Between Innovation and Safeguards: Analysing SEBI’s 2025 Algorithmic Trading Circular (Part II)

Introduction

The first part of this piece has explained that the rationale behind the introduction of SEBI’s new algorithmic trading framework is to reinforce SEBI’s mandate of investor protection by safeguarding investors against the misuse, manipulation, and risks associated with ATS. It has further analysed the positive implications of these guidelines which include enhancement of accountability, evolving a disclosure-based regime founded on transparency, containment of structural risk, and professionalisation of the trading ecosystem.

This second part continues this analysis and shows how in spite of the meticulously designed framework, negative implications exist in practice. These include compliance difficulties, absence of systemic resilience stipulations, stifling of the open-source innovation ecosystem, and ignorance of the behavioural dimension of retail investors. It consequently conceptualises a model, suggesting changes in the introduced guidelines to resolve these identified inconsistencies.

Evaluating the Potential Negative Consequences

The analysis until now in Part I of this piece has presented a rather positive picture of these guidelines. However, these stipulations are not without their drawbacks. Firstly, this framework places disproportionate responsibility on brokers, assuming they possess technical expertise and infrastructure to monitor both, retail-developed ATS, as well as empanelled third-party providers. Although larger technologically advanced brokers can adapt, smaller brokers might face compliance difficulties due to the lack of resources and experience, ultimately leading to uneven enforcement. This is evident from the recent trend of active investors exiting markets, according to the recent National Stock Exchange data which estimates that in February 2025, eight of the top ten brokers in India experienced declines in the total number of active investors. This reduces the income by way of commission of brokers, consequently leading to the exit of smaller brokers from the market, particularly if tasked with additional responsibilities and compliance burdens in relation to ATS. For example, brokers including Groww, Zerodha, and Angel One, although amongst the largest brokers in India, have experienced significant declines of over twenty percent in the number of active investors in the past few months of 2025. There are also studies explaining how the introduction of algorithmic trading has reduced the number of manual traders from around thirty thousand in the 2000s to only two thousand by 2014, forcing brokers to shut their jobbing desks and reallocate resources to technology. This is in addition to the regulatory and financial challenges that brokers are bound to face with increased taxes on trading and SEBI’s tighter controls on retail futures and options trading, which is expected to reduce their income by thirty to fifty percent. SEBI in October 2024 increased the minimum contract lot size in index derivatives to fifteen lakh rupees from the current five lakh rupees, which made trading costlier. Although the intention was to protect retail investors from increased losses in the futures and options segment, this regulatory change means reduced trading volumes with increased costs. The implementation of the current algorithmic trading circular has also been pushed to April 1, 2026 because of brokers and vendors demanding extensions. Compliance has already become difficult due to vast changes in the execution of orders and sluggish implementation by vendors. If large broking houses are drastically affected, the impact on smaller brokers is bound to be detrimental, consequently leading to consolidation. This creates a two-tiered market structure wherein larger brokers dominate retail algorithmic offerings while smaller brokers are forced to exit the space altogether. This centralisation also contradicts SEBI’s broader agenda of encouraging competition and diversity among intermediaries. SEBI’s Financial year 2025 report itself indicates high market concentration with the top hundred brokers continuing to account for over ninety percent of the cash market turnover.

Precedent in the United States Treasury market also reflects this phenomenon. The passage of the Securities and Exchange Commission’s Market Access Rule in 2010 had raised concerns regarding consolidation with expectations that new algorithmic trading risk-controls might have unintended consequences of driving out small liquidity providers from the market, negatively affecting marketplace liquidity. The results are evident with the past decade witnessing the shutting down of at least eight independent broker-dealers, with the number of primary dealers falling from forty-six in 1988 to twenty-four today, in spite of an increase in the market size. The reason behind this includes the high fixed compliance costs with estimates stating that on an average, brokers in the United States spend roughly 13.1% of their net revenue on compliance, amounting to around 1.2 million dollars per year. Algorithmic trading regulations add to these costs due to the imposition of additional obligations which are manageable for big firms but become overly burdensome for smaller dealers.

Parallels can be drawn with the European Union’s (‘EU’) Markets in Crypto-Assets Regulation, wherein smaller market participants struggled to meet compliance costs and procedures, contrary to the larger institutional giants. Secondly, the guidelines seem to delegate crucial decisions such as the finalisation of the order per second threshold, norms for empanelment and turnaround times for ATS registration to the ISF and stock exchanges. While flexibility is valuable, such delegation risks regulatory capture by dominant industry players. This in turn creates potential for misuse because the corresponding thresholds and empanelment criteria would be designed to suit respective personal market positions. Smaller fintech start-ups would be locked out due to the high compliance entry barriers, which would undermine the very innovation system that SEBI seeks to foster. This goes against the scheme prevalent in the European Union wherein under the Markets in Financial Instruments Directive II (‘MiFID II’), the thresholds and obligations are defined directly under these Regulations, which ultimately upholds transparency, fairness, and the prevalence of non-discriminatory standards among the various member states. This is evident from Article 17 of the MiFID II which explicitly requires investment firms engaging in algorithmic trading to make timely disclosures regarding trading strategies and parameters and maintain appropriate records. It also requires firms to ensure that their systems are resilient, have sufficient capacity, risk control mechanisms, business continuity arrangements, response and recovery plans, and are subject to appropriate thresholds and limits.

