The number of price updates is increasing as the hold time for quotes decreases.
This is partly the result of firms demonstrating their compliance with new last-look guidance, according to Kurt vom Scheidt, global head of FX at Saxo Bank – though faster market data also push the market in that direction.
“Firms are keen to adhere to the code’s guidance on last look, especially when that is defined to be an extra hold time for an order accept/reject decision beyond an immediate price/risk check,” says Vom Scheidt.
“Allowing the immediate check supports tighter spreads in a fragmented, decentralized global FX market, which can be beneficial when liquidity is appropriately managed and accessed.”
He adds: “Liquidity distributors and consumers are now working through their relationships, inclusive of differing potential technology constraints, to find the right balance to achieve minimal market impact of execution against a price stack where shortened hold time is a normalized aggregation factor.”
The Global Foreign Exchange Committee has indicated it will establish three new working groups at its next meeting in South Africa in the second quarter of 2018 – expected to address last look, cover and deal trading and disclosure.
Last look has long been one of the most controversial aspects of the FX market. It originated in the days of voice trading, when traders needed their brokers to hold prices for a short period while they checked them against others in the market – though it was not called last look until relatively recently.
Traders inevitably have to put quotes out to multiple liquidity providers simultaneously, despite having no intention of going through with every trade, because there are so many venues on which they can trade, offering different prices.
With the advent of electronic trading, and trading times being measured in milliseconds rather than seconds, last look has in some cases evolved into a form of latency arbitrage, with the possibility of exploitation for both counterparties to a trade.
In December, principle 17 of the code, which relates to last look, was revised after further market consultation. The changes provided that market participants should not undertake trading activity that utilizes information from the client’s trade request during the last-look window.
It also distinguished between this potential bad practice and more legitimate trading arrangements, such as cover and deal, where brokers trade for their clients on an agency basis. The former wording had cast doubt on the viability of a growing number of trades conducted on an agency, rather than a principal, basis, but the revision has clarified their position.
Christopher Matsko, head of FX trading services at Portware, a technology vendor owned by FactSet, says: “By removing the word ‘likely’ from the language and offering carve-outs of sorts for best practices in cover-and-deal workflows, I think that 90% of the market’s concerns will be alleviated.”
There were certainly voices during the discussions around the code calling for last look to be banned altogether. Others feared instead of ending the practice this would simply drive it underground. Such a measure would also be inconsistent with the principles-based code.
Most agree the principles approach is best, and that last look is defensible as long as the rules of engagement between the counterparties are clear and adhered to.
|David Clark, EVIA|
David Clark, chairman of the European Venues and Intermediaries Association (EVIA) – formerly the Wholesale Markets Brokers’ Association – says: “Last look is something that needs to be negotiated between counterparties.”
Of course, ensuring institutions disclose their policies around last look is only half the battle. Ensuring they do what they say they are going to do is something else, and a voluntary code, not enforced by any regulatory body, can only do so much.
However, participants agree the reputational risk for institutions not adhering to the code will be huge.
Reports are that large buy-side firms have already started asking their banks whether they comply with the code and the pressure will only increase after the deadline for signing up in May. The pressure from central banks could be even blunter if they decide to cut off non-compliers from their liquidity flow.
The expectation is that most leading institutions will sign up, despite only a handful having done so to date. The assumption is they are using their remaining months to ensure their legal and compliance departments have fully assessed the implications, and whether they intend to sign up at the institutional level, or multiple times at a more local level.
Addressing some of these questions will help ensure institutions get the most out of the code, rather than just ticking a box.
|Brad Bailey, Celent|
Brad Bailey, research director in the capital markets division at Celent, says: “Improvements in things like transaction cost analysis that highlight good execution will make it easier for firms to show – both internally and to investors and counterparties – that they are managing their best execution process and complying with these components of the code.”
A growing number of institutions now have systems that can analyze their market impact by liquidity source, showing where the market has moved away from them pre-execution, and the cost of being accepted or rejected in a trade, all of which will also incentivize good behaviour among counterparties.
Roger Rutherford, COO of ParFX, says: “The direction of travel is towards openness and transparency. That is why we think it is so important that traders know the identity of their counterparty post trade, as they do on our platform.
“It provides clarity and allows them to truly analyze the behaviour of their liquidity provider.”
The creation of the working group looking at last look suggests there will be more changes to come. Just as the ACI’s model code needed a refresh to account for market changes since it was first published in 1975, so the new global code will also start to age.
It will be subject to review every year, with a bigger review every third year, with rules around last look strong contenders for regular updates.
Saxo’s Vom Scheidt says: “Standardization of disclosure could be one way to make it easier for clients to compare the policies of their trading counterparties and assess their behaviour when it comes to matters like last look and order handling.
“Market participants can use very different language in their disclosures, which can also be quite lengthy, so it may be challenging for some clients to recognize what the differences actually are.”
Portware’s Matsko says: “Banks are now being transparent around the circumstances which they use last look, but the disclosure statements can range from five to 40 pages.
“Each of our buy-side clients have 10 or more counterparties in their liquidity pool and having to understand how each counterparty is handling your orders via these complicated disclosures can be a resource-consuming, if not difficult, task.”
EVIA’s Clark says: “The debate around last look is unlikely to go away. We will be discussing it as long as technology and market infrastructure continue to advance.”