© The Financial Times Ltd 2015 FT and 'Financial Times' are trademarks of The Financial Times Ltd.
April 7, 2011 4:32 am
Deutsche Bank has unveiled a new equity trading algorithm that mimics high-speed market-making strategies, as automated trading continues to constitute a larger part of market activity in spite of concerns expressed by regulators.
The new algorithm, a revised version of its previous Stealth product that has been in the works for a year, will use models that predict price changes over a period of microseconds and other complex quantitative tools to execute trades for institutional clients, such as mutual or pension funds.
“All the angst traders have had over the last three or four years has arisen from a mismatch between humans taking liquidity and machines supplying it. Until buyside traders start using the same quant models, they’re not going to see an improvement,” said Jose Marques, global head of electronic equity trading at Deutsche Bank.
Traditional investors, such as mutual funds and pension funds, have feared that high-frequency firms are taking the best prices from public markets. In a broad survey of asset managers by the Tabb Group last year found that 49 per cent of firms said the impact of “high-frequency trading” was their top worry.
That has spurred the development of dark pools and block-trading venues, such as Liquidnet, ITG’s POSIT, Pipeline, and BIDS Trading, that enable institutions to trade with each other rather than interact with high-speed market makers in public markets.
However, banks are increasingly proposing that they can execute such trades in public markets by using algorithmic tools that trade large blocks of shares in small pieces, so as to minimise the impact the order has on market prices and reveal as little as possible to high-speed liquidity providers.
“Market makers deserve a fair return for the risk they take, but they don’t deserve outsized returns. Levelling the playing field by bringing equivalent trading tools to the buyside restores equality to the marketplace,” said Mr Marques.
Deutsche began an aggressive expansion of its electronic trading group in the US two years ago. It ranked 11th in US trading commissions generated in 2010, according to Tabb. Mr Marques joined last year from Credit Suisse. Deutsche ranked first in a Bloomberg Markets magazine survey of best prices for clients’ stock trades last year, supplanting Goldman.
At the same time, scrutiny of algorithms by US regulators has intensified after a simple algorithm was at the heart of the “flash crash” report written jointly by the Securities and Exchange Commission and Commodity Futures Trading Commission.
Mary Schapiro, chair of the SEC, has proposed checks on algorithms before they are used, though no formal rule has yet been written. The SEC has also banned so-called “naked access” to markets by traders, and now requires basic checks on trades by brokers before they are sent out.
The CFTC has convened a technology panel, which recommended that clearinghouses, exchanges, traders and brokers should jointly oversee algorithms, and that all trades should have a “kill switch”.
Brokers-dealers, whose results have suffered in recent quarters when trading volumes and volatility fell to three-year lows, have been responding to changes with new technologies to entice buyside clients.
Earlier this year, Goldman Sachs launched its BlockStrike algorithm, which seeks to trade large blocks of shares using order types that do not run afoul of proposed SEC restrictions on what trades can be shown to “dark pools”, which are non-exchange markets that do not display orders.
Credit Suisse unveiled the Light Pool electronic marketplace, which will display orders but preference block trades by filtering out traders whose orders are quickly filled. In March, Deutsche also said it had created computer chips that can complete SEC-required pre-trade risk checks in microseconds.
Please don't cut articles from FT.com and redistribute by email or post to the web.