Why this approach
How this actually makes money
The job description leads with commercialisation and go-to-market strategy. This is the business case: where the revenue comes from, what it costs to run, what breaks it, and what would make me stop.
New to the wording on this page? 6 terms explainedShow
TreasurerCorporate treasurer
The person inside a large company responsible for its cash: making sure each part of the business has money when it needs it, spare cash earns something, and currency risk is managed.
For example: A group with offices in six countries has one treasurer deciding each morning which subsidiary needs funding, which has surplus cash to sweep back, and what to do about a weakening currency.
Nostro account
An account one bank holds at another bank, usually abroad and in that country's currency, so it can make payments there. Latin for 'ours' — our money, held at your bank.
For example: For HSBC to pay someone in Brazilian reais, it keeps a pot of reais in an account at a Brazilian bank. That cash sits there earning little, purely so payments can be made — money that is parked rather than working.
Cut-off time
The daily deadline after which a payment will not be processed until the next business day.
For example: A treasurer realises at 6pm Hong Kong time that the Singapore office needs cash. The payment window closed at 5pm, so the money cannot arrive until Monday — leaving Singapore short all weekend while the cash sits unused in Hong Kong.
RMRelationship Manager
The banker who owns the commercial relationship with a corporate client — the person who actually sells the product and fields the complaints.
For example: A Hong Kong manufacturer has one main HSBC contact who knows their business, brings them new products, and gets the call when a payment goes wrong.
RFPRequest for Proposal
A formal document a large client sends to several banks, asking each to propose a solution and compete for the business.
For example: A multinational sends the same 200 questions to HSBC, Citi and Standard Chartered about handling its Asian cash management, then scores the answers side by side and picks one.
Go/no-go gate
A checkpoint between phases where named people decide whether the product may continue. Not a status update — a decision with genuine power to stop the work.
For example: Before phase two, legal must confirm settlement finality in that specific country. If they cannot, the phase does not start — regardless of how much has already been spent.
All figures are illustrative and internally consistent for this case study. They are not HSBC numbers and are not a forecast. What matters is the shape of the reasoning — say this out loud before walking anyone through the numbers.
The one insight to lead with
A tokenised treasury product does not make its money on transaction fees. In the worked example below, fees contribute about US$86,000 per client a year while retained deposit balances contribute about US$450,000 — roughly five times as much. The commercial purpose of the product is to attract and defend operating balances. Everything about how you price it follows from that, and most candidates miss it entirely.
Where the revenue comes from
Five lines, and they behave very differently. Knowing which are predictable, which follow client volume and which follow interest rates is the difference between a business case and a wish.
Platform subscription
PredictableUS$4,000 per month · US$48,000 a year
A fixed monthly fee per client for access, entitlements and the support model, independent of how much they transact.
The only line that does not move with client behaviour. Useful for covering the fixed cost to serve, and the easiest line to defend in a downturn.
Per-transfer fee
Volume-linkedUS$8 per transfer × 400 a month · US$38,400 a year
A charge on each tokenised movement, tiered so heavy users pay less per transaction.
Deliberately priced below the wire it replaces. If you try to hold legacy pricing on a cheaper rail, the client simply does not migrate.
Foreign-exchange spread
Volume-linkedVaries by corridor and volume
Where a transfer crosses currencies, the bank earns the margin on the conversion — the same way it does today.
Often the largest fee line in cross-border, but it is not new revenue: it already exists on the legacy rail. Counting it as incremental overstates the case.
Retained deposit balances
Rate-sensitiveUS$30m retained × 1.5% net interest margin · US$450,000 a year
Operating cash that stays with the bank because the service is genuinely useful. The bank earns a net interest margin on those balances.
This is the real prize, and it is roughly five times the fee revenue. A treasury product's commercial purpose is to attract and hold operating balances — the fees are almost a rounding error beside it.
Implementation fee
PredictableUS$40,000 one-off
A one-off charge covering integration, testing and onboarding effort.
Often waived for anchor clients to win the reference. Worth being explicit that this is a customer-acquisition cost, not lost revenue.
Unit economics per anchor client
One mid-size multinational, eight entities, four markets
| Platform subscription | + US$48,000 |
| Per-transfer fees (400/month) | + US$38,400 |
| Net interest margin on US$30m retained balances | + US$450,000 |
| Cost to serve (support, compliance, operations share) | − US$60,000 |
| Annual contribution per anchor client | US$476,400 |
The investment ask and payback
- Build: US$6.5m over 18 months
- Platform engineering, core-banking integration, financial-crime controls, legal opinions per corridor and the operating model to run it around the clock.
- Run: US$2.8m a year
- Product and engineering team, infrastructure, 24/7 operational cover and ongoing compliance monitoring. The always-on promise is a real, recurring cost, not a one-off build.