The provision which restricts retail developed ATS from being used in the market is protective but it risks stifling open-source innovation ecosystems. This is because retail algorithm developers often collaborate in practice through the sharing of code libraries and backtesting frameworks. This becomes evident in the context of MetaTrader, an electronic trading platform, which fosters efficient algorithmic testing, shared code libraries, and rapid knowledge dissemination. Quant Connect, an open source algorithmic trading software, allows the sharing of over twelve hundred strategies through the forums, monthly creating around twenty-five hundred new algorithms and one million lines of code. By legally excluding the entire category of retail developed ATS from public utilisation, SEBI might end up inadvertently discouraging the very ecosystem that could potentially generate indigenous innovation in the algorithmic finance landscape. In the United States, alternative trading systems are permitted to flourish and effectively operate under disclosure-heavy regimes, merely with a few additional compliances, which contradicts SEBI’s approach in India. The introduction of this framework may compromise innovation to merely favour risk-aversion. The existing liability structure creates a blame-shifting triangle wherein the principal-agent relationship is introduced, ATS providers are only empanelled and not directly regulated by SEBI and stock exchanges supervise but do not bear liability for any systemic malfunctions or violations. It is unclear who ultimately bears responsibility, especially if a black box algorithm goes rogue. The absence of direct SEBI oversight over ATS providers creates a regulatory gap which could be negatively exploited.  This is unlike the MiFID II regime wherein investment firms deploying ATS are themselves directly held accountable by the European Securities and Markets Authority, EU’s securities market regulator. The absence of a similar provision in India creates a legal grey zone for investors seeking redress.

Additionally, SEBI’s framework is heavily individualised and operational in the sense that it deals with monitoring or supervision of individual ATS but lacks explicit systemic resilience obligations such as stress testing and capacity requirements, emphasised upon under the EU MiFID II framework which also provides detailed kill-switch protocols for high frequency trading. Kill-switches are mechanisms designed to immediately halt the activity of a malfunctioning algorithm or a trading system that goes rogue. They prevent broad market disruption because algorithms place thousands of orders per second and even a minor glitch can create flash crashes and destabilise prices. The Knight Capital incident in 2012 which caused the company to lose approximately four hundred and forty million dollars in around forty-five minutes was a direct result of incorrect algorithmic input triggering millions of erroneous buy and sell orders. This necessitates the presence of safeguards including kill-switches and stress-testing requirements. The absence of these provisions in India creates the risk of the market set-up remaining vulnerable to systemic, correlated algorithmic failures. Moreover, SEBI’s reliance on SOPs issued by the exchanges instead of clearly codified statutory rules introduces uncertainty and discretion, ultimately leaving systemic safeguards under-addressed. Lastly, the framework underestimates the behavioural dimension of retail investors. Similar to the 2008 subprime crisis wherein credit ratings created a false sense of security for retail investors, the introduction of a formalised, SEBI-approved algorithmic trading regime might create an illusion of safety with the belief that all registered or empanelled ATS providers are inherently profitable or risk-free. Therefore, in the absence of strong investor education and risk disclosure frameworks, the regulatory system might inadvertently lead to overconfidence and excessive risk-taking, directly contradicting SEBI’s mandate and objective. Parallels can be drawn with an experiment conducted by Ruben Cox and Peter de Goeij which found that highlighting regulatory approval in issue advertisements or the prospectus itself raised the willingness of investors to invest by around ten percent and lowered perceived risk by almost six percent. A similar experiment in Czech Republic found that consumer-protection measures including deposit insurance and licensed fund managers acted as a strong stimuli to buying decisions, having a correspondingly misleading effect. Similarly, Jiri Sindelar and Petr Budinsky have argued that consumers tend to over-rely on official protective measures and even inexperienced investors neglect risk when regulators are involved. This creates a halo effect wherein regulated status creates misplaced trust. The European Securities Markets Authority, in the context of cryptocurrency, has cautioned that even when a licensed cryptocurrency firm offers unregulated products, the firm’s legitimate image can mislead investors into perceiving the offerings as safe. This is linked to optimism bias wherein investors with a positive outlook are more likely to take excessive risks, overlooking market downturns and other negative market sentiments. In the Indian context, there is ample evidence of overconfidence in bull markets and anchoring/ herd behavioural tendencies, particularly when social cues or regulator’s words dominate actual data. SEBI has itself noted that around sixty-two percent of retail investors rely on social media finfluencers to make investment decisions and hardly one percent have formal investment education. These facts indicate that the securities market posits heightened behavioural biases wherein formal regulatory frameworks can induce a false illusion of safety.

