidder’s beliefs about my preferences. However directly lying to them is morally bad because I play on their sympathy and trust in order to extract concessions from them. I can expect to be deemed untrustworthy from that point on and moreover have harmed the reliability of the channels of communication upon which cooperation to achieve moral agency relies (Shiffrin, 2014; Sherwood, 2021, pp. 16–17). By contrast simply making large bids on Lot 1 is fully within the practice of auctions. Counterparties are not in any way morally entitled to pierce my inner intentions by watching my market behaviour, especially when doing so will be to my disbenefit. There is no wider harm to the practice of trust or reliance on testimony by this bidding strategy because someone placing an auction bid is not inviting trust or reliance from others beyond the fact that were their bid to win they would follow through and pay for what they have purchased. The fact that others try to garner private internal information from market offers is fundamentally their problem, not yours, so long as you are not actively lying to them and inviting their trust but simply engaging in the market practice of making bids (Strudler, 2010, p. 175).

None of this is to say that spoofed offers cannot mislead or cause others to believe falsehoods. As discussed above, HFATs and other speculators rely on inferring underlying preferences from market behaviour in order to profit by predicting them. Techniques such as iceberging and spoofing are designed to confound these inferences, and so by their nature are misleading. Hersch, therefore, argues that spoofers harm the market by going beyond the proper scope of the practice of market trading (Hersch 2020, pp. 213–214). In particular, he argues that spoofers are trying to manipulate public information, upon which others rely. In this, he distinguishes spoofing from iceberging and argues that the latter concerns private information but because spoofing can affect prices it concerns public information. Nevertheless, Hersch is incorrect here. In both iceberging and spoofing cases what the agent conceals is their preferences which sit behind their offer, this is quintessentially private information. In both cases revealing this private information to counterparties such as HFATs would cause them to alter the terms which they would contract on, affecting public prices. Following the offensive–defensive deception distinction, this is a case of defensive deception, deceiving, “… to protect proprietary information, information about which your audience has no right to know.” (Strudler, 2005, p. 465).

To address examples which fit the offensive deception allegations Hersch levels at spoofing, one could imagine either posting offers which cannot be filled or hacking into the computer terminals of market counterparties to falsify the market updates they receive from the exchange. In the first case the offer itself is fake– there is no actual offer of a binding contract. This distinction between fakes and bluffs is something Hersch’s analysis misses—by definition bluffs can be ‘called’. A spoofer who posts offers without wanting them to be accepted is still participating in the market practice of bids and asks—if a counterparty accepts their offer they are bound to fulfil them.¹⁴This is distinct from an outright fake offer which would subvert the actual practice and purpose of the market exchange as a forum to present legally binding offers. The analogous situation in poker would be a bet which when called the bettor simply refused to show their cards or pay their chips to the pot. The second example of hacking distorts information over which the hacker clearly has no ownership—the reliability of the information flow from the market to its participants. Hersch’s allegation of offensive deception is that spoofing distorts public information by placing orders they do not wish to be fulfilled, but this is not correct (Hersch 2020, pp. 213–214). The spoofed offers are real—they can be accepted, and just like any other order on the exchange they can also be withdrawn prior to acceptance. The potential for deception or misunderstanding only arises because other participants are used to forming assumptions and building trading models based on inferring the intentions which sit behind offers. The example of hacking makes clear what it would mean to actually distort information flows regarding the state of the market’s orders.

Nevertheless, it is worth noting that in contrast to the two-party auction example above financial exchange markets are large anonymised venues—speculators try to infer preferences from the aggregate activities of all anonymous market participants rather than specific identifiable agents.¹⁵ Recall, however, that the orders of single agents can influence the market—particularly at the margins. This is why hedgers iceberg their orders to try to prevent speculators profiting from them. Spoofers interfere with, and take advantage of, this process by mimicking the market activity of an agent interested in a large number of orders, and in so doing confound speculators’ abilities to read and profit from the underlying preferences of other market actors who may genuinely want to place such large orders. A positive externality of spoofing, therefore, is that the effects of its defensive deception extends to other market actors whose intentions similarly become opaque to speculators. Even if a spoofer is a speculator the veil of uncertainty they provide regarding underlying intentions protects hedgers who want to place large orders in the market without being so predictable that speculators can anticipate and profit from market movements precipitated by their trading preferences.¹⁶ Still, it is important to recall that the deception fundamentally remains one of concealing by misrepresenting underlying intentions, which counterparties have no right to know, otherwise it follows the market practice of making, accepting, or withdrawing offers. The market is still filled with legitimate orders which can be accepted by counterparties, or withdrawn by offerors. What has changed is the assumptions which can be made regarding the intentions which sit behind the offers.

