In brief. A SEBI matter does not have to end in a finding. Under Section 15JB of the SEBI Act and the Settlement Proceedings Regulations, 2018, most noticees can settle, often without admitting or denying the findings of fact and conclusions of law. The economic question is not whether settlement is available but whether it is the right move. The answer turns on a small, repeatable arithmetic: the probability of an adverse finding, multiplied by the expected magnitude of penalty and disgorgement and the cost of litigation through SAT, weighed against the settlement amount the regulator's framework would assess. Across the modern record, only about 1,949 of approximately 35,417 SEBI-related orders are settlements; the rest go through contested adjudication. That selection is not random.
The doctrinal piece on what settlement actually is, including the "neither admit nor deny" framework and the post-2018 architecture, is at Can You Settle With SEBI Without Admitting Guilt?. This chapter is the operational complement: when should you settle, and what does the math actually look like?
Why is settlement a smaller share of the record than people assume?
Because the regulator does not let every matter settle. Settlement under Section 15JB and the 2018 Regulations is available to the noticee, but the regulator retains discretion to refuse where the default is too serious or where settlement would not be in the interest of investors or the orderly development of the market.1 Across the data study of approximately 35,417 SEBI-related orders, settlements account for roughly 1,949, alongside roughly 6,339 WTM orders, 11,766 AO orders and 15,363 SAT orders, as set out in How Does India's Securities Regulator Actually Work?. The matters that go through contested adjudication are not always the most serious ones; they are often the matters where one side or the other concluded the math did not favour settlement.
What does the math actually look like?
A small expected-value calculation, applied honestly. Take the probability of an adverse finding on the merits, multiply it by the expected monetary direction (penalty plus disgorgement plus any debarment-cost equivalent), add the present value of the legal and operational cost of running adjudication through to a SAT appeal, and the SAT appeal's own probability-adjusted outcome. Compare that figure to the settlement amount the regulator's framework would assess, plus any non-monetary undertakings, plus the loss of the right to fight for a finding of no violation. Where the settlement number is smaller than the expected-value figure, settlement is rational. Where it is materially larger, contest is rational.
Why does the "neither admit nor deny" feature matter to the math?
Because the cost of an admission, especially for a regulated entity or a listed company, can be very large in collateral terms. A formal finding of violation flows into parallel proceedings, into bidding eligibility, into board-level reporting, into reputational hits that the settlement amount itself does not capture. The 2018 Regulations expressly allow a noticee to settle while neither admitting nor denying the findings of fact and conclusions of law, the route most settlements take.1 The value of that feature is the value of avoiding the collateral consequences of a contested finding, and for some noticees it dominates the rest of the calculation.
How is the settlement amount calculated?
By a structured framework. The 2018 Regulations set out a method that derives a base amount from the nature and gravity of the default, adjusted by factors and tables specified in the regulations, so that the settlement figure bears a defined relationship to the seriousness of the conduct.1 The proposal is reviewed by an internal committee and then by a high powered advisory committee, which can require revisions before the settlement is accepted. The number is not arbitrary; it is the regulator's structured estimate of what a fair settlement looks like for the conduct alleged. That structure also makes the settlement amount predictable enough for noticees to do the comparison this chapter is about.
When is settling cheaper than fighting?
When the probability-weighted exposure on the merits exceeds the settlement amount by enough to cover litigation cost and the option value of fighting. Routine disclosure defaults, technical compliance lapses, and matters where the substantive case against the noticee is strong but the conduct does not warrant a contested finding are the textbook candidates. The settlement amount is calibrated for these cases, and the probability of an adverse finding is high; the math favours closure.
When is fighting cheaper than settling?
When the regulator's case has visible weaknesses on procedure or evidence, when the substantive law is contested and the noticee has a real shot at vindication, or when the precedent value of winning is itself worth more than the settlement saving. Several of the cases covered in this series, including the eventual outcomes in NSE Co-location and Balram Garg, show that contested defences can succeed; in those matters, settling would have been the more expensive choice. The strategic test, once the math is done, is whether the substantive case is strong enough to bear an appeal through SAT to the Supreme Court, and whether the noticee's institutional appetite supports that timeline.
What about the non-monetary terms?
They are part of the price. Settlements routinely carry non-monetary terms, including voluntary debarments for fixed periods, compliance undertakings, and the appointment of independent reviewers or auditors. Those terms have real economic value: a voluntary debarment from a market segment can cost more, in foregone business, than the cash settlement amount itself. A disciplined settlement negotiation tests not only the rupee figure but the non-monetary undertakings, often trading one against the other to land on a package that the noticee can actually operate inside.
Why does this matter for how a SEBI matter is run?
Because the settlement-vs-contest decision is not a binary that gets made once at the end. It runs through the whole proceeding. The disciplined practice is to evaluate the math at the show-cause stage, again after the personal hearing, and again at any point where the evidentiary or doctrinal picture shifts. Settlement remains available throughout, and the eventual choice should reflect the up-to-date math at the point at which it is taken. How the broader machinery sits behind that decision is set out in How Does SEBI Actually Enforce the Law?.
Sources & citations
- SEBI Act, 1992, Section 15JB (settlement of administrative and civil proceedings), inserted by the Securities Laws (Amendment) Act, 2014 with retrospective effect from 2007; SEBI (Settlement Proceedings) Regulations, 2018 (in force 1 January 2019, repealing the 2014 Regulations), setting out the "neither admit nor deny" framework, the method for computing the settlement amount, and the internal and high powered advisory committee process.
- Legal Wires data study, How Does India's Securities Regulator Actually Work?, reporting approximately 1,949 Settlement Orders alongside approximately 6,339 Whole-Time Member orders, 11,766 Adjudicating Officer orders and 15,363 Securities Appellate Tribunal orders across the surveyed period.
About this article. Part of Legal Wires' SEBI Enforcement series, an analytical guide to India's securities enforcement record. This is general information and commentary, not legal advice; do not rely on it for any specific matter. Prepared with AI assistance and reviewed by the Legal Wires editorial team. Statutory provisions and regulations are cited above; the order-count figures are from the linked data study. Last reviewed: 28 May 2026. Spotted an error? Tell us and we will review it.