Noise Or Signal?: The Ground Reality Of The Jurisprudence Of Fraud

Past events always look less random than they were, it is called hindsight bias. Lending an attentive ear to someone's discussion of his own past can make one realise that much of what they were saying was just backfit explanations concocted ex post by their deluded mind.¹ The afore-stated caveat was conceptualised for participants in the securities markets all across the globe. However, it holds ample relevance for the enforcement mechanism of the Regulators of the very securities markets (presently, the Indian Securities Market and SEBI).

Contextually speaking, in the Indian jurisdiction, 'Fraud' inherently means one thing, but the narrative shifts based on the concerned law it finds a mention in, and the context it gains impetus from. For securities law, flowing from the parent act², most matters of PFUTP³ are theories told backwards: a scrip spikes, a chart/pattern looks odd, a handful of names surface based on limbs of connection and their trading activity in the said scrip, after the Investigatory dragnet is cast en masse.

More often than not, the investigation commences with a conclusion: "something is unusual here", and works on reverse-engineered narratives, stitching together a vocabulary of labels to allege fraud: connection, collusion, acting in concert, meeting of minds, nexus, artificial price, artificial volume, synchronized trades, reversal trades, no bona fide intention, cornering, sucking out liquidity.

Sometimes, the Investigation's theory is accurate. Often, the underlying empirical data is real, but the narrative layered over it is exactly what the abovementioned caveat raised alarms against: backfitted explanation, constructed ex post, with 'labels' sufficing for proof.

Over the last few years, the Hon'ble Securities Appellate Tribunal ("the Hon'ble SAT") has done something extremely valuable for securities law, almost without fanfare. The Hon'ble SAT has injected ample context using the irresistible inferences squeezed from foundational facts/factum of a given matter to the labels.

Jurisprudentially, "Connection" is no longer an assumption that SEBI inks in a show cause notice (be it a familial or a business/professional relation) without supplying the requisite context capable of crossing the threshold of 'collusion'⁴. "Artificial price" is not every price which may seem abnormal when divorced from the apt fundamental analysis of the scrip. "Artificial volume" is not any spike in the scrip viewed in a vacuum without the larger market dynamics at play. Without saying so explicitly, the Hon'ble SAT has been running a quiet quality-control exercise on the PFUTP vocabulary.

If one steps back from the individual appeals and reads the line running through the well-established jurisprudence of fraud, a coherent map emerges. The same labels recur across Orders, but they mean very different things depending on what sits underneath them. Where SEBI is perceiving randomness as noise, the Hon'ble SAT pushes back. Where there is a concocted and conniving scheme with the labels contextualised as signal, the Hon'ble SAT is equally willing to uphold serious findings.

The present literature takes birth from the need of the hour to distil the above-discussed map consisting of the PFUTP vocabulary which remains hidden in plain sight. The agenda of the present literature, neither is nor can be, to sneer at SEBI, but to politely nudge everyone in the enforcement and adjudicatory ecosystem to contemplate a harder question before reaching for the PFUTP vocabulary: what exactly does the label hold inside the very context it is weaving from the foundational facts in order to cement the allegation of fraud?

Connection

"Connection" is the regulator's workhorse. Allegations of "acting in concert" and "inter-connected entities" are now boilerplate. The difficulty is that in many cases "connection" is asserted at the level of social facts – someone knows someone, money moved at some point, a telephone number appears in more than one KYC, and then treated as if it automatically implies a meeting of minds in the execution of fraud.

The Biocon case, AKG Securities⁵, is an early and important course correction. SEBI tried to present a web: Sunil Gangwal ("Noticee 1/ Gangwal"), had taken a friendly loan from Minesh Mehta ("Noticee 6/ Mehta"); Mehta had financial dealings with 'Sand Enterprises', a firm linked to Vipul Trivedi, partner in Paramount Incorporation ("Noticee 5"); and there were banking transactions in July–October 2019 between Gangwal and Vihit Investment ("Noticee 2"). From this, SEBI treated the six noticees as a cohesive group that allegedly manipulated Biocon's price in the last half hour of futures expiry on a single day in June 2017.

