Clifford Chance

On 1 July 2025 the Upper Tribunal, Tax and Chancery Chamber, handed down a decision dismissing the references brought by Diego Urra, Jorge Lopez Gonzalez and Poojan Sheth. It directed the Financial Conduct Authority to impose prohibition orders on all three men and to levy individual financial penalties of £223,400, £100,000 and £57,600 respectively. The Decision is important because it represents a rare example of the Upper Tribunal conducting a detailed examination of so-called "spoofing" allegations.

What were the facts?

The relevant trading occurred during the summer of 2016 and was carried out by three members of Mizuho International plc’s European Government Bond ("EGB") desk in London – Diego Urra (desk head), Jorge Lopez and Poojan Sheth. The desk was a so-called third-tier market-making operation that serviced client requests for prices in Italian government bonds ("BTPs") and hedged its cash-bond risk with Long-Term Euro-BTP futures ("BTP Futures") on the Eurex electronic order book in Frankfurt.

Mizuho’s third-tier position in the market meant it saw only a fraction of overall BTP flow: it received fewer and smaller requests-for-quotes ("RFQs") from clients meaning it was operating at an information disadvantage in a competitive market, in which it was seeking to improve its position.

What was the alleged market abuse?

The FCA's case was that the three traders executed a co-ordinated spoofing scheme on 233 occasions between 1 June and 29 July 2016 in the BTP Futures contract on Eurex. In each episode, on one side of the order book they displayed one or more visible "Large Orders" of at least 200 lots (often 400-500 lots) which they did not intend to trade,  either shortly before or shortly after placing a much smaller "Small Order" (typically under 200 lots and frequently iceberged) on the opposite side, which reflected their true trading intention. The Large Orders were never iceberged, were placed a tick or two behind the touch so that they were highly visible yet unlikely to trade, and were almost invariably cancelled within seconds, often immediately after the Small Order had filled. The FCA argued that, by showing size that was exceptional for this market, the traders created a false picture of supply or demand that induced other participants (including high-frequency algorithms) to lift or hit the genuine Small Orders more quickly and at better prices than would otherwise have been available.

The FCA argued that this systematic pattern of placing and withdrawing Large Orders amounted to deliberate market manipulation under both section 118(5) FSMA (until 2 July 2016) and Articles 12 and 15 of the Market Abuse Regulation thereafter. Key hallmarks relied on included: (i) the statistically unusual size of Large Orders across the market and the fact that desk placed more Large Orders than any other market participants despite trading less than 0.43 per cent of the total traded volume of BTP Futures; (ii) the fact that none of the Large Orders was placed as an iceberg while many of the Small Orders were; (iii) positioning of the Large Orders close enough to the touch to influence price, but not close enough to execute; (iv) an execution rate of 0 per cent for virtually all Large Orders versus a 95 per cent execution rate for the overlapping Small Orders; and (v) repetition of the pattern both individually and in concert, which the FCA argued showed the scheme was intentional, dishonest and designed to secure faster fills and better economics for the traders’ genuine business at the expense of a truthful order-book signal.

How did the traders explain their trading?

Both Diego Urra, who managed the EGB desk, and the junior trader Poojan Sheth argued that the large orders they entered formed part of an "Information Discovery Strategy". In their account, Mizuho's EGB desk's third-tier position meant it operated at an informational disadvantage because, unlike primary dealers, it saw only fragmented client flow: a client could split a sizeable cash-bond order across several dealers, thereby masking real supply or demand. This could result in the EGB Desk having an incomplete understanding of the market, causing it to misprice a trade. To test whether such hidden interest existed, Urra said he would post a conspicuously sized futures order a tick or two behind the touch, reasoning that a rival market-maker pressed to hedge an undisclosed cash fill would aggress that liquidity at a premium; if hit, the desk captured a favourable position, and if ignored the absence of interest yielded intelligence about the true depth of the market. Sheth, who said Urra had demonstrated the technique to him after a 450-lot order was filled on 12 May 2016, adopted the same rationale, explaining his multiple overlapping 500-lot orders as clumsy price amendments rather than spoofing. Both men maintained that they always intended to trade the orders, saw them as medium rather than large size, and cancelled them only when the market failed to respond, denying any aim to mislead or to synchronise activity with the smaller hedge orders that happened to be working on the opposite side of the book.

Jorge Lopez advanced a slightly different explanation, describing his Large Orders as "anticipatory hedges". He said he had been recruited to improve the desk’s hit-ratio for medium-sized (€20-30 million) RFQs. He would use the Bund futures market as a reference point during the day and sought to identify resistance and support levels in the Bund/BTP spread and anticipate the timing of predictable client demand. When he expected imminent buying, he placed a visible bid in the futures market at a price he regarded as attractive but unlikely to trade instantly; if executed, the fill would leave him pre-positioned and able to quote clients an aggressive cash price while already long futures, and if unexecuted the order could be cancelled before market risk crystallised. He emphasised that none of his orders were iceberged because he wanted the full size to be hit in one clip, that the size sat comfortably within his individual VaR limits, and that the overlap with the desk’s smaller hedging orders was an operational coincidence rather than evidence of spoofing or concerted action with his colleagues.

