Don't Blame Algos for Market Crashes

Algorithms cannot be blamed for stock market crashes; traders need to understand algorithmic trading is the future

Pravin Palande
Updated: Apr 25, 2012 06:32:36 PM UTC

The last two years have seen many incidents where computers have been blamed for extreme volatility in the financial markets across the globe. Indian markets have also seen some incidents where algorithms have been blamed for volatility. The most recent one happened on April 20, 2012.

Last Friday, the Nifty was down by 7 percent or 300 points which panicked the equity markets throughout the week. It was also the day when Infosys futures went down by 20 percent in a single day before it recovered. Traders blamed algorithms and called this the Indian version of the Flash Crash that was experienced on May 6, 2010, when the Dow Jones crashed by 1000 points or 9 percent, but managed to recover within minutes.

The official report by the SEC blamed a single large trade in E-mini S&P 500 Futures by a mutual fund that led to large buying and selling of the same contract by high frequency traders due to which the E-mini price was down 3 percent in four minutes.

In the Indian context, the blame has fallen on a broker who put in large orders on the Nifty contracts without specifying price limits. These were large orders that ended up as market orders and due to the lack of specific price details, these trades were executed at the best available price. A mint report (http://www.livemint.com/2012/04/23204454/Another-flash-crash-another-a.html) states that it was a foreign broker who put this order and it was unlikely that this order was sent manually into the system.

The name of the broker is not yet public. But another story goes that the broking firm is basically a prop trading house which operates out of Mumbai and specialises in derivatives products. It is believed that this firm ran an options strategy which led to the present crisis. It was an arbitrage between futures and options where orders were executed wrongly.

Though the stock of Infosys also crashed by 20 percent on the same day and the exchange wants to play it safe by not connecting the stock future and the Nifty futures, it cannot be dismissed that this could have been a part of the strategy. In any case, Infosys has the highest weightage at 8.47 percent in the Nifty Index and this connection in the present crash cannot be ignored.

Sandeep Tyagi of Estee advisor runs an arbitrage fund out ofDelhiwhich is completely based on algorithms. He feels that algorithms are being unnecessarily blamed for the present situation. When the markets started to crash on Friday afternoon, his algorithm did not go overboard to sell the Nifty when the market witnessed a free fall. “I’m running a fund that is based on algorithms. We did not really experience any extreme reaction from our fund so I won’t really blame the algorithms. This looks more like a manual error to me,” he says.

Tyagi feels that if all brokers or traders follow a simple checklist for algorithms, then situations like the present flash crash can easily be avoided. He puts thrust on price, value, open positions, real time risk and profitability and number of orders in a given time frame as his main parameters in the checklist. More importantly, this checklist needs to be managed by someone responsible. According to him, such simple measures can go a long way to create a good trading environment where algorithms will not be unnecessarily blamed in case of market crashes.

Lokesh Madan of Algo Trading India LLC feels that the problem was caused due to human error. In a Business Standard story (http://business-standard.com/india/news/sebi-to-probe-flash-crashes/472305/) he states that the volatility was caused by someone who tried to change the variable settings of an algo programme.

The future of markets is algorithms and high frequency trading. Indian exchanges understand this truth and are getting ready to operate in this environment. Many software vendors and high frequency traders are eyeing emerging markets such asIndiaas they believe there is money to be made out here due to the fundamental growth in the economy.

Algorithms should not be considered as the ghost in the machine. It should not be taken as the sole factor that causes market crashes as it only reacts to situations. Many times these are just bugs in the system. These bugs can be solved by stress testing and market simulators. However, there will still be errors and market crashes. But going back on algorithmic trading will be difficult. We will have to accept this future and tread carefully.

Keeping a check on algorithmic trading is a continuous process. The regulator obviously knows who was responsible in placing the orders that caused a flash crash on the Nifty. If an example is made out of the firm which caused this problem, it will only help other players to improve and understand algorithms better.

On a lighter note, we should be happy that the Flash Crash did not happen a week earlier or the superstitious trader would have blamed Friday 13 as the reason for the crash. If the incident had happened on May 6, the May 6 Flash Crash virus that strikes every two years.

Readers can also go through this link http://newsfeedresearcher.com/data/articles_b12_2/bats-trading-exchange.html . The story talks about BATS Global markets one of the biggest electronic exchanges, which had to stop trading its shares due to technical problems or errors in the trading system. It was the first day of trading for its IPO and everything went wrong. BATS had to cancel its IPO. It was again a Friday. BATS has now solved its problem and remains true to the field of electronic or algorithmic trading with the lowest error rates on any exchange.

 

The thoughts and opinions shared here are of the author.

Check out our end of season subscription discounts with a Moneycontrol pro subscription absolutely free. Use code EOSO2021. Click here for details.

Post Your Comment
Required
Required, will not be published
All comments are moderated