The Hon'ble SAT refused to accept that kind of elastic reasoning. The alleged fund flows surfaced almost two years after the period of investigation. In this regard, the Hon'ble SAT demanded an irresistible inference from the foundational facts ("material/evidence on record"), which was conspicuously absent, and duly observed:

The Hon'ble Tribunal called it "absurd" to use ex post transactions to establish that the noticees may have "acted in concert" on an earlier date, and opined verbatim:

The supposed group had been inferred from weak connections and then used to colour ordinary trades as manipulative. There was no evidence of communication around the trades, no contemporaneous instructions, no bridging narrative between those financial relationships and the intraday activity in June 2018. Therefore, the SEBI Order⁶ did not survive.

Smitaben Shah performs a similar operation in a different fact pattern. SEBI alleged that a group of "connected buyers" had provided an exit to a QIB in Nissan Copper's IPO and "sucked out liquidity" by cornering its shares on the listing day. The evidence of connection was a scatter of family ties, old directorships, shared brokers and a few financial links. The Hon'ble SAT, far from being convinced, observed on the contrary, for the want of context to the label of connection, as:

The Hon'ble SAT refused to be constrained by SEBI's curated subset and held that the complete trade and order logs were essential to assess whether the appellants' pattern was in any way unique. The buyers were day traders who squared off by close; they did not carry positions. The IPO was oversubscribed; the free float and unmet demand were enormous. Further, the Hon'ble SAT pointedly noted that there could have been many other traders with similar or larger orders, which would have emerged had SEBI produced the full trade and order logs, yet only these 'connected buyers' were targeted. The Hon'ble SAT travelled a mile further to introspect SEBI's theory in terms of the impoverished quality of the Investigation conducted for a grave charge of fraud, and upon extreme dissatisfaction, the Tribunal remarked:

Therefore, once the Hon'ble SAT unveiled the context and none surfaced, the "connected buyers" label lost its bite.

Set those against Nilesh Kishanbhai Pandya⁸ ("Nilesh Kishanbhai") and Menika⁹, where the Hon'ble SAT was unambiguous in affirming SEBI's view. In Nilesh Kishanbhai, the Hon'ble Tribunal was willing to treat the nineteen accounts/noticees as a tight cluster because the evidentiary spine was hard, not impressionistic. SEBI did not stop at saying that the noticees "knew" each other. The Ld. AO traced common telephone numbers and shared addresses across UCC and KYC records, picked up at least one off-market transfer in another scrip tying a peripheral player back into the chain, and then overlaid this connectivity map on a trading pattern in which the same group executed 3,266 synchronised trades over thirty-nine days, accounting for roughly 12.86% of the total market volume, with every account closing the period flat.

The Hon'ble SAT simply upheld that the trading pattern gave "a clear indication of the connection" between the appellants and the other noticees and that the synchronized trades led to "an irresistible inference that there was a meeting of mind" between them. The Hon'ble Tribunal went on to reject the "less than 1%" defence, holding that in such a case it is the cumulative percentage of trades that matters, which was "substantial" and supported the conclusion that the purpose was to increase the LTP and give a "misleading appearance of trading without any transfer of ownership."

In Menika, thirteen entities carried out hundreds of synchronized and reversal trades¹⁰ in the scrip of Swan Energy on both exchanges i.e. NSE and BSE, with each appellant accounting for a non-trivial slice of the total synchronized volume¹¹ and sharing contact credentials with others. In fact, the evidentiary floor for "inter-se connection" in Menika was strikingly low. The Hon'ble Tribunal recorded that the appellants did not dispute the connectivity mapped in the Ld. AO's Order and, noting that they all shared a common mobile number and a common email ID, was content to treat them as "inter-se connected" on that basis alone.

