Riding the Wave: Speed, Algos and Consolidation
By Ivy Schmerken, Editorial Director
A wave of consolidation has hit high-speed trading firms and brokers in the past few months, and many are blaming low volatility and weak trading volumes in the face of rising compliance and market data costs.
Starting with electronic market-maker Virtu’s acquisition of rival KCG in April, Two Sigma Securities buying options market-marker Timber Hill in May, and Cowen gobbling up Convergex a few weeks ago, trading firms are reassessing their business models, including the need for speed.
“While speed is still important, and is the way to capture liquidity, high-frequency trading firms are moving away from speed and diversifying into longer duration strategies,” said Rob Shapiro, former Chief of Staff at Bloomberg Tradebook, speaking at the Intelligent Trading Summit organized by A-Team Group, held on June 8. The two culprits are regulatory compliance and the increase in market data fees, he said.
“The HFT firms have long trafficked in their doing their own thing. Because their models are no longer working, they have become service providers because they need to subsidize their tremendous infrastructure,” commented Shapiro. Now high-speed trading firms are going after the buy side and the sell side as service providers of electronic trading for high performance tools, said Shapiro.
However, not everyone agrees that consolidation is a bad thing.
“The mergers and acquisitions that we’re seeing now are not a sign of weakness,” said Bill Harts, CEO of Modern Markets Initiative, who pointed to a misunderstanding in the press around the role of electronic market makers.
“It’s not that the algos weren’t working. It’s that the smarter operators want to extend those algos, extend their capital to more customers, more asset classes,” said Harts.
“Certainly times are hard and a lot of that has to do with costs, simply, exchange costs,” noted Harts. “Overall, whether you are a bank, a high-frequency proprietary trader, or an exchange, all the people in the trading ecosystem are constantly thinking about costs,” said Harts. “It’s the trade-off: is one microsecond extra worth it, especially in the context of exchange costs? “Every dollar that is spent on exchange fees or connecting to exchanges at high-speed is one dollar less that can be available to go into investors’ pockets,” he said.
Cost pressures are leading banks to consolidate trading desks across multiple asset classes, said technology executives on the panel, Managing Change and Disruption: Trading in the New Normal Environment.
The pool of U.S. equity commissions earned by brokers contracted last year by more than 11% to $8.4 billion in the first quarter of 2017, from $9.7 billion in Q1 2016, according to a report by Greenwich Associates.
To Reduce Costs, Banks Leverage Algos Across Asset Classes
“What I am experiencing is consolidation across the business lines. No single business line can go off and build a massive low-latency infrastructure and support that with revenues. They have to collaborate with the other businesses,” said David Winig, Global Head of Market Data Services and Electronic Trading Infrastructure at JPMorgan Chase & Co.
Rather than build proprietary execution systems from scratch, large banks are working with third-party vendors and developing knowledge-sharing relationships.
“Really these vendors are becoming extensions of the banks,” said Haim Bodek, managing partner of Area 11 Research. Launched in 2017, the firm is focused on providing high-frequency trading tools and strategies to institutions and boutique broker-dealers. “We think of ourselves as the Intel-inside,” said Bodek, referring to the chipmaker’s popular slogan.
On the futures and options side, a few years ago Citi leveraged third-party providers to create a best-of-breed solution with algorithms and analytics on the back-end, said Andrew Keane, Global Head – Listed Derivatives Algo Trading at Citigroup Futures & OTC Clearing.
“It really helped us grow our business, invest wisely and not have this huge overhead in electronic trading in futures, options and switches, which is substantial,” said Keane.
To differentiate itself, Citi has put together certain components on the back-end, as a whole package, especially around its algos that are focused on best execution, he said. “When you go into the CME to trade a 10-year [Treasury bond futures], the customer can see “this is how your algo performed in the market,” said Keane.
Buy-Side Firms Turn to Automation
Another seismic trend is the shift of assets from actively managed to quantitative and passively managed funds. The context of traders’ orders from hedge funds and traditional long-only investment management firms is different than the orders of quantitative traders, said Shapiro. “You get massive program trades, cash constraints, dynamic risk optimization, and poor commissions,” he said.
One of the areas firms are questioning is the value of speed.
“From an automation standpoint on the buy side, everyone is looking at cost, return on investment, and asking, ‘Do I need to be high frequency?’ ” said Keane. Some buy-side firms are asking if they can get away with using the broker’s DMA FIX infrastructure, he said. “Will the broker’s DMA FIX infrastructure be faster than high frequency trading?” No, but it depends on the game they are in, said Keane.
“For quant funds, the bread-and-butter is their alpha-generating signals. It’s surprising that many of them are automating simple spreads. They are automating across the board,” said Keane. “With fewer people on their side, the return on investment is higher and they’re adding more alpha and then giving back to their clients,” said Keane.
Cost-Cutting: Sharing Technology from Equities to FX
In response to the buy side embracing automated trading, the sell side is looking to leverage pieces of their technology across different asset classes.
“As more markets electronify, we’re borrowing technology from our equity cash desk and moving it over to our rates desk,” said Citi’s Winig. The FX desk is also trying to share this technology, said Winig. “Along with that, we’re trying to automate rather than take up trader’s time and effort to price smaller deals,” he said. This consolidation across desks “is actually good from a software engineering and management standpoint,” he said.
However, cutting costs is difficult in a bank where different asset classes want to build their own technology. “Some of these builds across multi assets are big builds,” said Keane. “It’s not one-size fits all,” he added.
What JPMorgan Chase has done is leverage other pieces of the bank’s technology across equities and FX and work with other third parties such as exchange data providers. “For example, we borrow the equities stuff, not just the algos, to build best-of-breed,” Keane said.
Meanwhile, after global stock markets rallied to new highs for the first half of 2017, there were signs that turbulence in the last week of June could be a harbinger of higher volatility, reported The Wall Street Journal in “Global Stocks Post Strongest First Half in Years, Worrying Investors.” Uncertainty over Brexit, U.S. corporate earnings, and central bankers adjusting their bond-buying programs, are some of the factors that could lead to choppier markets.
Even if volatility picks up, cost control will continue to consume mind share on Wall Street.
How FlexTrade Can Help
Past Blogs Related to Algorithmic Trading Issues