Panellists at TradeTech FX in Amsterdam addressing motivations for choosing between different algo offerings agreed that when it comes to selection, how liquidity interacts is a key factor.
Buy-side panellist Gordon Noonan, head of FX and rates trading at Schroders confirmed that the firm only uses bank algos as the team seeks to “do more with less people” and highlighted the importance of connectivity and full functionality.
Similarly, Patrice Guesnet, co-head of fixed income trading at Pictet Asset Management, also shared that many bank algos are used at Pictet, explaining: “Major algo providers can take an algo and customise it to your needs […] connectivity is very important for us as well – the banks already have all the different types of clients we want and they can basically internalise a lot of flows.”
With this in mind, Kimiya Minoukadeh, head of eFX quant trading at ING, highlighted that though connectivity is indeed a key driver, ultimately if the name of the game is to get a better price at the end of the day, more is not necessarily beneficial.
Minoukadeh shared that on ING’s side the firm internally used algos before beginning to offer this out to clients also, having seen the benefits first hand. However, she highlighted, though bank algo benefits are clear, less can often be more.
“Obviously, there’s been a lot of conversation in the past years about the fragmentation of the market so that does give a big benefit, but if the ultimate reason to use algos is to get improved slippage, a better price, etc. I think everybody is aware more is not necessarily better.”
She added that it is important not to forget the prevailing fact that often one does not know how risk is managed behind the scenes.
“You really have to do a lot of the analysis yourself and that’s what we do […] we advise clients whether they could hit our internal pool or hit the external venues as well […] we’re able to have that visibility on where there is impact and we can sort of manage that for them as well.”
Jeremy Smart, global head of distribution at XTX Markets, concurred, asserting that how liquidity interacts is the bottom line.
“It’s not about understanding whether you’re connected to lots of people, but how your liquidity actually interacts with all of those venues and all of those clients […] for example, understanding how you show skew, to which clients you show skew, to which ECNs you show, and how you interact with those markets – that is what’s actually critical.”
Addressing where firms are finding algos proving most effective, Guesnet was quick to highlight emerging markets, asserting that that is where Pictet uses these most and where the firm “sees the most value”.
He added: “Even with a small size you see you can tighten the spread.”
Smart agreed, but added that from his perspective, algos can definitely be equally effective across emerging markets and the G10, with the key being to look specifically at how individual currencies trade with the markets, and focusing on where skew if being shown.
He added: “The amount of alpha you have as a market maker is also very important in the process of avoiding adverse selection in the market and actually picking good trades to do, rather than just trades to do.”
Read more – Beyond the Data: Long-only managers more optimistic than ever when it comes to their algo providers
Elsewhere, the panel addressed the differentiating aspects of algo providers, and what stands out when it comes to selection.
“A few things but data is important, including pre- and post-trade analytics […] every different provider has their own opinion when it comes to their approach,” said Guesnet, while Noonan affirmed: “For us with algos, and the smart order router and the analytics the different banks provide with regard to access – that’s key for us.”
“Another thing we lean on as well is the banks natural franchise which is quite important.”