Why did the Tribunal reject the traders' explanations?

The Tribunal dismissed the traders’ "Information Discovery" narrative because it did not explain the pattern of their trading activity. The objective features of their trading were incompatible with genuine liquidity-testing. Their bid / offer orders of 200–500 lots were displayed for only a handful of seconds and were withdrawn almost immediately after the opposite-side Small Order had executed, giving them no practical window to discover hidden depth. The pattern recurred more than two hundred times, frequently at illiquid moments (including after the cash market closed) and, in multi-trader episodes, required split-second co-ordination of pricing and cancellation across the desk, a coincidence the Tribunal regarded as striking. The strategy rested on the improbable premise that a single counterparty would hedge a much larger split cash order by running through dozens of visible lots and then paying a premium for the traders' clip; the Tribunal described that assumption as "bizarre… unseen in the market". No notebook, record or supervisory discussion evidenced the design, implementation or approval of, or information gained from, the supposed experiment and the trades were all undertaken outside the desk mandate (which forbade speculative futures positions). Taken together these factors led the Tribunal to conclude that the Large Orders were never meant to fill and were simply there to mislead other participants.

The separate "anticipatory hedging" justification offered by Lopez was also rejected. The Tribunal found the 93 RFQs he relied on as evidence of RFQs requiring hedging with 200 lots or more were too small a slice of total client flow and arrived at unpredictable times, so any pre-positioning in 200-lot clips was speculative rather than a "highly likely near-term" hedge envisaged by the desk mandate. None of the 88 Large Orders that he allegedly placed in pursuance of his "anticipatory hedging" strategy ever traded; if the aim were genuinely to secure inventory at attractive levels he would have shown the size as an iceberg, left it on the book longer or tested smaller clips, yet instead he published the full quantity two or three ticks back where it was certain to shift the touch against him. Many of the orders were submitted late in the day when no fresh RFQs could reasonably be expected, and he often cancelled them within three to five seconds of the opposite-side hedge filling, displaying the same "striking coincidence" identified in his colleagues’ activity.

The Tribunal therefore rejected his argument and held that his conduct too fell outside the mandate and concluded that all three traders had knowingly placed false signals with no intention to trade the size shown.

Lone Large orders

The traders relied on what they called the "Lone Large Orders" - thirty-four visible futures orders of 200 lots or more that did not overlap with any opposing Small Orders. Defence counsel argued that these clips, entered and then cancelled without an opposite-side position, undermined the FCA's theory that every Large Order was an exercise in spoofing designed to move the price toward a smaller genuine order. Because the Lone Large Orders formed a "meaningful" proportion of the Large Orders the desk placed, defence counsel said they could only be explained by the traders pursuing their legitimate trading strategies. The FCA was criticised for offering "no explanation at all" for these apparent outliers and for selecting only the overlapping orders when presenting its case; the defence said that selectivity demonstrated the Authority’s "Texas-sharpshooter" approach to pattern-building.

The Upper Tribunal rejected that submission. The Tribunal acknowledged the FCA had not pleaded those orders as separate acts of abuse and that they were not otherwise explained on the FCA's case, but found the small number of such orders (particularly for Lopez, being only 3 of 88 orders) to be significant. The Tribunal considered that while the Lone Large Orders may have been consistent with the traders' professed trading strategies, the Lone Large Orders were not sufficient to persuade the Tribunal that such strategies actually explained their trading activity. The existence of the Lone Large Orders, whilst supportive of the Traders' case, was not fatal to the Authority’s case and was heavily outweighed by evidence in favour of the Authority.

Observations

The case provides important guidance for those who investigate potential instances of spoofing.

The Decision demonstrates that, in principle, "information-discovery" or "anticipatory hedging" could provide sufficient explanation for potential instances of spoofing, but such explanations cannot be accepted at face value. When a trader claims an information-discovery rationale, it will sensible to assess (i) whether a written strategy was drawn up and approved under the desk mandate, (ii) whether the size, visibility and life-span of the orders would reasonably allow another dealer time to aggress them and reveal hidden liquidity, and (iii) whether there are contemporaneous notes of the information discovered.  For anticipatory hedging, it will be helpful to consider whether the hedge is linked to a likely near-term client exposure.

In this case, Lone Large Orders were a very small proportion of the overall population of large orders and very few executed. In another investigation, Lone Large Orders could carry genuine exculpatory weight if they have objective features showing that these were not simply other examples of potentially manipulative orders. Those features might include, for example, a higher number of Lone Large Orders as a proportion of the overall number of large orders, or examples of Partial- or full-fills of the Loan Large Orders.