Although spoofers help all market participants protect their own proprietary information—their preferences—they are doing so in order to profit from those whose anticipatory trading strategies rely on clearly identifying future market trends based on the intentions they infer from existing offers. For some readers this may jar with the concept of ‘defensive’ deception. This intuition is possibly what causes Hersch to incorrectly differentiate iceberging as defensive deception and spoofing as offensive deception when both of them concern the actor’s personal preferences. It may be argued that there is something less than wholesome at profiting by inducing misunderstanding in others. However, the converse view should also be considered—why should certain speculators be free to profit unchallenged because other market participants, and in particular hedgers, allow their private information to be read so easily? This is what Hersch effectively implies when he suggests that perhaps spoofing and iceberging should both be prohibited (Hersch 2020, p. 214). Traders lose out when they are predictable and their preferences are discerned by speculators, this is why hedgers adopt strategies such as iceberging to avoid HFATs anticipating the market impacts of, and profiting from, their orders. In fact, it seems reasonable to suggest that traders, were it legal to do so, should adopt a mixed strategy of fully desired offers and spoofed offers in order to make their own activities less predictable. I have outlined above why market participants, and spoofers in particular, do not act improperly by misleading counterparties as to the preferences underlying their offers. This logic applies irrespective of whether the strategy produces a profit for the user or not, and therefore, whether speculators or hedgers are actually adopting such spoofing strategies. The suggestion that spoofing be permitted in order to help hedgers avoid being such easy prey for HFATs raises the further question of whether this should be permitted from the perspective of the good of the market as a whole.

A Vision of Spoofing-Permitting Markets

The above section outlines why market participants, including spoofers, do not owe one another clarity as to their own preferences and should be free to misrepresent them so long as it is through channels endogenous to the market practice of making or withdrawing legally binding offers. However, when proposing regulatory change one needs to consider not only market participants’ duties to one another but also the impact on the market as a whole and its economic purpose. Recall that this economic function of markets is what separates the analysis of markets from that of gambling or games of chance. From a regulatory policy perspective we are not only concerned with agents’ duties to one another but also the aggregate outcomes of their interactions. My analysis here involves a distributional argument to the benefit of hedgers, who—as Hersch (2020) correctly identified—are the more important agents of concern in these markets. These markets, if they are to be distinguished from straightforward gambling markets, need to support economic hedgers. This can be distinguished from Cooper et al.’s argument that straightforward quality arbitrage—encouraging ongoing competition and evolution of strategies—should be the goal of regulation (Cooper et al., 2016, p. 12).¹⁷ Cooper et al., although they argue for the permissibility of spoofing, maintain that we should not prefer low-frequency traders as investors compared to high-frequency speculators (Cooper et al., 2016, pp. 15–16). They conclude that the marketplace exists to provide an effective place for investors to buy and sell financial instruments and that, “… no [type of] firm has a moral right to have its continued viability guaranteed.” (Cooper et al., 2016, p. 15). In contrast to this view I follow Hersch in identifying the purpose of financial exchange markets as enabling the hedging of risk. Without hedgers a speculator-only market would have no role from an economic perspective beyond that of straightforward gambling, whereas a hedgers-only market could exist and fulfil an economic purpose beyond gambling but would suffer from the lack of liquidity provided by speculators (Hersch 2020, pp. 215–216). The two groups are in a symbiotic relationship, but hedgers provide the raison d'être for the market. Nevertheless, unlike both Cooper et al. (2016) and Hersch (2020) I see spoofing as playing an important role in the market in support of hedgers.