Further, the Ld. AO's matrix then layered on the rest of the web: the same contact details recurring across multiple noticees' UCC/KYC records, common addresses and introducers at the same broker, and bank transactions linking Menika and Akash to another noticee, Akshit Gupta. In that factual setting, the Hon'ble SAT had no difficulty treating the thirteen entities as a connected cluster and analysing their synchronized and reversal trades at the group level rather than as isolated books.

Sanjay Poddar¹² (Maa Jagdambe Tradelinks) gives a rare, explicit unpacking of how "connection" morphs into collusion. The Hon'ble SAT starts by spelling out that "connection" is not a term of art but SEBI's own shorthand for any "link between entities": personal relationship, off-market transaction, financial relationship, even sheer location proximity. On the foundational facts, that connection-web was hard: Maa Jagdambe routed funds to Ridhi Hi-Fashion and A1 Century; there were overlapping directors across these entities; some appellants were employees within that web; and six core appellants repeatedly acted as first sellers in the scrip, placing tiny sell orders when large buy orders at higher prices were already sitting in the book, and together contributing close to 70% of positive LTP in the first patch and nearly 24% in the second.

The Hon'ble SAT rejected the argument that these were "far-fetched" links, noting that common directorship is "not a fleeting relationship like a co-passenger sharing a common seat in a passenger bus but one of long time trust", and held that the non-overlapping, choreographed trading pattern "indicates prior meeting of minds which cannot be termed as a coincidence".

Equally important is where the Hon'ble SAT puts a full-stop. Three other noticees also showed up as top contributors to positive LTP, but there was admittedly no connection with the company, no group case set up inter se, and no linkage to the entities that had placed the aggressive buy orders. For them, the Hon'ble SAT held that their conduct, "though highly suspicious, is short of finding their culpability on preponderance of probabilities", and applied Nishith M. Shah HUF¹³ to let them out. The judgment, therefore, hard-coded that "connection" is the evidentiary gateway; it turns into collusive fraud only when fund flows, governance links, and a structured trading pattern together make a common design more probable than coincidence. Absent that bridge, even sharp LTP-positive trading remains suspicious, not proof of a violation of PFUTP.

In a Nutshell: "Connection" in the PFUTP vocabulary has quietly shifted from being a label to being a threshold. It has to be contemporaneous, behaviourally relevant, and capable of supporting an inference of coordination, not just an excuse to join names in a cause-title.

Artificial Price/Volume

The expression "artificial price" appears with intimidating regularity in show cause notices. It needed a proper legal anchor; Jagruti Securities¹⁴ provides it. In that case, the broker was accused of raising the price of a scrip by placing buy orders in small quantities at rates higher than the previous close. The Ld. AO asserted that the said orders "artificially" pushed the price up. At the same time, the Ld. AO never sustained the specific charge that the trades themselves were not genuine.

The Hon'ble SAT took the abovementioned inconsistency head-on. If a trade is accepted as genuine: arms' length, executed through the exchange, Delivery-Versus-Price¹⁵ compliant; then, as the Hon'ble SAT put it in substance, a genuine trade produces a genuine price. Net-net, one does not get to call the price 'artificial' while treating the trade as real.

To close the loop, the Hon'ble SAT walked through how the anonymous screen works: price–time priority, best bid–best offer, the fact that a match only occurs when a willing buyer's maximum meets a willing seller's minimum. When that system spits out a price, that price is "truly the price which a willing buyer would pay to a willing seller". In that setting, artificiality cannot be attributed to the system itself; it must be attributed to human intervention that subverts it. In this regard, the Hon'ble SAT exponentially explained:

Further, the Hon'ble SAT then wrote the sentence that now sits at the centre of any discussion on PFUTP:

