Nyse drops curbs on program trading
29 October 2007 | 3979 views | 0
The New York Stock Exchange (Nyse) is set to drop a rule limiting program trading which was put in place after the 1987 crash on grounds that it is no longer effective in preventing market volatility.
According to a Reuters report, the rule change will remove current buying and selling curbs - called "trading collars" - on brokerages using computer-assisted program trading when the Nyse Composite Index moves up or down more than two per cent.
These "trading collars" had been in place since 1988 and were triggered 17 times this year, according to the report, which cites a SEC filing. Volatility is neither restrained nor enhanced by the collars, says the exchange.
Nyse defines computer-assisted trades as the buying or selling of a basket of at least 15 stocks from the S&P 500 Index valued at $1 million or more.
The current rule only applies to some index arbitrage trades on stocks in the Standard & Poor's 500 Index executed at the exchange. But trading strategies have multiplied in the past several years, says Nyse, and Index arbitrage is an increasingly smaller part of daily stock trading.
However the move comes as questions arise over the suitability of algorithmic trading in volatile market conditions. According to a Reuters report, traders say current volatility is showing the limitations of algo trading in equities and FX.
This seems to the be case for Morgan Stanley which reported a $480 million third quarter loss from its in-house equities trading desk, which was using computer-generated models to drive returns.
Hedge funds using algorithmic trading methods are also likely to have been hit, says the report.
Furthermore, a large number of trader and hedge funds are thought to have ditched algorithmic methods and reverted to more traditional trading procedures when markets were at their most volatile in August.
The Reuters report, which cites bank sources, says this may have been because the technology available was unable to deal with the immense volume of tickets.