“Consolidation of brokers is happening, but it’s been slow to happen and is not as extreme as many had predicted,”
“There are too many algos,” said Squires. “They are not differentiated enough.”
These days, you could be forgiven for thinking that the public discussion around financial services is little more than a broken record, endlessly repeating debates on regulation and technological change. But behind the scenes, there is something much more interesting going on.
Buy-side firms are busy building a new market model – one that rejects the previous paradigm which was largely built by the sell side. The platforms that evolved over the last 20 years increasingly seemed to favour participants that contributed high volumes, regardless of their impact on the market. However, the buy-side are now collaborating with the sell side and with trading venues to build services that genuinely aim to help the long-only asset manager trade in larger blocks.
“When I first started trading, any significant flows came from a merchant bank or major asset manager, who had the block liquidity you needed. Today, participants are much more varied and you are unlikely to know who you are transacting with,” said Squires.
Squires has an interesting story to tell.
In November 1993, when he joined the industry as a UK equity trader for Mercury Asset Management, in some ways it was a different world. That month, the Maastricht Treaty took effect, formally establishing the European Union; in the UK, John Major’s government passed the Railways Act 1993, which privatised the country’s railways; in the charts were songs by Tina Turner, Whitney Houston and Bryan Adams.
Fast forward to 2016, and buy siders sometimes express concern about the rising proportion of automated and quant flow in the market (as high as 90%, according to some buy side estimates). But Squires insists that even in the 1990s, things were already well underway in that direction. “It’s definitely true that the rise of HFT is a factor in the more recent market structure evolution, but it really first changed a long time ago – I’m thinking back to 20 years ago (when order books replaced quote driven prices).” he said.
So how should the buy-side trader of today approach the market?
One response might be to withdraw from the market and trade less, maximising the use of tools such as dark pools to minimise market impact and protect against the ravages of aggressive predatory HFT. But for Squires, there is a more optimistic option available: he believes the buy side should take control of its own destiny.
What that means in practice is using technology and making the most of collaborative efforts such as non-profit buy-side and sell-side platforms Plato in equities and Project Neptune in fixed income. Both platforms have been jointly created by banks and asset managers to address the needs of the buy-side. They are part of a new wave of initiatives aimed at addressing the balance, in a market that has spent years catering mainly to the demands of quant flow, arguably at the expense of the long-term investor. Some platforms have introduced innovations not widely seen before – notably in the US, where the venue IEX has introduced “speed bumps” i.e. random delays in a bid to discourage HFT and prevent HFTs from gaming or otherwise exploiting asset managers.
“These platforms have added an ethical dimension,” said Squires. “They took action to make the experience better for the buy side, even if it was at the cost of market share. They took a commercial hit by closing themselves off to some forms of HFT, when they could have made a lot of money. I support an initiative that is willing to do that. These are the kinds of platforms that people should want to trade on.”
AXA Investment Managers has been heavily involved in the Plato initiative. Aside from a collaborative trading utility, the project will use the revenue generated via its trading to commission academic research, to identify better ways of executing trades, lowering cost and improving quality of data and processes to support execution. The aim is to “create a virtuous circle of improvement for the benefit of all market participants.”
It is not just the new venues that are of interest, either. Some existing platforms have introduced new features targeting the buy-side, such as Turquoise’s block discovery service (a platform ultimately owned by the London Stock Exchange Group), which was put together following extensive consultation with the buy side, and which promises larger order sizes than the lit markets.
“All these platforms share a theme,” said Squires. “A community of buy side and sell side collaborating within a utility structure. That’s the new dynamic, where the buy side are more empowered to determine market evolution without disintermediating their brokers .”
AXA Investment Managers sends orders direct to both IEX and Turquoise. The choice itself has a bearing on the relationship between the buy side and the sell-side, because it effectively neutralises the smart order routing aspect of trading execution. While this is arguably positive in the sense that the buy side is taking greater control over its own execution, it also requires a degree of independent-mindedness by the buy side, which can be a frustration for those who feel that other buy-side firms are dragging their feet.
“It’s strange that the buy side has been slow to use these kinds of platforms more as it illustrates a positive choice in the process of due diligence,” said Squires. “I want to see more support from the buy side for these kind of platforms. Unfortunately, it’s still hard for some of the buy side to take responsibility for their actions and feel it is safer to default to the historic standard which is to keep using the broker’s smart order router.”
“The excuses I’ve seen for why people are not using the new platforms can sound a bit flimsy,” he added. “For example, backing away from blocks for fear of being run over or having difficulty implementing the required technology. We love the two main ECNs. Our average execution size is close to €1m on one and €500k on the other. That’s a huge difference from the average market fill we get through a broker algorithm, even if we select a BCN and dark strategy.”