Moving further, Vikas Bengani¹⁶ and Rajiv Maheshwari¹⁷ are the twin stress-tests for SEBI's use of "artificial price", "artificial volume" and "manipulative trades" when collusion is missing. In Bengani, the WTM took a run of delivery-based buys in the scrip of Brijlaxmi Leasing ("BLFL"), one individual and his related entities repeatedly purchasing in an illiquid, loss-making scrip (as described in the WTM Order), often above LTP, across two investigation periods, and labelled it as "unusual (and apparently irrational) behaviour" which "gives rise to conclusion" that he was "acting to artificially prop up the prices" and was "instrumental in creating artificial volume" and "part of a larger operation". Yet the WTM Order recorded that the investigation "could not establish his links with any other person to complete the overall pattern of market manipulation" and that all trades were on-market and delivery-based.

The Hon'ble SAT started from those findings: Bengani was only on the buy side, his orders went through the exchange in the ordinary way, and there was "no connection with the persons on the other side of the trades executed by him". On that footing, the Hon'ble SAT was categorical: "Merely because an investor like the appellant placed orders for the purchase of a scrip at a price higher than the last traded price does not by itself lead to the conclusion that he was manipulating the price of the scrip... one cannot jump to the conclusion that the price of the scrip was being manipulated merely because a buy order was put in at a price higher than the last traded price. Such an action is only indicative of the desire of the purchaser to buy the shares for whatever reason." The Hon'ble Tribunal then crystallised the rule: "in order to establish the charge of price manipulation collusion between the buyer and the seller is necessary", and in Bengani's case "it has not been alleged that the appellant was colluding with the counterparty (seller) or with his broker."

Maheshwari pushes the same principle into a group-manipulation setting. SEBI's Investigation in the scrip of BFL Asset Finvest involved 15 noticees, with three (Noticees 1, 2 and 5) found to have generated artificial market volume through synchronized, circular and reversal trades and others, including Rajiv as Noticee 6, fitted around that core. Rajiv's alleged role was tightly framed: he placed the first order of the day on seven trading days, in minuscule quantities, at prices higher than LTP, contributing Rs. 5.08 to a total market positive LTP of Rs. 117.50, with Jaipur Infragold (a Group I entity) as counterparty on all seven days.

On that basis, the Ld. AO held that the concerned trades "for miniscule quantities contributed to the increase in the price of the scrip of the company with the intention to manipulate the price upwards" and branded them "manipulative", while at the same time recording that Rajiv was "not connected with entities of group 1 or group 2" and had "not executed any manipulative trades with any connected entity". The Hon'ble SAT called that out "at the outset", holding there was "no connection of the appellant with the other noticees who had manipulated the price and contributed to artificial market volume" and narrowing the issue to whether his standalone trades can, by themselves, be fraud. The answer was no: Rajiv placed lawful limit orders (permitting a buy up to 5% above LTP), and "such orders when executed automatically raises the price of the scrip which was permissible under the limit orders"; his trades "did contribute to the increase in the price of the scrip... but there was no intention to manipulate the price upwards". The scheme and "meeting of minds" lay with others.

In a Nutshell: Genuine on-screen trades yields a genuine price, and "artificial price/volume" only begins where there is a demonstrable meeting of minds between buyer and seller. Above-LTP buys, miniscule first trades and eye-catching LTP contribution numbers, standing alone and without collusion, are noise, not fraud.

Investigations & the Quest for Context

There is a quieter, but equally important, theme running through the decisions and observations of the Hon'ble SAT, and it has less to do with charts or formulas and more to do with the quality of investigation. The Bottomline is that the Allegation of 'Fraud' is context-sensitive (emphasis supplied). The same trading pattern can be innocent in one setting and damning in another. A scalper's blotter will look nothing like a positional investor's; an algorithmic strategy can resemble a machine gun compared to a retail punter's occasional shots. That difference only emerges when one forces to contemplate the most basic, and most neglected, question in matter of market manipulation ("Matters"): why was this trade put on at this time in this size?

