Galen Stops charts the ups and downs of FX prime brokerage over the past 20 years and looks at how this segment of the market might be ripe for innovation going forward.
Ask anyone who has been around in the FX market for the past 20 years to list the key developments that have shaped the industry and inevitably FX prime brokerage (FXPB) will feature on this list.
PB had been around prior to Profit & Loss’ launch in 1999, but it was only after this date that it, like the FX market itself, began to go electronic and blossomed into a much more mature business. For the buy side, the rise of FXPB created new capital margin efficiencies and opened up new liquidity sources in the market that they could access, namely via the ECNs. For the banks, it offered an opportunity to vastly expand an existing revenue source and a way to become stickier with clients.
Crucially, the development of FXPB also changed the composition of the firms that were active in the foreign exchange market.
“Before FXPB really came into play, you only had bilateral trading and therefore the participants were corporates, financial institutions and governments,” remembers Simon Wilson-Taylor, head of EBS Institutional at CME Group. “As soon as FXPB came in, that all changed, it was the break from only being able to trade with someone that you had credit with, so it opened up the market to a wider range of participants. It also freed people up so that they could deal with their credit separately from their liquidity provision and allowed the growth of electronic trading. As a result, you saw more speculative trading, more retail flows coming in, and as you added those segments it led to a higher volume of tickets at a smaller size.”
Thus, it’s no surprise that the electronic evolution of FXPB coincided with a massive increase in FX trading volumes, but following a period of rapid expansion in the early and mid 2000s, FXPBs were hit with a number of shocks that fundamentally changed these businesses. The first of these being, unsurprisingly, the global financial crisis.
FXPB grew as a segregated business, it was a way of adding to client franchises and the overall FX business, but after 2008 and 2009, the market began to realise that it has capital and risk implications. One market source explains that obviously the FXPBs were aware of this before, but that it just became much more pronounced after the crisis and so costs were focused on more closely, and slowly pricing began to change.
In addition to this, new regulations following the financial crisis, and in particular Basel III, introduced new requirements which forced banks to begin the process of re-examining the profitability of their existing client base and business structure under this new regime.
FXPB, however, was still seen as such an attractive business back then that an increasing number of banks wanted to compete in this space, and these new entrants ensured that fee levels were maintained, and then, as increased competition started to bite, moved ever lower as banks tried to build market share.
While this might have seemed like good news for clients at the time, with some paying less than $1 per million, the pricing battle amongst banks ultimately proved unsustainable. Pure economics alone was forcing some FXPBs to increase their fees and re-assess their client base, the Swiss National Bank (SNB) removed its peg to the euro in January 2015, wrong-footing the market and causing significant losses amongst both FX trading firms and the PB businesses servicing them.
One source, who was working in a PB business at the time, recalls: “The SNB event was huge for the FXPBs – and also for the market in general – because it accelerated FXPBs looking at the costs of the business and the risks that they were bringing on. It also highlighted some of the tail risks involved with this business and was the start of a shake-out where PBs put a minimum balance sheet level requirement in place for clients and I think that it did have a negative impact on the overall credit capacity of the marketplace.”
A combination of the new regulatory requirements, a better understanding of the costs and risks associated with FXPB and the losses from the SNB event led to some banks exiting the business altogether and others significantly scaling back the number of clients that they were willing to service. In a 2017 special report looking at the prime services segment, Profit & Loss documented this all in great detail, but also made the case that the pendulum was swinging back post-SNB as a number of banks were once more looking to grow their FXPB businesses again.
One by-product of the shake-out that happened amongst FXPBs was that the business became very concentrated, with Citi becoming by far the largest player in terms of client base. Which made it all the more significant when Citi – following a reported $180 million loss at the end of last year – announced both a repricing of its FXPB business and culled a number of clients, including some of the largest non-bank market makers in FX.
