Automation “Not the Holy Grail” For Buy-Side Trading Desks

As asset managers across the industry ramp up their preparations for the arrival of Mifid II, automation has been a top priority across many front, middle and back offices on the buy side.

The benefits – reduced costs, the removal of manual processes and streamlined workflows – are both apparent and enticing for many asset managers, particularly when it comes to regulatory-related operations, but not everyone believes automation is necessarily a golden ticket to success.

“There isn’t any right or wrong about seeing more or less automation, as this very much comes down to strategy,” said Eric Boess, global head of trading at Allianz Global Investors. “Some people think that automation is the Holy Grail and you can get rid of most of the human traders and just have one monitoring a laptop. I don’t think that’s the way forward.”

Boess was careful to stipulate that while automation is not without its advantages, a “sense of human oversight” is necessary to maintain cost-income ratios for a viable business model. One of the drawbacks of implementing what Boess referred to as a “monolithic IT infrastructure,” is a loss of flexibility, something that multi-asset trading firms like Allianz benefit from in changing or volatile marketplaces.

“It might work for some firms – CTAs (commodity trading advisors) or very quant-driven or high-frequency strategies – but for us as a broad-based strategies, multi-asset firm, I definitely need the flexibility of human traders to change [their trading decisions] depending on which strategies we are currently adopting,” Boess said. “Automation is needed, but it’s not the Holy Grail.”

Erik Koenig, global head of trading at GLG Partners, part of London-based Man Group, Europe’s largest hedge fund entity, agreed that while there is value in automation to optimize processes, as an active fund manager the human element of trading is a vital component, due to the level of informational feedback derived from the trading desk when searching for liquidity.

“One of the biggest problems that we face is liquidity and the hope is that in the post-Mifid world we will actually be interacting with the right people to make sure we can get that,” Koenig said.

Automation will still be top of many wish lists across the buy side, particularly when it comes to optimizing problematic areas in the middle and back offices; however, the front office is still where the human trader holds sway unless there is particular call for an automated trading strategy.

“Stupid” Data

One part of the Mifid II regulation that is of particular concern for asset managers is the change to best execution requirements, which in turn is significantly impacting how transaction-cost analysis (TCA) is utilized, as firms adapt to demands from regulators and end-investors for more transparency.

Both Boess and Koenig asserted that through the establishment of a holistic best-execution committee, there was no need to “reinvent the wheel” in what remains a “grey, opaque area,” regardless of how a firm chooses to approach its TCA adoption.

“We have in-house and external TCA,” said Koeing. “Our view is that we want to overlay both to make sure that we have the purest outcome, that the facts and information overlay there.”

Boess meanwhile asserted that he preferred that TCA providers were external parties, as it provides a minor level of credibility, as opposed to doing the job in-house, in discussion with external clients and internally between traders and portfolio managers.

“There has to be a certain amount of proprietary intellectual capital going into this, because best execution is not a one-size-fits-all exercise,” he said. “Trading is always the extrapolation of investment strategy; hence what is best execution for one strategy does not necessarily work for another, so you have to ask the right questions to a TCA provider and the data. Data itself is stupid – you have to ask the right questions to get the right answers.”