UBS uses iPad to let clients help design trading algorithms

April 19 (Bloomberg) -- UBS AG will let clients design trading algorithms on Apple Inc.’s iPad, putting tailor-made tools in their hands more quickly in an effort to capture equities market share, according to executives at the Zurich- based bank.

The move extends a trend toward customization that began about five years ago. Algorithms, used by U.S. asset managers to trade 19% of their volume, according to Greenwich Associates, are automated tactics that break larger buy or sell orders into hundreds or thousands of smaller pieces over a specified period to mute price impact and mask activity. Customization allows traders to fine-tune automated strategies.

“We can now build an algorithm in days and we won’t have the lost-in-translation problem about what will ultimately be built -- what’s lost in phone calls and e-mails,” Owain Self, global head of algorithmic trading at UBS, said in an interview. “We’re eliminating the slippage and improving execution.” The electronic strategies will be “more client-centric, more scientific and ultimately more scalable,” Self said.

UBS’s Direct Execution trading group, the 100-person unit in the bank’s global equities division, will extend technology- driven tools to customers as part of a new service called Quant on Demand, said Self, who co-runs the electronic group with Charles Susi. The first product is the QUOD Studio offering of individual algorithms in equities and futures, available today to clients in the U.S., Canada and Europe. It will be accessible to customers in the Asia-Pacific region later this year.

‘Way Less Scary’

The bank has offered individualized algorithms to clients for several years. The iPad system is more visual and intuitive, Susi said. “It is super-powerful and way less scary,” he added.

“It allows clients to have the perfect algorithm for them,” Susi, managing director in equities, said in an interview in UBS’s New York office. “They get the benefit of our research lab. We make clients happy and the ultimate goal for us is to get more clients and market share.”

The bank’s electronic trading group has built 25 to 30 algorithms using the QUOD Studio tools for at least 20 clients in the U.S., excluding UBS’s own trading desks, since late 2011, Self said. Until now customers haven’t been able to use the iPad application to help devise their own algorithms and the bank has done it for them to make sure the process worked, he said. Managers and quantitative analysts will begin visiting clients in the U.S. and Europe today to show them the product, he said.

Performance Targets

Traders use different algorithms to achieve varying performance targets, denoted by the average price per share, based on factors such as the stock’s daily volume, market capitalization, volatility and moves during the day. They often adapt the strategies based on how aggressively they want to buy or sell, broader market conditions that affect individual stock behavior, and activity in the shares.

Goldman Sachs Group Inc., Credit Suisse Group AG, Bank of America Corp., Barclays Plc and others offer algorithms that can be customized or built to clients’ specifications. Others such as Orc Group AB and Tbricks AB, both in Stockholm, and Actant AG in Zug, Switzerland, are specialist firms that allow asset managers and professional traders to develop their own tactics.

Asset managers traded 19% of the dollar value of their U.S. equity orders through algorithms in the year ending mid-February 2011, according to data from Greenwich Associates. Institutions paid their brokers an average commission of 2.69 cents per share while hedge funds paid 2.29 cents, the Stamford, Connecticut-based research firm said in a June report. Securities firms in recent years have often charged less than 1 cent per share for so-called low-touch algorithmic trades.

Page 1 of 2 >>

Copyright 2014 Bloomberg. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.

comments powered by Disqus
Check out Futures Magazine - Polls on LockerDome on LockerDome