In umpteen matters, the requisite context that ought to be infused in the labels is missing. In AKG Securities, each noticee offered a trading rationale that was, on its face, coherent: large-book investing driven by in-house software, news-based intraday positions, day trading built around specific limit and stop-loss logic. The Final Order of the Ld. AO barely engaged with those explanations; it sidestepped them and went back to the trading pattern. In Bengani, expressions like "unusual behaviour", "no bona fide intention" and "larger operation" took the place of a serious interrogation of strategy. Why this scrip? Why keep buying a falling stock? Why above LTP instead of waiting to be hit? In Maheshwari, a string of tiny, first-of-the-day limit orders was labelled "manipulative trades" while the same order recorded that he was not connected with any of the manipulative groups and had not traded with any connected entity. The gap is the same: labels where there should have been a reasoned dismantling of the noticee's own account.

Those are not tick-box questions. There can never be a hardened checklist for fraud in a market where schemes mutate and strategies differ. That is precisely why context matters so much. If a regulator is going to accuse someone of fraud, it must be able to say, after hearing the noticee's explanation, why the fraud narrative is still the most plausible explanation, not simply the most convenient one. It has to show, on a preponderance standard, that it is much more likely than not that the conduct in question was part of a scheme, not just an uncomfortable outlier.

Seen properly, securities markets are a risk–reward bargain. Every genuine trade involves both buyer and seller putting capital at risk on the basis of their view of value. Fraudsters try to invert that equation. The architecture of a classic PFUTP scheme: circular trades, synchronized trades, reversals with no beneficial change in ownership, is designed to strip risk out of the operator's own book while still generating price and volume "signal" on the screen. Shares are rotated within a closed cluster, positions net out flat, and the only participants actually left holding risk are the outside investors who enter on the strength of that manufactured tape. Economically, that is the core of "artificial price" and "artificial volume": the appearance of activity without a corresponding transfer of risk.

There is, of course, a flip side. Trading pattern is not irrelevant; in the right fact set, it is the loudest voice in the room. When a handful of accounts keep meeting each other on both legs, day after day, at pre-arranged prices; when their aggregate trades command a double-digit share of market volume; when their books repeatedly end flat with no real exposure; at that point the pattern itself starts to scream collusion. In those cases, (Nilesh Kishanbhai and Menika) the tape is not noisy data waiting to be explained. It is the primary signal that a scheme exists, and the PFUTP vocabulary is simply catching up with what the market microstructure already shows.

That is why there can never be a mechanical "2 (connection) + 2 (trading pattern) = 4 (fraud)" identity. Fraud in the market is closer to a linear equation with variables than to an arithmetic sum. Call it x + y = allegation of fraud, where x is the context and quality of the connections: are these links, in their nature, timing and depth, such that it is more likely than not that the noticees acted in concert; and y is the trading pattern: does the structure, synchronisation and economic irrationality of the trades, taken as a whole, speak and scream scheme rather than mere noise? In some matters, x is weak and y is weak; the PFUTP vocabulary has nothing solid to rest on. In others, x and y both point the same way with enough force that the only coherent explanation left on the table is design. That is the space in which "connection" legitimately hardens into "collusion" and where a trading pattern stops being background noise and becomes actionable signal.

Viewed against that frame, the Hon'ble SAT's repeated insistence on connection, collusion, structure and materiality is less hostility to enforcement and more a reaction to investigative gaps. Half-baked, conclusory treatment of loose connections that have no real bearing on the trading activity clubbed with a context-free reading of charts is not a recipe for cementing an allegation of fraud. Investigations that only scratch the surface of foundational facts, or treat those facts as ceiling rather than floor, invite exactly the sort of pushback from the Hon'ble SAT. When that understanding is missing, hindsight bias takes over, and everything retrospective starts to look like a scheme.

