Search interesting materials

Friday, August 07, 2015

SEBI's new "trade annulment" policy

by Nidhi Aggarwal, Chirag Anand.

One of the main functions of organised financial markets is efficient price discovery. Under certain circumstances, prices fail to reflect the correct information. These include erroneous trades or market manipulation. Trades at non informative prices impose significant costs on other market participants and the public at large. There can be a market failure in the form of a negative externality imposed on other market participants by way of distorted prices.

As has been found around the world, incidents of erroneous trades and market manipulation have been experienced in India, which have adversely affected the functioning of markets. Erroneous trades can occur either due to the so-called fat-finger trades, trades caused by bad algorithms, or buggy software. With the growth of algorithmic trading and an overall increase in the number and pace of orders entering the exchanges, the probability of such occurrences may increase.

In an attempt to address these concerns, in a circular dated July 16, 2015, the Securities and Exchange Board of India issued guidelines for annulment of trades by stock exchanges. These new provisions are required to be implemented by the exchanges within one month of the issue of the circular. The policy comes after a discussion paper released by SEBI in October 2013 on the same topic. The paper had sought public comments on a policy that proposed trade annulment on occurrence of erroneous trades under "exceptional circumstances". The Finance Research Group at IGIDR had responded to the discussion paper with an analysis of the proposed intervention by SEBI. We argued that the costs of cancelling trades under any circumstances are much higher than the benefits envisioned. The published regulation suffers from the same flaws that were present in the discussion paper.

The new framework

Traditionally, stock exchanges have been empowered to annul trades either suo moto, or on receipt of requests from stock brokers. Under the new regulation, the regulator has directed the stock exchanges to consider a trade annulment request if it is sent within 30 minutes of execution of the trade(s). This time limit can be extended to 60 minutes in case of "exceptional circumstances". The stock exchanges are required to communicate the receipt of such requests in a time bound manner to all market participants and convey a decision on the request not later than the start of next trading day. With respect to examining such requests, the circular mentions:

"2.5. .. While examining such requests, stock exchanges shall consider the potential effect of such annulment on trades of other stock brokers/investors across all segments, including trades that resulted as an outcome of trade(s) under consideration."

"2.7. Stock exchanges shall undertake annulment or price reset only in exceptional cases, after recording reasons in writing, in the interest of the investors, market integrity, and maintaining sanctity of price discovery mechanism."

In addition, stock exchanges are required to define "suitable" criteria so as to discourage frivolous trade annulment requests from stock brokers. The regulator has asked exchanges to charge stock brokers an annulment application fee which could range between Rs 1 lakh to Rs 10 lakh. The circular also says that the exchange shall penalise brokers who put in erroneous orders.

Evaluating the new framework

A clear policy on erroneous trades is a welcome step from the regulator, since a) it addresses the market failure induced by hurting the normal course of price discovery process, and b) it removes the uncertainty for other market participants on how such trades will be handled in the event of their occurrence. However, as has been argued on this blog before (see here, here and here), trade annulment is a bad solution for numerous reasons:

  1. Moral hazard: Bailing out trading firms by cancelling trades introduces moral hazard. These firms are supposed to have adequate risk control systems. With such an option in place, trading firms will be less careful in building high quality algorithms or trading systems.
  2. Deters market stabilising trading strategies: In the case of extreme events, two types of trading strategies help markets to recover: First, strategies that place orders far away from the touch, and second, the presence of active traders who come into the market to take opposite position. These strategies are often high risk strategies. Trade annulment will deter these traders to enter the market during stress events. This will reduce market resilience. The idea should instead be to make markets more resilient to such shocks.
  3. Hurts other market participants: Several market participants (especially the liquidity providers) generally take positions across several asset classes (example: equity spot and futures). Cancellation of trades on one asset class leaves them exposed to risk on the other leg.
  4. Leaves space for regulatory capture: The current policy essentially leaves the decision for trade annulment to the subjective satisfaction of the exchanges. Such vague powers give the exchanges undue power and leaves space for regulatory capture -- a dominant group of traders or large firms with large share in trading will stand to benefit under such a system, while day traders and other small traders will lose out. A previous instance of trade annulment by BSE after a trading error shows how giving powers to the exchanges to decide on trade annulment can result in undesired outcomes.
  5. Moral hazard in trading strategies: It is difficult to ascertain whether trade cancellation requests are made in good faith. Once the law allows that fat finger trades can be annulled, rogue traders can take advantage of that rule to enter into trades and get them cancelled subsequently. As an example, imagine the following steps: Long nifty vol followed by a big fat finger trade on Nifty spot followed by closeout of the options position.
  6. Ambiguity in language: Section 2.7 of the regulation uses the phrase "in the interest of the investors, market integrity". These are very broad terms, and, as discussed above, open to subjective interpretation by the exchanges. Such words should not be used when drafting law.

The issue of erroneous trades is analogous to the issues of industrial safety, where a failure occurs when a firm fails to deploy adequate safety measures to prevent catastrophic events. The Bhopal gas tragedy in 1984 is an example of such a failure. The gas leak accident at Union Carbide India Ltd. caused several deaths and affected many thousands of people. The firm failed to deploy enough resources to have developed a safety mechanism to avoid such a catastrophe.

By this reasoning, trade annulment should be prescribed only if the exchange's order matching software fouls up, or if there was a systematic breakdown of connectivity to the exchange. In all other circumstances, trade annulment is a poor strategy. It messes up the sanctity of order matching processes and questions the finality of trades. It provides wrong incentives to the doers of such acts and sets all the wrong precedents. A trading firm should be mandated to place proper checks and balances in their operations, and in the event of such a failure, should be penalised to compensate for the damage caused to other participants in the system.

Taking the example of other jurisdictions, internationally, such trades are cancelled only if the price movements are beyond certain thresholds. Even though such a practice is not recommended and often debated, there are clearly laid rules on when a trade will be termed as clearly erroneous. This leaves no ambiguity with respect to an exchange's decision on when a trade will be annulled versus not. The question of such thresholds does not arise in India since Indian exchanges already have hygiene checks in the form of margin money, price bands, circuit filters, pre-trade order limits, that should not allow large price movements beyond thresholds. Exchanges are responsible for maintaining and running perfectly functional systems with all these hygiene checks well in place. These systems should guarantee functionality and is a service provided by the exchange to the investors. Any failure in a service should be considered a breach of contract between the exchange and the investor, and the exchange should be penalised subject to the terms of the contract.

Good governance practices

The Handbook on adoption of governance enhancing and non-legislative elements of the draft Indian Financial Code issued by the Ministry of Finance in December 2013 states the good governance practices for issuing regulations. Section 4.4 'Comments on draft Regulations' of the handbook says:

"The regulator has to publish all representations received, and at least a general account of the response to the representations while publishing the final regulations."

SEBI has failed to publish an account of the representations made and explain why the clear criticisms were rejected.


SEBI's new policy on trade annulment does not address the issue of erroneous trades appropriately. If anything, it only leaves a great deal of ambiguity. Public policy thinking and regulation-making should be done in a more rational way.

No comments:

Post a Comment

Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticism is perfectly okay; uncivilised language is not. We delete any comment which is spam, has personal attacks against anyone, or uses foul language. We delete any comment which does not contribute to the intellectual discussion about the blog article in question.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.