SEBI has voiced concern about retail traders using algo strategies they don’t understand well, often having purchased it from third-party vendors.
On August 11, a bunch of freak trades in Nifty Bank put options of 45,700 strike caused an over 90 percent drop in the premia briefly, before those trades were reversed. The identity of the brokers and the clients behind the trades could not be confirmed, but Moneycontrol has learned from derivative market players that the trades were put by multiple clients, all of whom were using the same algo strategy.
These clients ended up selling (or writing) put options for a premium of Rs 90 when the prevailing premium at that point was around Rs 1,300. The highest premia for that day was Rs 1,371. A put option is a financial contract that gives the buyer of the option the right to sell a specific asset, like a stock or an index, at a predetermined price, to the writer of that option. In this case, buyers of the Bank Nifty 45,700 put can sell the index to the writer of the put at 45,700, irrespective of the market price.
Derivatives players say the clients who sold the put options would have suffered a heavy loss because they would have ended up covering their positions at a much higher price. And because the futures market is dominated by high-frequency trading (HFT) firms that use algo strategies, the rival algos would have initiated their own trades to capitalise on the situation, increasing the losses for the people who put in the wrong trade.
For the writer of a put to be profitable, he needs to collect the highest possible premium that the market can bear. That’s because for option writers—both call and put—theoretically the losses can be unlimited if the price of the underlying moves in the opposite direction. Option sellers’ profit is limited to the amount they collect as premium.
It is not clear how the error happened, but derivative players say most likely the actual strategy would have been to sell Bank Nifty 45,700 strike call options which were quoting around Rs 90. Instead, the algos ended up firing orders to sell Nifty Bank puts of that strike price. Again, it could not be confirmed if it was a human error (known as fat finger trade in market parlance) at the time of feeding the strategy in the excel file driving the algo, or if it was a case of the algo having gone rogue.
In April this year, traders using the app of discount broking platform Shoonya were unable to put in their trades, and at the same time, saw their accounts showing trades they had not done.
In the erroneous Nifty Bank trade, market players said that besides the clients, other traders too would have lost money because of their stop losses getting triggered during the free fall in the premia. When a stop loss is triggered, the trade is automatically squared off.
Many traders use pair strategies wherein they initiate two different positions simultaneously depending on their outlook on a stock or the market, with stop losses built in. Sometimes traders take up simultaneous positions just to lower their margin obligations, which is a function of how much they lose if the market moves against them. By taking two bets in opposite directions at the same time, they can reduce the effective amount at risk in the event of an adverse price move.
“Because there are so many algo strategies at play, it is possible that a fall below a certain threshold would then trigger a fresh round of selling from algos looking to capitalise from the downtrend,” said a derivatives trader. This could have implications for many traders depending on where they have placed their stop losses.
At the same time, the freak Nifty Bank trade would also have resulted in windfall profits to many traders as they would have been able to buy the put at a throwaway price and then sell it quickly at a profit. Talk in the market is that some traders have profited to the tune of Rs 10-15 crore from the error.
Chinks in the system?
Derivatives traders are surprised how the freak order managed to find its way into the system bypassing the NSE’s Limit Price Protection (LPP) mechanism, considering that the price was 90 percent below the prevailing market rates. The NSE had introduced the LPP in October last year
Unlike in the case of stocks where sharp price moves are kept in check by intra-day circuit filters that set a floor and a cap, the LPP range in the options segment is dynamic and may change throughout the day based on market activity.
“Any incoming limit order placed beyond LPP range shall automatically be rejected by the Exchange,” said the NSE circular addressing the frequently asked questions about LPP.
During trading hours, the LPP is the simple average of trade prices of that contract in the last 30 seconds. For contracts that have traded in the last 30 seconds, the reference price shall be revised throughout the day at 30-second intervals.
“It is possible that had there been a single order, the system would have rejected it,” said a derivatives trader. “But because there were multiple orders at the same price coming in from different brokers, the system may have been fooled into thinking that the erroneous price was actually the fair market price,” said the trader.
In response to Moneycontrol’s query on how the system accepted such a low price, the NSE spokesperson said: “NSE has various pre-trade risk controls including Limit Price Protection for ensuring orderly trading, effective risk management and price discovery. As a process, NSE monitors all trades which may have taken place away from the prevailing market price.”
In the past SEBI has voiced concern about retail traders using algo strategies they don’t understand well, often having purchased it from third-party vendors. Using application progamming interface (API), traders can connect the algo strategy to their brokers’ software and get the trades executed.
“There needs to be a clarity on whether the services offered by the third party algo providers/vendors are in the nature of investment advisory services as the nature of their services includes providing strategies to the investors based on research and analysis done by them,” the SEBI consultation paper said.
“Since there is limited understanding with respect to the nature of services provided by various algo providers, brokers may obtain from their clients, details of nature and type of services taken from algo providers along with a confirmation as to whether the said services are in the nature of investment advisory services,” the SEBI paper said.