The message is narrower and more valuable than a blanket defence of traders or a blanket endorsement of SEBI. The words used in the PFUTP vocabulary are not free-floating. "Connection" ought to metamorphose into "collusion" only when the evidentiary bridge is there. "Artificial price" has to be tied to collusive trades that subvert the system, not to any price move the regulator happens to dislike. "Artificial volume" must reflect a real scheme, a manufactured appearance of liquidity with no transfer of risk, not just busy intraday books. "Non-genuine buyer" cannot mean "someone whose trading style I do not like". Context has to be built into the evidentiary story from the start, not bolted on at the end. Statements have to be used to narrow the field of plausible explanations, not to tick a procedural box. LTP charts and volume tables are starting points; they are not conclusions.

Past trades will always more intentional and more structured than they actually were. That is the nature of hindsight. It is in such situations that the guidance of the Hon'ble Supreme Court becomes most apposite. In Hanumant vs. State of Madhya Pradesh¹⁸, the Hon'ble Court cautioned against the danger of conjecture or suspicion being elevated to the standard of legal proof, recalling Baron Alderson's warning in Reg. v. Hodge ((1838) 2 Lew. 227):

All in all, the law's job is not to rebadge market noise as fraud, but to separate noise from signal and reserve the language of fraud for those situations where randomness has genuinely been displaced by design.

Footnotes

1. Fooled by Randomness, Nassim Nicholas Taleb (Page No. ix, Preface)

2. Securities and Exchange Board of India Act, 1992 ("SEBI Act")

3. SEBI (Prohibition of Fraudulent and Unfair Trade Practices) Regulations, 2003

4. "An agreement to defraud another or to do or obtain something forbidden by law." Black's Law Dictionary Tenth Edition, Bryan A. Garner.

5. AKG Securities and Consultancy Ltd. v. SEBI [Appeal No. 368 of 2021 decided on 28 July 2025]

6. Final Order in the matter of Biocon Limited [Ref. No. WTM/AB/IVD/ID-3/06/2021-22 dated 5^th May 2021]

7. Ms. Smitaben N. Shah v. SEBI [Appeal No. 37 of 2010 decided on 30^th July 2010]

8. Nilesh Kishanbhai Pandya & Ors. V. SEBI [Appeal No. 56-58 of 2023 decided on 18^th December 2023]

9. Menika v. SEBI [Appeal No. 468 of 2022 decided on 30^th November 2022]

10. The reversal trades on BSE alone as 2,263 trades with 7,03,023 shares and artificial volume of 14,06,046 shares (12.89% of market volume).

11. For Instance: Menika: 160,089 shares, 176 synchronized trades, 11.52% of total synchronized traded quantity on NSE; 141,299 shares, 161 trades, 10.57% on BSE.; Akash: 124,604 shares, 121 trades, 8.97% on NSE; 174,440 shares, 619 trades, 13.05% on BSE.; Monika: 70,907 shares, 93 trades, 5.10% on NSE; 144,892 shares, 177 trades, 10.84% on BSE.

12. Sanjay Kumar Poddar HUF v. SEBI [ Appeal No.326 of 2020 decided on 24^th August 2021].

13. M/s Nishith M. Shah HUF v. SEBI [Appeal No.97 of 2019 decided on 16^th January 2020].

14. M/s. Jagruti Securities Ltd. V. SEBI [Appeal No. 102 of 2006 decided on 27^th October, 2008].

15. In a delivery versus payment (DVP) system, the exchange of securities and cash happens simultaneously, so the buyer receives the assets only when the seller receives payment, creating a safeguard against default.

16. Vikas Ganeshmal Bengani v. WTM SEBI [ Appeal No. 225 of 2009 decided on 25^th February, 2010].

17. Rajiv Maheshwari v. SEBI [ Appeal No. 358 of 2023 decided on 18^th May 2023].

18. [1975 AIR 1083]

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.