For years, banks operated with interest rates close to zero, and the cost of funding a loan was almost invisible in its margin. When central banks raised rates sharply in 2022 and 2023, net interest income (NII) climbed across most of the industry, and the results looked strong. McKinsey's Global Banking Annual Review 2024 estimates that without that tailwind, industry return on tangible equity in many geographies would have been about 8%, near or below the cost of capital.
That gap left many institutions with a question their reported figures could not answer: how much of the income came from commercial activity, and how much from the level of rates. The accounting margin mixes the two. Separating them is the job of funds transfer pricing (FTP), a mechanism often read as a technical treasury exercise that changes little in how the rest of the bank is run.
This article examines what FTP actually measures, how the margin on a loan or deposit is built from it, and why a margin calculated without it describes a business the bank does not really have.
Every bank charges interest on the loans it grants and pays interest on the deposits it takes. Those two rates are the visible part of the business: the customer negotiates them, the contract records them and the accounting margin is built from them. Behind each of them sits a second price that the customer never sees, the cost the money has for the bank itself.
FTP is the process that sets that second price. The interagency guidance on FTP issued by the Federal Reserve, the Federal Deposit Insurance Corporation (FDIC) and the Office of the Comptroller of the Currency (OCC) defines it in terms of allocation:
A process performed by a firm's central management function that allocates costs and benefits associated with funding and contingent liquidity risks … to a firm's business lines, products, and activities.
That definition is accurate, and it describes where the cost goes once it is known. What FTP produces first, however, is the cost itself. Every loan and every deposit receives an FTP rate, the real cost the money represents for the bank once the different factors that make up that cost are taken into account. The bank can then set that cost against the rate it charges or pays, one contract at a time.
Thirty-year money costs a bank something quite different from what the customer rate on a 30-year mortgage suggests, and FTP exists to say how much. Whatever rate the customer is offered, and whatever the bank paid on average for its funding last quarter, FTP answers a narrower question: what does money committed for the full life of that loan really cost this institution?
The question is narrow on purpose. An average funding rate describes the bank's liabilities as a whole, while the mortgage commits funding for three decades, gives the borrower the right to repay early and uses up room in the bank's regulatory ratios. None of those costs is visible in the rate the customer signs for, and none of them is captured by an average.
With that internal price in hand, the margin on a product becomes a subtraction: the interest rate the bank applies minus the FTP rate. Suppose the 30-year mortgage is priced at 4% against an FTP rate of 3%. It leaves a contribution margin of 1%. These are round numbers chosen for illustration, not market rates.
The two sides of that subtraction behave differently. The bank sets the first one, because pricing is a commercial decision in its own hands. It cannot set the second one. It has to measure it, and the FTP rate is that measurement.
The 3% FTP rate on the mortgage is a base rate plus a set of FTP spreads, each paying for something different that the bank commits when it lends. One illustrative breakdown looks like this:
This breakdown is only an example. Each institution defines its own FTP spreads, and its set may be longer, shorter or cut differently. Whatever the set, each component answers its own question about the same contract, and none of them can be read off the customer rate.
Together, the base rate and the FTP spreads form the FTP rate, the full cost of the money behind the contract. The rate the bank earns minus the FTP rate is the contribution margin, a figure some institutions call the FTP margin. Read this way, the 1% at the end of the calculation is a residual rather than a starting point: if the base rate or any FTP spread rises while the customer rate stays where it was, the margin falls by exactly as much.
The separation between the base rate and the liquidity spread is the least intuitive part of the calculation, because the two can follow different horizons on the same loan. The interagency guidance makes the point with a five-year commercial loan whose interest rate component is set against a three-month repricing horizon while its liquidity component is set against the full five years. One asset carries two separate transfer price components.
On a fixed-rate 30-year mortgage the two horizons coincide, but the components stay separate. The base rate pays for fixing the rate over the life of the loan. The liquidity spread pays for keeping the funding in place for just as long.
The contribution margin is only as complete as the list of costs subtracted to reach it. Take the mortgage again, priced at 4% against an FTP rate of 3%, and leave the optionality spread out of that rate. The FTP rate falls below 3%, and the reported margin rises above 1% by exactly the amount left out.
The mortgage has not changed, and neither has the customer or the funding behind it. Only the measurement has. The bank now reports a margin it does not earn.
The same gap opens when a bank sets its accounting margin against one average funding rate. The average charges the 30-year mortgage what it charges a one-year loan, so the liquidity spread for money tied up for three decades drops out of the calculation without anyone deciding to drop it. Álvaro Nieto, Head of Product at Mirai, puts the consequence in plain terms:
In recent years FTP has become a key factor in margin management. If we don't take it into account, the margin we are calculating isn't real, because it ignores the cost of funding.
A wrong margin on a single product rarely stays on a single product. The first consequence is the most direct: a product can look profitable on the margin the bank measures and still lose money on the margin it earns. Nieto describes the risk through a client case: a bank that does not charge its mortgages the regulatory cost of granting them, nor account for how each new mortgage affects its ratios, can end up granting a very high volume of them that turns out to be negative for it. The larger the commercial success, the larger the loss, and nothing prompts anyone to look, since every report shows the product in positive territory.
The second consequence is harder to see. An incomplete or averaged FTP overcharges some contracts and undercharges others, and the ranking of products, business areas and branches shifts with it. When the yield curve slopes upward, one average rate charges short-term lending more than its money costs and long-term lending less, and credits stable deposits with less than the funding they provide is worth. A branch whose customers hold large, stable balances can rank near the bottom of the network, while a branch lending heavily on underpriced mortgages ranks near the top. Nieto sees both failures in the same measurement gap:
Getting FTP wrong can lead us to sell products at a loss, or to lose focus on the products, areas and locations where we are most profitable.
The third consequence turns measurement into behavior. Budgets, targets, branch rankings and bonus pools are all built on the margin the bank measures, so FTP decides what each unit is seen to earn and, sooner or later, what it is asked to deliver. Targets set on volume push growth. Targets set on contribution margin push profitable growth.
While rates were rising, the increase in NII lifted reported income across most of the business at once. A product priced too close to its FTP rate still appeared to contribute, and a branch whose lending barely covered its base rate and liquidity spread still showed growth in income. That lift covers weak lines as readily as strong ones.
In Nieto's reading, the environment now points to a possible new rise in rates. The direction matters less than it seems. Whichever way rates move next, the question for management is which products, business areas and branches remain profitable once the rate effect is taken out, and which were profitable only because rates were high.
Only the real margin, with the base rate and every FTP spread subtracted, can answer it.
Knowing what FTP measures is the first step, although a definition on its own does not change a pricing decision. The margin becomes usable only when the base rate and every FTP spread are calculated for every contract, on the bank's real curves and assumptions, and when the result is ready in time to act on. The whitepaper Funds Transfer Pricing as a Steering Mechanism builds from this definition to the automated, contract-level process that makes the real margin usable, and then to projecting that margin forward and testing decisions against it before they are taken.
Institutions that measure the real margin contract by contract will know which products, areas and branches deserve their capital and effort. Those working from the reported margin will learn the same thing from their annual results.
The Mirai Platform is one example of how this works in practice, bringing FTP, balance sheet management and regulatory compliance together on one platform, fed by one data load and run on one set of assumptions. At one global systemically important bank client, the calculation processes millions of contracts in about 30 minutes.
For the full framework, from the base rate and FTP spreads to contribution margin by product, area and branch, calculated overnight and tested with what-if strategies, see Mirai RiskTech's whitepaper Funds Transfer Pricing as a Steering Mechanism: how banks measure and steer the real margin.