Speaking of brokers, broker consolidation is another theme that has increasingly become a topic of discussion in recent years. It is widely believed that the increased regulatory pressure on the sell-side since 2008 combined with declining broker commissions has hurt the sell-side, and none more so than the mid-tier brokers. The bifurcation of the market between large global houses that have advantages of scale and the smaller, local brokers that trade on their deep local expertise in each niche, for example Brazil or Turkey, leaves brokers in the middle struggling.
Different asset managers have differing views when it comes to broker lists. At one extreme, some firms have cut their broker lists dramatically, citing the unnecessary duplication of products such as algos, the lack of differentiation of services, and the need to cut costs. At the other end of the spectrum, some firms have refused to cut any brokers and are instead occasionally adding new ones in local niche markets, such as Latin America or the Far East. “Consolidation of brokers is happening, but it’s been slow to happen and is not as extreme as many had predicted,” said Squires.
One of the other major concerns for a head of trading today is the question of how the trading desk should be organised, and specifically, how the different asset classes should be managed, organised and traded. In theory, the choice is either to segregate the different asset classes and have them traded separately by specialists, or to bring them together and have multiple asset classes traded by the individual traders. Of course, there are an almost limitless number of possibilities between these two extremes.
According to Squires, his own personal trading experience suggests that performance improves, the greater the focus of the trader – so AXA uses dedicated traders. Traders specialise in the UK, for example, or specialise by asset class. These distinctions are important, because a generalist might struggle to know the difference between a UK small cap versus an Italian mid-cap, for example. However, that does not mean there is segregation.
“The benefits of cross-asset are mainly market intelligence and operational efficiency,” said Squires. “I think it is mutually beneficial if bond traders and equities traders increase their understanding of each others markets. For example I like seeing the credit traders chat with equity traders to gain insight.”
“It’s not a case of trading German bonds one minute and UK small cap equities the next. It’s about information sharing and using common processes to evidence controls. And the same practices can be applied. Benchmarking across asset classes is very useful – it’s great information. We go through the TCA each month and look at the trends or patterns: why brokers were expensive, which algos were effective and why. Analysis prompts that discussion and that’s very helpful, especially when the understanding is spread across asset classes.”
Technology has been a key component of AXA’s plans – particularly the use of FIX. AXA uses a FIX connection from the OMS to its fixed income RFQ systems to limit so-called ‘fat finger’ risk, in which the human trader makes an error, as well as to gain an audit trail both for internal and for compliance purposes. The company did the same for FX, primarily using the platform FXall. In addition, AXA expanded out its TCA system to cover FI and FX as well as equities. AXA has decided to outsource some of its fixed income benchmarking to gain operational efficiencies. However, not everything about the way technology has evolved is useful. “There are too many algos,” said Squires. “They are not differentiated enough.”
In terms of the mix of participants and the types of flow in the marketplace, Squires does not agree with some of the stereotypes about HFT and long-only asset managers, especially the depiction of a good versus bad struggle in the market. For example, one prominent buy-sider some years ago described HFTs as predatory “vultures” who pick off asset managers and undermine the ability of the buy-side to do its job.
“It’s hard to be objective about the optimal trading mindset,” said Squires. “You end up sounding either like a 90s block trading technophobe or a quant fanatic. We are all inclined to drift into old stereotypes – HFT is bad, dark pools are toxic and so on. But the market is not so binary and when you delve deeper those preconceptions get challenged. I met a large ‘HFT’ which described itself as an ‘old fashioned market maker’ which just happens to use smart technology. How can that not be a model worth trying to adapt to? ”
In the end, perhaps it ultimately rests on the buy side to make the best of the situation and work together to create a brighter tomorrow. Squires is adamant that the buy side should make the most of the opportunity presenting itself and be proactive. It’s not so much a question of disintermediating brokers, he says, as of taking the lead on issues like best execution, algos and harmonising TCA benchmarks, as well as pushing for a consolidated tape and lowering the cost of market data. There is a need for buy-side unity, however, because without that it becomes harder to push for positive change.
“It can happen in a meaningful way with a change of mindset but not in the current dynamic,” he said. “Projects such as Plato – with innovative ideas about market structure and a mechanism to develop ongoing improvements – should be in everyone’s interest. It may require the buyside having skin in the game but why not be able to evidence a tangible commitment to getting the best outcome for your clients?”
A new dawn for the buy side?
Paul Squires, head of trading, AXA Investment Managers
Head of Global Buy-side Research