Some readers may sympathise with the idea of assisting hedgers at the expense of HFATs but nevertheless query whether there are not alternative reforms which might achieve similar goals. This article does not purport to comprehensively survey all possible changes to financial exchange markets which might alter the balance between HFATs and hedgers. One reform designed to rein in the profitability of HFATs has been the adoption of ‘speed bumps’ by certain financial exchanges (Osipovich, 2019). These are measures which slow down the execution of either all, or particular types of, market orders. Given the well-known advantages HFATs derive from maximising the speed with which they receive and send information, for example by co-locating as close as possible to market exchanges (Arnuk & Saluzzi, 2009; Angel & McCabe, 2013, pp. 590–591; Lewis, 2014), it is no surprise that the speed of information flows has been a target for reform. However, speed bumps are typically considered to work to the benefit of fellow market makers providing liquidity, essentially other speculators, as they are the ones primarily engaged in latency races where speed of execution is crucial (Aquilina & O’Neill, 2020). Spoofing, like iceberging, responds to anticipatory trading by HFATs and other speculators along the same axis by which hedgers can be exploited—by the predictability of their intentions. Furthermore, while the systemic impact of spoofing is important, it is important to emphasise that as argued above it should be permissible on the grounds of defensive deception concerning the spoofer’s private intentions. These considerations provide two important reasons to favour permitting spoofing in particular as a means to address the ability of HFATs to adopt anticipatory trading strategies at the expense of hedgers.

Spoofing has been illegal in many jurisdictions for a number of years, but we can observe an admittedly selective sample of recent instances of spoofing through the regulatory actions which have been brought against spoofers. One recurring feature of modern spoofing is that it tends to take place at the temporal margins of price fluctuations—offers are made and withdrawn in milliseconds (FCA 2013, Para 10; U.S. CFTC V. Eric Moncada et al. 2014 S.D.N.Y.; U.S. CFTC V. Igor B. Oystacher et al. 2016 N.D.Ill.) or within a handful of seconds (7722656 Canada Inc v. Financial Services Authority 2013, p. 23; FCA 2017, Paras 4.24–4.48).¹⁸ This gives us a strong indication of who spoofers are targeting. Spoofing works because speculating counterparties such as HFATs try to anticipate shifts in the market, often fractions of a second before they occur. This is distinguishable from the macro level at which hedgers generally operate. Our oil company which needs to hedge price volatility against its future deliveries is less interested in the second-by-second price because its needs are in the form of large bulk orders. Certainly the oil company would prefer to choose the optimal—meaning the cheapest—moment to actually purchase its hedging contracts but given the longer timescales they work at this is far more likely to be influenced by significant macro events affecting the oil market rather than second-by-second micro-fluctuations in contract supply and demand. Spoofers interact primarily with speculators who are day traders—market participants who seek to profit from intra-day price movements. These actors provide important liquidity for markets but importantly aim to do so at the expense of participants who want to place large orders. Recall that spoofers traditionally mimic a large swell in demand over the course of seconds—if they were a real hedger then the HFATs would be making money by pre-empting the price moves caused by their large orders. Instead the spoofer does not intend to go through with the large order, but by causing the HFATs to react as if there is a large order the spoofer can profit off of them instead.

A critic of spoofing might argue that the question of who the spoofer is targeting is beside the point, the problem being that they are somehow interfering with the quality of the market itself—its efficiency and, therefore, its price-determining function (Hersch 2020, pp. 217–218). Angel and McCabe assert that strategies such as spoofing move prices away from their ‘fundamental values’ (Angel & McCabe, 2013, p. 594). However this is to overly idealise what occurs in non-spoofed markets. At the micro-level, there is no market ‘undistorted’ by strategic activity because actors are continuously trying to outmanoeuvre one another.¹⁹ Speculators are continuously engaged in a zero-sum game of trying to buy low and sell high from both each other and hedgers—including by constantly placing and withdrawing orders.²⁰ Hedgers are conscious that speculators are seeking to anticipate and profit from their market activity. Therefore, they engage in strategic behaviour, such as iceberging, to try and protect their position and minimise the extent to which they are taken advantage of. It is important to remember that the market as represented by the bid-ask spread at any one point in time only reflects public demand in light of the strategic positions different actors have taken vis-à-vis one another. For example, the knowledge that HFATs are highly active in a certain market may dissuade certain hedgers from making large orders at various points in time. This demand simply will not appear on the exchange but that is not to say it does not exist, and in fact it might enter the market were conditions to be different. Prevailing prices are, therefore, driven by the interplay between those participants willing to enter into the strategic competition and trade and this in turn depends on the prevailing structure of the rules, including whether spoofing is permitted or not.