Covers run cost
About 6 anchor clients
Full payback over 3 years
About 11 anchor clients
Six clients cover the annual run cost. Eleven cover run cost plus repayment of the build investment across three years. That is a realistic ask for a bank with HSBC's corporate treasury footprint, which is what makes the case credible rather than aspirational.
What breaks this case
Volunteering these before you are asked is what makes the rest of the numbers believable.
Interest rates fall
CriticalNet interest margin compresses from 1.5% to 0.75%
Deposit contribution halves from US$450,000 to US$225,000 per client. Annual contribution falls to about US$251,000 and break-even roughly doubles, from 11 clients to about 20.
So what: The single most important thing to understand about this business case: it is driven by interest rates, not by the product. If you present it as a technology investment you will be caught out by the first person who has run a transaction-banking profit and loss account.
Clients migrate but do not deposit more
CriticalVolume moves from wires to tokens with no change in balances held
Fee revenue per client falls, because tokenised transfers are priced below the wires they replace. Net revenue goes backwards.
So what: This is the cannibalisation trap. The product only pays for itself if it wins or defends balances. Migration alone is a revenue loss dressed up as innovation.
Adoption is slower than planned
MaterialClients take 9 months to go live instead of 4
Run cost continues while contribution is delayed, pushing payback out by roughly a year.
So what: Argues for a small number of deeply committed anchor clients over a long list of interested ones. Depth of commitment beats breadth of pipeline here.
A corridor's legal opinion does not clear
MaterialOne planned market cannot go live
Addressable volume shrinks and some already-spent integration cost is stranded.
So what: Why legal readiness is a go/no-go gate before build spend in that corridor, rather than a parallel workstream you hope catches up.
How you would price it
Three approaches, with a clear recommendation. Being able to say why you rejected the other two is worth more than the recommendation itself.
Cost-plus
Work out what it costs to run, add a margin, charge that.
For
- + Simple to defend internally
- + Protects against loss-making clients
Against
- − Ignores what the capability is actually worth to the client
- − Early on, cost per transaction is high because volume is low — so this prices the product out of the market exactly when it needs adoption
Wrong for a launch product. It prices against your own inefficiency rather than the client's benefit.
Value-based
Price against what the client saves — idle cash released, cut-off costs avoided, reconciliation hours removed.
For
- + Aligns price with the benefit the client actually receives
- + Supports a genuine commercial conversation rather than a line-item comparison
Against
- − Requires a credible baseline of the client's current costs, which many treasurers have never measured
- − Harder to standardise across a sales force
Right in principle, but only workable if discovery has quantified the client's current pain. That measurement work is part of the product job.
Penetration pricing into balance capture
RecommendedPrice transactions close to cost to remove every barrier to migration, and take the return on retained balances instead.
For
- + Removes the client's reason to hesitate
- + Aligns directly with where the money actually is — the deposits
- + Makes the product a defensive moat around existing balances, not just a new fee line
Against
- − Requires the bank to accept thin fee margins early and hold its nerve
- − Only works if balance retention is genuinely measured and attributed to the product
The recommended approach. The economics above show deposits are roughly five times the fee revenue, so competing on transaction price to win balances is the rational trade — provided you can prove the attribution.
The four questions a finance director will ask
Is this new revenue, or are we just moving existing revenue onto a cheaper rail?
Honestly, at first it is mostly migration, and per-transaction revenue goes down. The case rests on defending and attracting operating balances. If we do not build it, the balances still leave — to a bank that did. I would rather cannibalise our own fees than lose the deposits.
What happens to this business case if rates fall?
It weakens materially — halving the net interest margin roughly doubles the number of clients needed to break even. That is why I would hold the fee lines as a floor rather than discount them to zero, and why I would not approve the full multi-corridor build on a rates assumption alone.
Why now, rather than in two years?
Because balance relationships are sticky once a client integrates their systems into a bank's rails. The cost of being second is not a delayed launch — it is a client whose treasury systems are already wired into a competitor.
What would make you stop?
If anchor clients migrate volume but we cannot demonstrate balance retention attributable to the product, the core premise is wrong. I would stop at the end of the pilot rather than fund a corridor expansion on hope.
Which clients to target
Being able to name who you would decline, and why, is worth more than a long list of who you would sell to.
Multi-entity regional groups with an Asian treasury centre
Target firstSix to fifteen legal entities across three or more time zones, a central treasury in Hong Kong or Singapore, and frequent intra-group funding. Typically manufacturing, shipping, commodities or regional conglomerates.
Why them
The cut-off problem is a daily operational reality for them, not a hypothetical. They already move money between their own entities constantly, so migration needs no new client behaviour — only a better rail for behaviour they already have.