“Every time there’s a big loss in prime brokerage, it rocks the whole system,” says Jill Sigelbaum, head of FXall at Refinitiv. “Citi changing their business model and pulling out from servicing the non-bank market makers, that’s a huge event for the market. These firms will find homes elsewhere, and I’m sure that they already have, but we’ve just seen this happen over and over again. For example, Rabobank had an issue with a rogue trader and they shut down their entire FXPB business.”
Time for a Change?
Whereas the decade before offered unbridled growth for FXPB businesses, the last 10 years have proven tumultuous to say the least. All of which begs the question: is it still a viable business for the banks to be in?
“You see different views on its viability, but I think it’s a viable business and it always has been,” says Andy Coyne, co-founder and chief product officer at Cobalt. “I think that the recalibration that we’ve seen is really down to certain institutions changing their opinion on risk rather than problems with the mechanics of FXPB.”
The former PB opines: “FXPB works on several levels and a lot of banks are heavily invested in scale, processing and risk management. They are much more aware of the costs and the risks associated with this business and so they’ve adjusted their pricing and put more processes in place to manage those risks. So PB will definitely continue to be an important component of the FX market.”
Similarly, Gio Pillitteri, head of e-FX market making at HC Technologies – which was one of the firms given notice by Citi’s FXPB business earlier this year – says that the demand for FXPB services means that they will very much continue to have a role to play in the market.
“I think that the demand for FXPB services is always going to be there, and if anything it’s going to expand,” he says. “People are working on solutions that will help include the people that are currently disenfranchised and cannot trade, and the market will be better as a result. I actually think that what has happened in the last six months will prove to be beneficial for the market because after the SNB event there was too much concentration risk.”
This is not to say that there aren’t significant changes that FXPBs need to make to their businesses going forward. For starters, Coyne points out that some areas of these businesses that haven’t really changed for 20 years, most notably the models around credit and credit distribution.
“People carve credit limits out, but you can’t react quick enough and carve-outs are static limits at the end point that bear no resemblance to the true credit availability,” he says. “Carve-outs and designation notices are both outdated, and things really haven’t changed to correct this. All we’ve done is made a process that was previously manual electronic, but that doesn’t necessarily make it any better. These are the things that need addressing, but the nature of the market infrastructure will need to change in order for this to happen.”
A New Credit Structure
Sigelbaum sees the potential to improve the existing credit structure “I think there’s still work to be done,” she says. “FXall is more involved in relationship trading and so prime brokers aren’t as essential to the business, but what we’re seeing is that not having a prime broker inhibits what execution venue you can trade on and what liquidity providers you can execute trades with. If we simplified and consolidated our credit allocation and monitoring infrastructure, the buy side could have access to all liquidity sources, while choosing who they face for settlement. So where the market seems to be evolving is around the concept of segregating liquidity provision from credit provision, by utilising existing bilateral credit lines.”
Currently, Sigelbaum explains, banks use different monitoring tools to manage the credit lines they give to their PB clients versus cleints they have a direct relationship with. She says that asset managers don’t want to use PBs because they don’t need the credit, they don’t want to incur PB fees, it complicates how they monitor their own credit appetite with the banks (ie. a PB line and a direct line) and it makes their post-trade workflow more complicated. Buy if this could be simplified for the buy side then Sigelbaum says it would enable certain clients to access liquidity in a more flexible way.
Indeed, Sigelbaum says that for the first time in years there appears to be more creative ideas being proposed around improving the credit structure of the FX market and predicts that five years from now the industry will have a very different credit framework.
Expanding upon this, she comments: “I think that when we get to a place where we’re comfortable with a centralised data model (which due to the blockchain hype, the industry seems more open to than ever) then it will make sense to have a low latency, pretrade credit monitoring in a centralised database. Think about how many problems that would solve.”
Drivers Towards Clearing
Even if FXPBs find new ways to drive efficiencies, reduce risk and expand their businesses, there is increasingly the question of whether more FX flows will move towards a centrally cleared model.