The acceptance of spoofing as legitimate would mean that the informational content of large orders is reduced because counterparties, and in particular speculators and HFATs, could no longer rely on them to easily predict large upswells in demand. One effect of this would be to encourage hedgers to act more freely and place more orders as they know speculators and HFATs will have to be more wary of anticipatory trading at the expense of their orders. Hedgers might themselves sensibly engage in a degree of spoofing to help prevent speculators and HFATs from reliably predicting, and profiting from, orders which they intend to fulfil. Alternatively, even if spoofing was primarily undertaken by speculators seeking to profit from other speculators the anonymised nature of the market would mean that hedgers would still benefit from the resulting uncertainty surrounding the intentions behind large orders.²¹ Hedgers would expect their large orders to have less of an impact in terms of moving prices against them. As a consequence, we would expect a relative reduction in the profits being made by HFATs who would otherwise adjust their prices more dramatically in response to large hedging orders. Ultimately this reduction in HFAT profit should effectively reduce the cost of liquidity for hedgers.

HFATs would be left with two choices in a spoofing-enabled market: work harder or give up on anticipatory trading strategies altogether. The former is most likely. Markets are complex self-adapting systems because strategic actors continuously adapt to stay ahead of one another (Davis et al., 2013, pp. 853–854). The predictive power of HFATs and their ability to trade by anticipating price movements from orders currently provides them with a lucrative revenue stream despite the best efforts of hedgers and competing speculators to overcome them. Spoofing would represent another twist in the strategic arms race between HFATs and their intended targets, but it would not represent an end to the practice of speculation. First, there are limits to spoofing. As discussed above spoofing constitutes a ‘bluff’ and comes with attendant risks. The market may move against a spoofer due to unrelated causes before their spoofed orders can be cancelled and instead the orders may end up being accepted.²² Therefore, the spoofer is limited by the fact that they have to be in a position to fulfil spoofed orders which are accepted.²³ Second, the existence of spoofing does not erase the fact that hedgers do have a practical need to make orders they want to be filled and which, therefore, convey real information about the short-term future state of the market. Just as in poker, the existence of bluffing reduces the easily interpretable informational content of a raise, it does not render raises informationally inert. Parties still draw inferences from raises, just as they would from orders in a spoofing-enabled market. The consequence would be that such inferences would be more uncertain but not impossible to make. Third, there are still ways for HFATs to make money, and therefore, provide liquidity, from speculative market making. Although individual orders would be harder to informationally parse, other strategies such as event-driven trading would remain fully viable. For example while a large bid on oil futures might become more difficult to read, the news of a large Saudi oil refinery being rendered inoperable would still have a predictable effect on the price of oil futures which HFATs and other speculators would race to profit from.

Therefore, while HFATs and other speculators might become less profitable—and even that is not a given—in a world with legalised spoofing, it seems unlikely that they would withdraw from markets altogether. It is difficult to predict exactly what effect this would have on market liquidity for hedgers who, after all, still need counterparties for their trades.²⁴ Nevertheless, as outlined above spoofing would not pose an existential threat to HFATs and other speculators—it would require them to work harder to analyse the implications of the order book and to continue to update their trading strategies. For example, they may become adept at recognising patterns of spoofing and, therefore, not react to it at all, rendering spoofing redundant and requiring a further update in strategy. It is difficult to predict exactly what the outcomes of these iterated strategic interactions will produce.

