FX is often thought to be a simple market with a homogenous product where buyers compete on price alone. In fact, winning and retaining business is much more complex, and
starts with an understanding of costs.
The Cost Conundrum
Allocating and splitting costs across different asset classes and parts of the business is not an exact science and remains a challenge for banks with a distribution franchise. The distinction between some capital expenses (Capex) and operating expenses (Opex) may be unclear, and middle and back-office costs can be hard to split out. The FX business may even be a loss leader to win business elsewhere.
A key cost for longer-dated instruments is X-value adjustment (XVA) – account funding, credit risk and capital costs. Along with balance sheet costs, these need to be recorded separately from sales margins, not only to build pricing accurately but also to ensure that margin is assigned post trade to the correct areas.
Staffing costs also need to be considered. To achieve profitability, banks aiming to grow and retain business don’t want staff costs to grow in line with volumes. The clear trend today is to try to automate each part of the front-office process: almost all clients can be automatically priced and are able to self-service their trading and post trade needs. There may be a good reason why some larger swaps or certain pairs can’t be automatically priced but this should be the exception and should be proven to add more value than automation.
A final cost to focus on is that of a bank’s distribution channels. A bank’s larger clients will tend to seek liquidity via a multi-bank platform (MBP), whereas smaller clients will use a single-dealer platform (SDP). Costs for the bank differ greatly between the two. The SDP’s GUI or API channel represents a low fixed cost whereas costs via an MBP can be as high as 10$/M. Under pressure to reduce these costs, banks have in the past been able to negotiate reduced MBP fees. They may also be able to incentivise some clients away from the MBP and back to the SDP/API by proving best execution and improve pricing and by offering features and services not found elsewhere.
Liquidity FX from smartTrade Technologies offers a transparent, fixed-cost service covering the full front office workflow. The end-to-end fully-hosted SaaS model allows the costs of software and hosting to be blended into a simple single annual fee. Aggregation, pricing, risk management, distribution and post-trade are integral functions of the platform.
Some vendors have different charging models such as asking Liquidity Providers (LPs) for rebates in proportion to flows or charging banks $/M fees but this leads to conflicts of interest and hidden and uncertain costs which are difficult to quantify.
Having looked at some of the key costs, what factors should a bank take into account when pricing clients?
First, pricing is relative, so it is important to know where other providers are in the market. Firms fortunate to have flow on both sides might have the option to take on some risk and lay it off later to capture bid/offer spread. Returning to the theme of automation, a question here is whether an algo can and should allow these skews to be automatically driven by positions.
A topic often overlooked in pricing discussions is pricing improvement or slippage and whether the benefit can be passed onto clients when a price and an order cross on the wire in their favour. If client pricing can be improved from time to time, should this happen, or should the price a client gets be fixed, based on the price on the screen? If price improvement is allowed, then so should slippage be – but consumers may find this less palatable. Naturally, whichever route is taken should be clearly disclosed to clients and symmetrical in nature to avoid regulatory risk.
What do clients want?
To say that winning and retaining client business is dependent on price alone would be missing the point. Banks should also consider whether it is it easy for clients to see their pricing – i.e. whether they are pricing to the right channels and venues. Similarly, they should consider whether they are pricing the right bands of liquidity – pricing only 5,10 and 50M is no good if most orders are sub 1M. Format is also important – clients may be comfortable with RFSQ or they may need ESP quotes.
A further consideration is how clients prefer to be charged. In order to sustain business, both sides need to come away happy from each transaction. Clients might be happy to be charged in spread (i.e. as a pips or percentage rate), or they might see the value of having charges split out and clearly labelled as post-trade commission.
Pricing is not set in stone
Finally, pricing should never be a static concept. As the market changes, banks should regularly re-evaluate their choices, considering whether costs or client preferences have changed, and whether they can justify increased margins. If they decide that margins should be tightened, is the lower per transaction profit offset by proportionally higher volumes or not – i.e. are client prices elastic or inelastic?
These are not easy questions, but the answers will be driven by data so it is essential to deploy a trading platform with a solid analytical framework or the ability to extract data to a good third party platform. Data and analytics are becoming ever more important to support Banks in generating reports enabling them to analyse and improve trading performance and pricing while segment their client base.
In conclusion we have seen that pricing and retaining clients relies on banks knowing their own costs, recognising how and where their clients want to consume liquidity and understanding how they like to be charged. Only when they have all this information can banks hope to retain and build their FX businesses.