Algos: Do They Need to be Independent?

July 7, 2016

Algorithmic trading is under the microscope with U.S. regulators looking into the roles and responsibilities of algos and their providers. It once again raises the important question, who is best placed to build algos for bank clients?

The choice seems clear for banks: build your own algos, or outsource to an independent provider. Not so fast. Mizuho is one of the firms that believes it’s possible to get the best of both worlds, partnering with a quantitative trading technology provider and customizing it as-needed for their own purposes. It may be a smart approach, especially given the slip-ups that have taken place with several of the dark pools at other major Wall Street banks.

According to Matt DeSalvo, Mizuho’s head of U.S. Equities, the firm selected Pragma Securities and after a six-month process went live earlier in 2016. “Even if Pragma provides its product to another broker dealer, each dealer is able to put their own fingerprint on it,” said DeSalvo in a May 2016 interview with Markets Media. “We’re using the same thing, but can select different routing destinations or they allow us to tweak it so that it is customized to us.”

The benefits of algos are clear-cut for investors: they enable them to efficiently execute and route trades, while providing access to a range of trading strategies. Of course, this all depends on which strategy is used, which is where the science of algos takes off. Numerous variations on the quantitative strategies underpin these twenty-first century trading tools, some of which are developed internally by banks, others of which are marketed by third-party providers of quantitative tools.

While the rise in the use of algos is fairly recent, it has been dramatic. In fact, algorithmic trading now represents 80% or more of the entire U.S. equities market, according to industry estimates, of which a high percentage is likely to be from high-frequency trading firms. But algos are extremely valuable for long-term institutional investors as well, as they provide a mechanism to help fulfill best execution requirements while increasing performance by limiting price slippage.

For banks like Mizuho, the use of algos is not the question, it is how best to deploy them to clients. There is a cost and time associated with developing and maintaining proprietary algos, as well as potential reputational risk. On the other hand, using third-party tools can eradicate competitive advantage, and still has a related cost. Mizuho elected a hybrid option, but there are plenty of other banks that choose the homegrown route and likely many more that offer and run internal and third-party models side-by-side.

While the use of algo trading in the equity markets, as the successor to the idea of program trading, dates back to the late-1990s, its deployment in other markets is more nascent.

In fixed income – a largely over-the-counter market – there is still no widespread use of algos by the buy-side, although there is an opportunity for this to change on the back of real-time dealer order books, which have only recently been made available to the buy-side. In the inter-dealer markets, algos have been employed for several years, however.

Like the inter-dealer market, exchange based futures markets are well-suited and have been seeing a high level of demand for sophisticated algo strategies for a number of years.

While people may argue about the best type of algo, or how best to source them, in an uncertain market, one thing’s for sure: algos are here to stay.

Simon Hylson-Smith is a former financial industry editor and currently CEO of Paragon.

Simon Hylson-Smith

CEO, Paragon

Simon Hylson-Smith is a former financial industry Editor and currently CEO of Paragon.

Back to top