Watch out
Confirm the treasury centre has authority over subsidiary cash. If each subsidiary controls its own balances, the central buying decision does not exist and the deal stalls in a committee.
Asset managers and funds needing subscription and redemption settlement
Second waveFund operators moving cash against unit creation and redemption, increasingly against tokenised fund units.
Why them
Directly adjacent to the Hong Kong Monetary Authority's own pilot work on tokenised funds settling against tokenised deposits, so the regulatory path is already being trodden.
Watch out
Settlement timing is driven by fund rules and cut-offs you do not control. The benefit is real but the sales cycle runs through the fund administrator too, not just the client.
Digital-native platforms and marketplaces with 24/7 payout cycles
Second waveBusinesses whose own customers transact at all hours, so treasury needs to fund positions outside banking windows.
Why them
The always-on proposition is worth most to a business that is itself always on, and they typically have the engineering capability to integrate quickly.
Watch out
Balances tend to be operational and thin rather than parked. Strong volume, weaker deposit retention — which undercuts the part of the business case that actually pays for the build.
Single-entity domestic corporates
Decline for nowOne legal entity, one market, straightforward payables and receivables.
Why not
No genuine use case. Without multiple entities there is no intra-group transfer to improve, and existing domestic rails already settle quickly.
Watch out
Easy to onboard and therefore tempting for a pipeline number. Resist it — cost to serve exceeds any plausible value and it distorts the pilot evidence.
Clients seeking crypto exposure rather than treasury efficiency
Decline for nowInterested because it is blockchain, not because they have a cash-mobility problem.
Why not
Wrong motivation produces wrong requirements. They will ask for public-chain interoperability and asset trading the product is not designed to provide.
Watch out
These conversations are enthusiastic and can look like demand. They are the fastest way to pull a roadmap off course.
The qualifying checklist for a first client
If a prospect fails more than two of these, they are a second-wave client however enthusiastic they are.
- Multiple legal entities in corridors where a legal-finality opinion already exists
- At least 200 intra-group transfers a month, so migration produces a measurable signal
- Meaningful operating balances already held with the bank, so retention can be attributed
- In-house capability to integrate by interface rather than manual portal use
- A treasurer who has already measured their own cut-off and reconciliation cost - without a baseline there is no way to prove value
- Willing to be a named reference if the pilot succeeds
What would prove this should scale
Leading metrics tell you early, outcome metrics tell you whether it worked, and guardrails tell you when to stop even if the first two look good.
The scale decision itself
I would take the scale decision only when three things hold together: above 60% of eligible flow migrated, attributable balances tracking towards target, and guardrails stable or improving as volume grows. Any one alone is misleading. High migration with flat balances means we have given clients a cheaper rail and gained nothing. Good balances with degrading exception times means we are about to damage the relationships we just won.
Metrics I would refuse to be judged on
- Total value settled - impressive to quote and almost meaningless; a single large transfer can dwarf a month of genuine adoption.
- Number of pilots or memoranda signed - measures sales activity, not client value received.
- Cumulative transaction count - only ever rises, so it can never tell you to stop.
- Press coverage and awards - relevant to the thought-leadership objective, but never evidence that the product should scale.
Who inside the bank decides whether this succeeds
This role owns the outcome without owning the teams. Four of these can stop the product outright, so knowing what each one cares about is the actual job.
Treasury / Asset and Liability Management
Can stop the productOwns: The balance sheet, and the internal funds transfer pricing that decides how much value a product is credited for the deposits it raises.
What they care about
Deposit quality, stability and duration - not transaction volume.
What you need from them
An agreed attribution method, so balances held by entities active on the platform are credited to this product rather than disappearing into a general pool.
How they block you
Without an attribution agreement the entire business case is unprovable. You will have raised the deposits and be unable to demonstrate you did.
How to bring them with you
Engage them before build, not at the first review. Frame the product as a deposit-stability instrument in their language, and agree the measurement method while it is still a hypothesis rather than a claim.
Global Payments Solutions sales and relationship managers
Can stop the productOwns: The client relationships and the revenue conversation.
What they care about
Whether this helps them retain and grow their accounts, and whether it is explainable without technical training.
What you need from them
Qualified introductions to the right clients and honest feedback on objections.
How they block you
Quietly. They simply do not raise it in client meetings, and the product dies from absence rather than rejection.
How to bring them with you
Give them a one-page value narrative in client language, a qualifying checklist, and early wins they can point to. Never send them into a meeting needing to explain a ledger.
Financial Crime Compliance
Can stop the productOwns: Screening standards, sanctions policy and the threshold for holding a payment.
What they care about
That faster settlement never means weaker control, and that every hold is defensible to a regulator.
What you need from them
Agreement on pre-settlement screening design and the exception path.
How they block you
A late objection here can invalidate the core atomic-settlement design, because controls must complete before value moves.