Conceptualisation of the Way Forward Through a Model

It is undisputed that the introduction of this framework is bound to have beneficial consequences. However, the analysis above indicates a few possible drawbacks including systemic vulnerability and behavioural complacence that might arise during or post the implementation stage. To counter such inconsistencies, this paper has introduced a model which serves as a potential judicial checklist or a set of legislative guiding principles to guide future law-making or the passage of decisions. This model has been titled – ‘The Layered Accountability with Sandbox Flexibility Framework’ to accurately reflect its various provisions. It is not intended to completely replace the introduced guidelines but to supplement them. However, in case of any discrepancy, this model should take precedence. The stipulations of the model include the following:

Principle 1: Direct Registration of ATS Providers and Shared Liability and Oversight

Instead of mere empanelment with exchanges, SEBI should introduce a tiered registration system with varying compliance burdens, ranging from light-touch in case of white box algorithms to full research analyst style compliance procedures for its black box counterparts. Similar to the EU MiFID II approach, ATS providers should be directly regulated by the SEBI without placing accountability solely upon the brokers. SEBI’s oversight should complement broker supervision and not replace it, which would mitigate behavioural and other risk without absolving intermediaries. This fills the gap identified earlier wherein brokers were burdened with disproportionate liability, despite lacking visibility into third-party ATS logic. Liability should also be jointly shared between brokers and ATS providers. This would reduce the possibility of regulatory arbitrage, consequently resulting in more efficient outcomes. Clear statutory rules should be introduced with specific provisions for the respective allocations of liability.

Principle 2: Regulatory Sandbox for Retail Innovation

Building on the United Kingdom Financial Conduct Authority’s sandbox which provides a controlled environment for firms to test innovative financial products, services, and technologies with real consumers under the Authority’s supervision, SEBI could permit community-driven or retail-developed ATS to go beyond mere family use in a controlled sandbox environment. This would effectively balance innovation with investor protection, addressing the critique highlighted earlier pertaining to the restricted application of self-developed ATS to personal or family use.

Principle 3: Algorithmic Transparency Index

Pursuant to the concerns highlighted earlier related to investor misperception and complacence, and information asymmetry, a disclosure metric should be introduced wherein white box algorithms are rated in terms of replicability and transparency and black box algorithms are mandated to be accompanied with detailed risk disclaimers. This would enable investors to then make more efficient and informed decisions.

Principle 4: Equal Access without Entry Barriers

SEBI should treat retail algorithmic trading as a question of equal access, alongside ensuring that safeguards don’t transform into entry barriers. By restricting retail self-developed ATS to the investor’s family, alongside placing heavy compliance responsibility on brokers, large institutional players dominate whereas smaller brokers and innovators face entry barriers. A better approach would be to design safeguards which preserve integrity without excluding innovative retail or small-scale participants from accessing the benefits of algorithmic finance. A judicial analogy can be drawn with National Stock Exchange v. SEBI. This case involved the National Stock Exchange introducing the co-location facility in 2009, pursuant to which brokers were able to install their servers in the Exchange’s data centre. In 2015, it was claimed that the co-location centre was subject to market manipulation because non-empanelled internet service providers were allowed to lay fibre cables on its premises for a few stockbrokers. In this context, it was clearly held that no class of participants should be given preferential treatment that undermines a level playing field. All types of investors should be given an equal opportunity of participating in the securities market and to achieve this end, exchanges and regulators like SEBI must ensure that although the market infrastructure fosters innovation and efficiency, it does not compromise systemic fairness or stability.

Principle 5: Bridging the Structural Resilience Gap

SEBI should introduce system-wide resilience safeguards, necessary to prevent occurrences like the 2010 United States Flash Crash, instead of solely focusing on an individualised conception of ATS providers. This would include the incorporation of stress testing mechanisms, kill-switches, adequate capacity requirements, and codified systemic controls, similar to MiFID II’s Article 17 stipulations, discussed earlier.

Principle 6: Supervisory Enforcement and Regulatory Capacity

To ensure that this proposed model is not merely aspirational, SEBI should explicitly build its supervisory architecture. This includes the incorporation of algorithm audits, escalation protocols, and structured penalties for non-disclosure or systemic risk creation, among other enforcement mechanisms.

Conclusion

SEBI’s circular is a decisive regulatory step that acknowledges the inevitability and benefits of retail algorithmic trading while simultaneously embedding SEBI’s statutory mandate of investor protection. Although it marks an appropriate step in the correct direction, amidst its benefits, there are a few drawbacks and loopholes that have to be filled. This has been done by the model conceptualised in this paper which suggests reforms to align India’s framework with international best practices. It is hoped that alongside the efficient implementation of these useful guidelines, the principles provided in the model are incorporated at the earliest to enable India to create a robust, fair, and innovative algorithmic trading ecosystem that protects retail investors without stifling technological progress.

*Manav Pamnani is a penultimate-year B.A. LL.B. (Hons.) student at the NALSAR University of Law, Hyderabad, and a qualified company secretary.