“What happened with Citi with that loss at the end of last year was a game changing event, but what does it lend its hand to?” asks the head of a multi-dealer platform in the US. “It lends its hand to the strategy of the exchanges. In the future, you’re going to see the OTC marketplace operating as it has done historically, with intermediated bilateral credit settlement, but through regulatory changes, capital allocation changes and changes to the cost of the extension of credit for the prime banks, clearing is going to become much more important in the marketplace, not just for listed instruments, but for OTC ones as well.”
The volume of NDF products going to central clearing has risen significantly in recent years; some interbank FX options trading is now just beginning to move that way as well, and it seems that forwards might be the next FX product set targeted by clearing houses, as LCH indicated in regulatory submissions earlier this year that it wants to remove the current restrictions that allow it to only clear spot or forwards products as hedges for options. As more FX products and flows move into central clearing, the case for pushing more towards these venues becomes stronger as the netting and cross collateralisation benefits increase as a result.
Perhaps a more immediate driver towards central clearing for certain products could be the Uncleared Margin Rules (UMR), which will require firms to post initial margin for uncleared swaps, including FX forwards and FX swaps.
The implementation of the UMR was pushed back by one year this summer so that now firms trading over $750 billion of these products will come under the rules in September 2020 and then this threshold will drop down to $8 billion in September 2021, which is when a lot more buy side firms will be impacted by them. As a result, some buy side firms are thought to be looking at using central clearing when trading these products to avoid being caught up in the UMR rules.
“If you’re required to have collateral for certain instruments and the regulations are such that there is a built-in tax for not using a clearing house, then people will gravitate towards using a clearing house and clearing those instruments. Standardised NDFs are perfect for that model,” says Chip Lowry, senior managing director at State Street Global Markets.
The flip side of this though is that some firms are also thought to be evaluating whether they should be using some form of intermediary services, such as a PB. Whereas asset managers traditionally have not needed PBs, with the UMR in place it might make sense for some of them to consolidate their portfolios and centralise any margin payments with a PB. Such a move could also help them from an operational, processing and cost perspective.
“I think we will eventually get to the point where segments of the market that haven’t traditionally used PBs, such as institutional investment managers, will be more interested in doing so. There are actually some firms out there already looking at this, it’s typically not the behemoths doing this but rather the small-to medium sized managers who have sophisticated trading operations and understand that they can get additional efficiencies here,” says Lowry.
Separating Price and Credit
This is partially why some market participants envisage a future scenario where firms are using both services. The range of products impacted by UMR is relatively narrow, so just looking at the PB versus client relationship, then taking out NDFs and options and putting them into a cleared environment may not make sense from a netting point of view because it involves splitting two different pools. But the second order effects of UMR, such as the PB’s cost to the executing bank, mean that if this cost gets passed on then perhaps clearing could become more economical versus FXPB. This suggets that having a multi-structured FXPB, listed and OTC clearing solution in place is might be the way forward.
Part of the reason why PB was such an important innovation for the FX market was that it separated price and credit for the first time, and Lowry argues that this concept still has further to go within the industry.
“I don’t think we have seen the complete story of the desegregation of price and credit, which is a concept that originally applied mostly to hedge funds.” he says. “But we are starting to see the next stage in this story with firms using algos much more than they have in the past, and really FX algos are just PB in disguise for the institutional market.”
While the outlook for FXPB businesses might still be good, however, it’s hard to shake the nagging feeling that, after 20 years, some of them are still struggling with a fundamental question that needs answering.
As Lowry puts it: “The market is going to have to figure out what the right clearing price is for FXPB. It has been historically way underpriced – in the time that I was at Currenex, the pricing battles that we saw amongst the FXPBs just drove it to levels that didn’t seem to make economic sense. So now, especially in the context of what banks have to deal with in terms of stress tests and capital requirements, I think that we’re still wondering: what is the right price of credit? We are also seeing for the first time some PBs beginning to charge executing brokers to recoup costs.”