How to bring them with you
Involve them in designing the unhappy path first. Arriving with the exception flow already drafted signals you understand their constraint is the product, not an obstacle to it.
Legal
Can stop the productOwns: Settlement finality opinions per corridor, client terms and the enforceability of the arrangement.
What they care about
Whether the bank's position holds if a counterparty fails, and whether obligations are clear.
What you need from them
A finality opinion per corridor before that corridor is built, not after.
How they block you
A corridor without an opinion cannot launch, stranding whatever integration spend already went into it.
How to bring them with you
Sequence the roadmap so legal readiness is a gate ahead of build spend, and give them long lead times - opinions are slow by nature.
Operations
Owns: Exception handling, reconciliation and the round-the-clock service model.
What they care about
Whether they can actually staff and run the promise the product makes.
What you need from them
A costed, staffable operating model for 24/7 cover.
How they block you
An unrunnable promise gets diluted after launch, and the client experience quietly degrades to match.
How to bring them with you
Cost the operating model honestly in the business case rather than treating always-on as free, and design exceptions with them rather than handing them over.
Technology and Engineering
Owns: Build, integration to core systems, resilience and recovery targets.
What they care about
Clear requirements, realistic sequencing and non-functional targets set early rather than retrofitted.
What you need from them
Delivery capacity and honest estimates.
How they block you
Integration to core banking is usually the longest pole; underestimating it moves every date.
How to bring them with you
Prioritise a thin end-to-end path over breadth of features, so integration risk surfaces in month two rather than month ten.
Finance and Product Control
Owns: The profit and loss account, pricing approval and cost allocation.
What they care about
Whether revenue is incremental or cannibalised, and whether cost to serve falls with scale.
What you need from them
Pricing sign-off and agreement on how deposit contribution is recognised.
How they block you
They can refuse a price that undercuts existing payment revenue unless the balance argument is made explicitly.
How to bring them with you
Bring the cannibalisation point yourself, with the balance retention case attached. Being the one who raises it earns far more credibility than being the one who omitted it.
Risk (Non-Financial Risk) and the New Product Approval forum
Owns: Formal risk acceptance and the authority to permit launch.
What they care about
That risks are identified, owned, controlled and evidenced - not that they are zero.
What you need from them
Approval to launch, and to each subsequent phase.
How they block you
No approval, no launch. This is an absolute gate rather than a negotiation.
How to bring them with you
Treat the control framework as part of the product from the start. A risk register with named owners and tested controls moves far faster through this forum than a strong commercial case.
Regulatory Affairs and the regulator relationship
Owns: The bank's supervisory dialogue, including with the Hong Kong Monetary Authority.
What they care about
No surprises, and consistency with what the bank has already told its supervisor.
What you need from them
A path into pilot programmes and early read on supervisory expectations.
How they block you
Going around them damages a relationship worth far more than any single product.
How to bring them with you
Give them early visibility and let them lead the conversation. The job description names regulator engagement directly, so demonstrating this instinct matters in the interview itself.
Marketing and Communications
Owns: External positioning, media and thought-leadership opportunities.
What they care about
Credible, approved messages that strengthen the bank's position.
What you need from them
Support for the thought-leadership objective the job description sets out.
How they block you
Rarely block, but can overstate a pilot as a launch - which creates a client and regulatory problem you then own.
How to bring them with you
Give them precise language about what is live versus piloted, and review claims before they go out.
Four judgement calls I would make, and defend
Few deep pilots, not many shallow ones
Three genuinely committed clients produce better evidence than twelve curious ones. Shallow pilots generate activity metrics and no decision-grade signal, and they consume the same onboarding effort.
Refuse client-specific customisation in phase one
The first bespoke build feels like commercial flexibility and becomes a permanent maintenance cost that slows every later client. Say no early, and explain that a standard product is what makes it cheap for them.
Start where legal finality already exists
Corridor choice should follow legal readiness, not client enthusiasm. The most excited client in an unopined corridor is a slower path to revenue than a lukewarm one in a clear corridor.
Agree deposit attribution before building anything
If the internal funds transfer pricing model will not credit the product for balances it raises, the business case cannot be proven regardless of how well the product performs. This is an internal negotiation, and it decides the outcome more than any feature does.
What would make me stop
Naming your own kill criteria unprompted is one of the strongest senior signals available in an interview. It says you own an outcome, not a project.
- Pilot clients migrate transaction volume but retained balances do not move — the central premise fails and no amount of further build fixes it.
- Cost to serve per client stays above US$150,000 after the first year, which would mean the operating model does not scale.
- Legal finality cannot be established in at least two priority corridors, capping the addressable market below break-even.
- A shared industry utility emerges that offers the same client outcome at lower cost, making a single-bank build the wrong structure.