Safeguards needed to prevent another flash crash



‘As volatile as the stock market’ is a phrase in the English language; therefore, big fluctuations in the stock market are not treated as abnormal events. Major stock market crashes, however, are rare and invite a lot of attention and scrutiny since the consequences of such crashes can be far reaching. The Great Depression of the 1930s was preceded by a huge stock market crash. In recent times, the global stock market meltdown of 2008 led to the recession of 2009.

Technological advances and innovations have thrown up immense possibilities for all market participants in general, and for the capital market as an institution, in particular. However, technology has to be used with utmost caution, because, modern algorithmic trading, if misused either by design or default can unleash hitherto unimagined consequences. The flash crash on the NSE on Friday (October 5 ) has to be seen in this context.

On October 5 morning, the market opened normally and the Nifty opened at 5,815. At 9.50.58, the Nifty circuit filter got triggered and the index crashed 15.5 per cent to 4,888. The market was reopened by the Exchange with a pre-open phase at 10.00.22; trading resumed at 10.05.00 and the market functioned normally. The Nifty recovered at the end of day's trade to close at 5,747, down 41points. The Sensex closed 120 points down at 18,938. What caused this flash crash? What can be done to prevent such instances in future?

Unfortunately, we have limited facts on the flash crash in the public domain. According to the NSE, the crash was caused by an erroneous order executed by one of its trading members, Emkay Global. This erroneous order on 59 stocks resulted in multiple trades for an aggregate value of over Rs 650 crore.

These non-algo market orders have been entered for an erroneous quantity, which resulted in executing trades at multiple price points across the entire order-book, thereby causing the circuit filter to be triggered. It was reported that a young dealer mistook the value of the order for the number of shares and instead of selling stocks worth Rs 34 lakh, ended up selling stocks worth Rs 650 crore.

The flash crash triggered stop-losses for several investors for no fault of theirs. Emkay admitted that one of its dealers entered a wrong value while placing a basket sell order on Nifty stocks, on behalf of an institutional client. SEBI has initiated an investigation into the affair.

This affair raises some important questions: why did the Nifty crash by 15.5 per cent when the circuit filter should have been triggered at 10 per cent ? Was there a systemic failure? Were there no checks and balances at the broker’s end? If a 650 crore sell-order can cause market crashes of this magnitude, how deep are our markets?

Normally, brokers have caps for each terminal, dealer and client. Therefore, abnormal or freak trades automatically get flagged. Such safeguards at broker levels can prevent mishaps like these.

We need safeguards at the market level to prevent possible repeats of such instances. One safeguard can be narrowing of the price limits for order placement. Presently, for a large number of stocks, traders can place orders +/- 20 per cent of the previous closing price. This is too wide and can be narrowed to around 10 per cent. A deep market can absorb the shocks of such freak trades. In the US, there had been instances of major crashes caused by freak trades (on May 6, 2010 Dow Jones crashed by 998.5 points but recovered within minutes), but their market depth absorbed the shock facilitating the recovery. Since our markets are not that deep, we cannot afford to take instances like this lightly. Therefore, we have to give priority to deepening the markets with policy initiatives. We have to learn from mistakes and take corrective action.

(The author is Investment Strategist, Geojit BNP Paribas Financial Services Ltd. The views are personal)

Read the rest of this article by Signing up for Portfolio.It's completely free!

What You'll Get





This article is closed for comments.
Please Email the Editor