One concern might be a distributive one that the complexities of spoofing-enabled markets may ultimately be to the benefit of dominant market actors. This level of analysis in terms of predicting market outcomes on differently sized participants is beyond the scope of this article, but nevertheless two points can be made. The first is that from the admittedly small and selective sample of public enforcement actions both large and small actors have been shown to be capable of spoofing. As cited above these range from JPMorgan or Bank of America Merrill Lynch employees to self-employed individual traders like Coscia, Sarao, and Igor Oystacher. This is not necessarily predictive of what market practices in spoofing-enabled markets would look like, but it does strongly suggest that the ability to spoof is accessible to the full range of market actors irrespective of size. The second is that the anti-spoofing status quo is not neutral insofar as it already benefits HFATs who form some of the largest and most powerful market participants and have an interest in maintaining their profitable strategies. HFATs include some of the world’s largest hedge funds, such as Two Sigma Investments and Citadel. For example, although HFATs might prefer to keep their role in regulatory enforcement actions out of the press it is known that Citadel both encouraged the CFTC to pursue Oystacher as his spoofing activities harmed their trading profits and went so far as to provide witnesses in the case against him (U.S. CFTC V. Igor B. Oystacher et al. 2016 N.D.Ill., 15–22). Similarly in 2010 when the CFTC sought to crack down on spoofing it approached premier HFAT firm Jump Trading for advice (Slide, 2021).

On the upside spoofing holds the potential to incentivise economic actors to act more freely on financial exchange markets. By reducing the predictable profits of speculators the economic hedgers are effectively paying less for liquidity. As the productive activities of economic hedgers provide the normative rationale for financial exchange markets in the first place this should be seen as a welcome development.

Conclusion

This article has explained that spoofing works by obfuscating the preferences behind certain orders on financial exchanges—in particular the intention that orders be cancelled before they can be filled. In doing so spoofing makes it harder for speculators, and in particular HFATs, to anticipate market moves at the second-by-second micro-level based purely on changes in order flow. In current practice speculators and HFATs make money from anticipatory trading by using the information provided by orders against those placing them. Orders which indicate preferences towards significant buying or selling demand prompt speculators and HFATs to react and adjust their prices against those placing the orders. Spoofing makes it harder for speculators and HFATs to readily infer these preferences from the mere fact that certain orders have been placed.

In carrying out this analysis I have argued that market participants do not owe it to one another to reliably signal their preferences. I have further argued that inducing others to form incorrect beliefs about one’s preferences is permissible as a form of defensive deception so long as one uses the medium of an accepted part of the practice itself—making, accepting, or withdrawing of legally binding offers. This is because bluffing by placing market orders does not invite trust from market counterparties beyond a willingness to complete such orders if they are accepted. It is, therefore, distinguishable from problematic deceptions which might damage trust. The upshot of this is that spoofing does not violate duties held by market participants towards one another and in fact should be permitted as a way for parties to be protected from their intentions being too easily discerned and profited from.

From a system-wide perspective spoofing holds the potential to redress the balance of power away from speculators such as HFATs and towards hedgers who wish to place large orders and whose economic interests are what make such markets more normatively significant than straightforward betting markets. This shift in the balance is unlikely to eliminate speculators and HFATs from the market completely, which would undermine the important function they provide in adding liquidity to the market. Instead it would require speculators and HFATs to find ways to profit from market-affecting events other than trying to immediately trade ahead of buy or sell orders, or they will need to do so in a more sophisticated manner which incorporates the risk that such orders might be being spoofed. In shifting this balance of power spoofing offers the promise of more freedom of action for hedgers and encourages the placing of economic orders.

Spoofing, therefore, does not deserve the normative condemnation and legal prohibition which it has received. Instead, it has the potential to serve a salutary purpose in the proper running of financial exchange markets. As well as providing an analysis of financial market practice this article contributes to the wider business ethics literature on bluffing in an economic context. It illustrates how bluffing through the mechanisms of a market’s practice can be conducive to the positive functioning of the market itself. Financial exchange markets represent an economic interaction where communication is restricted to market bids from which other parties can draw inferences. There is here a close analogy with the misrepresentations of mental states contained in the bluffs central to the game of poker. The analogy becomes more complicated when it is applied to the types of verbal discussions and negotiations which poker bluffing is often compared to in the business ethics literature. In these verbal instances the analysis of what discussions form part of the market practice and what can be considered exogenous deceptions or misrepresentations is more complicated. However, this analysis of spoofing helps by providing a clear limiting case whereby bluffing through market practice should be permissible, both in its own right and because of the positive market outcomes it can facilitate.