What is funds transfer pricing? A practical guide for banks

What funds transfer pricing actually is#

Funds transfer pricing is the internal mechanism by which a bank charges its lending businesses for the money they use and pays its deposit-gathering businesses for the money they raise, at rates set centrally rather than negotiated between them. It is an accounting and management construct, not a market transaction: no cash moves outside the institution and no external counterparty is involved.

What moves is responsibility. Every loan is funded, notionally, by the treasury at a stated rate, and every deposit is sold, notionally, to the treasury at a stated rate, so that each business unit is left holding only the margin it can actually influence.

Numbers drawn from noise#

Without such a mechanism, profitability inside a bank is attributed by accident. A branch network gathering low-cost current accounts appears extraordinarily profitable when rates rise and mediocre when they fall, though nothing about its work has changed. A corporate desk writing long fixed-rate loans looks brilliant in a falling market and hopeless in a rising one, for the same reason.

Both results are driven by an interest-rate position neither team chose, cannot measure and has no mandate to hedge. Any incentive, pricing decision or strategic conclusion drawn from those numbers is being drawn from noise.

What the mechanism transfers#

What FTP does, therefore, is transfer risk before it transfers margin. Interest-rate risk, liquidity risk and the funding mismatch between short liabilities and long assets are moved out of the business lines and into the treasury or asset-and-liability management book, where a single team can see the aggregate position, measure it and hedge it.

What each business keeps afterwards is the spread it genuinely earns: for a lender, the credit spread over its transfer charge; for a deposit-taker, the spread between the transfer credit received and the rate paid to the customer. Those two numbers are decision-relevant. The unadjusted ones are not.

What a transfer rate contains#

A transfer rate is rarely a single number. In a developed framework it is assembled from several components, each answering a different question:

• The base or term component comes from the FTP curve and reflects the cost of funds for that maturity.
• A term liquidity premium reflects what it actually costs the bank — not a theoretical risk-free issuer — to raise money for that tenor.
• A contingent liquidity charge covers the buffer the bank must hold against commitments that may be drawn and deposits that may leave.
• A basis adjustment handles currency and index mismatches.
• A capital component charges the business for the regulatory or economic capital its product consumes.
• An option cost prices the customer’s right to prepay a loan or break a deposit, which is a real cost even when nobody exercises it.

Matched maturity and the rate fixed at origination#

The organising principle behind the base component is matched maturity: a transaction is priced against the point on the FTP curve corresponding to its maturity, and that rate is fixed for the life of the transaction rather than refreshed as markets move.

Fixing the rate at origination is what makes the resulting margin meaningful, because it locks in the funding cost that applied when the business made its pricing decision.

A framework that re-rates the back book every time the curve shifts hands business units a profit or loss they did not create and cannot manage — which is precisely the problem FTP was introduced to solve.

A stable core and a volatile portion#

Products without a contractual maturity are where the mechanism meets reality. Current accounts, savings accounts and revolving facilities have no maturity to match, yet they behave as though they do: a large share of a sight-deposit book stays for years, while another share is genuinely volatile.

Behavioural models split those balances into a stable core, priced against a longer point on the curve, and a volatile portion priced short.

The credit a deposit business receives therefore depends almost entirely on a modelling judgement about customer behaviour — which is why those models belong under the same governance, validation and challenge as the ones supporting the liquidity assessment, and why the two should never quietly use different assumptions about the same depositors.

Pricing, value, hedging and the residual#

A mechanism that works produces four visible effects.

• Product pricing starts reflecting real cost, so a long fixed-rate facility priced off a short funding assumption becomes visibly unprofitable rather than invisibly so.
• Deposit-gathering acquires a defensible value, which changes how branch networks and digital channels are assessed.
• The treasury holds a mismatch position it can see and hedge, rather than one distributed across a hundred business decisions.
• The residual left in the ALM book after everything has been transferred becomes small and explicable — a deliberate position taken by the people mandated to take it, not an accumulation of accidents.

Governance#

Governance is what keeps the mechanism honest, because FTP redistributes reported profit between people who are measured on reported profit.

• The curve and the methodology need a named owner, normally treasury, with the asset and liability committee approving the framework and any change to it.
• The rates need to be transparent to the businesses being charged, published on a stated schedule, and accompanied by an explanation of what moved and why.
• Changes need version control and an effective date, so that no transaction is silently re-rated.
• There needs to be a route for a business to dispute a charge and have the dispute resolved on methodology rather than on seniority.

Where those things are missing, the mechanism gets renegotiated informally, and once that starts the numbers stop meaning anything.

Where a mechanism goes wrong#

The failure modes are consistent enough to be worth naming.

• A single pooled rate applied to every product and maturity, which transfers no term risk at all and quietly subsidises long lending at the expense of short.
• A curve that has not been recalibrated since the last time rates were materially different.
• Liquidity priced at zero, so undrawn commitments and volatile deposits appear free.
• Capital left out entirely, so the products consuming the most capital look the most profitable.
• Behavioural assumptions in the FTP framework that contradict the ones in the liquidity assessment for the same balance sheet.
• Most common of all, a technically sound mechanism whose outputs reach no one — no product pricing changed, no incentive adjusted, no strategic decision reconsidered.

An FTP framework that has never made anyone uncomfortable is not measuring anything.