Funds Transfer Pricing: How Banks Put a Price on Internal Liquidity

What Funds Transfer Pricing Is and Why Banks Use It

Funds Transfer Pricing (FTP) is an internal accounting mechanism that allocates liquidity costs and funding credits across a bank's business lines. Some firms call it liquidity transfer pricing. Whatever the label, the purpose is the same: make the cost of liquidity visible at the point where it is created, not buried in a Treasury residual.

Without FTP, a lending desk that books a five year fixed rate mortgage consumes five years of funding at no visible internal cost. The margin looks attractive. Treasury absorbs the mismatch invisibly. FTP changes that by charging the lending desk for the funding it uses and crediting deposit gathering desks for the stable funding they provide.

The PRA, through its supervisory expectations on liquidity governance, looks for evidence that liquidity costs are embedded in pricing and performance measurement across the organisation. A robust FTP framework is how most banks demonstrate that.

Treasury as the Internal Liquidity Market

The easiest way to understand FTP is to think of Treasury as running an internal money market. Every other business line is either a borrower from or a depositor into that market.

The two sides work like this:

• Lending desks (mortgages, corporate loans, trade finance) borrow funding from Treasury to support the assets they originate. Treasury charges them an FTP rate for that funding.
• Deposit gathering desks (retail current accounts, savings, corporate deposits) provide funding to Treasury by attracting customer liabilities. Treasury pays them an FTP credit for the stable funding they generate.

Treasury sits in the middle. Its role is to manage the aggregate mismatch between the assets funded and the liabilities gathered, using wholesale markets to fill gaps and hedge the interest rate and liquidity risk that arises at the centre.

This structure has an important consequence. The net interest income that Treasury reports is not product profit. It is the residual from managing the overall balance sheet. The product profit sits in the business lines, net of the FTP charge or credit they have received. That separation is the point.

How Charges and Credits Work in Practice

Take a simplified example to make this concrete.

A retail lending desk books a two year personal loan at 7.5%. Treasury applies an FTP charge of 4.5%, reflecting the two year funding cost including a term liquidity premium. The lending desk retains a 3.0% net interest spread, or product NIM, which is the margin before credit risk costs and operating costs. That 3.0% is the number that tells you whether the product is actually profitable.

On the other side, a retail savings desk gathers a one year fixed term deposit at a customer rate of 3.0%. Treasury pays that desk an FTP credit of 3.8%, reflecting the value of the one year stable funding provided. The savings desk earns a 0.8% positive margin on every pound of deposit it gathers, which incentivises it to grow that book.

Without FTP, both figures are invisible. The bank's net interest margin is an aggregate that tells you nothing about which products or businesses are driving value and which are destroying it.

Building the FTP Curve

The FTP rate for any transaction is built from components stacked on top of a base curve. The typical build looks like this:

1. Reference rate curve: a risk free rate curve built from overnight index swap rates, such as those derived from SONIA, reflecting the time value of money at each tenor.
2. Term liquidity premium: the additional cost of locking in funding for a longer period. A one year funding cost is higher than overnight because the bank gives up flexibility. This premium widens in stress and is observable in bank wholesale funding spreads.
3. Additional regulatory costs: some banks add an explicit charge for the cost of holding HQLA buffers under LCR requirements, or for maintaining the stable funding ratios required by NSFR. The cost here is the difference between the yield on HQLA eligible assets such as gilts and the marginal cost of the funding used to hold them. If that carry is negative, it is a real cost and FTP should allocate it somewhere.
4. Contingency funding cost: a smaller component reflecting the cost of maintaining committed facilities and standby liquidity arrangements.

The resulting FTP curve gives a rate for each maturity from overnight out to whatever the longest tenor the bank funds. In normal curve conditions, a five year mortgage will attract a higher FTP charge than a one year overdraft, because five year funding costs more and carries more liquidity risk. In an inverted yield curve environment the rate relationship can reverse, though the liquidity risk point holds regardless.

Treasury teams typically review and recalibrate this curve at regular intervals, though the frequency is a design choice. Lagging market conditions is one of the most common failures in practice, and it matters because an FTP curve that does not reflect current wholesale funding spreads will misprice product profitability.

Matched Maturity vs Pooled Approaches

There are two broad approaches to how the FTP rate is applied to individual transactions, and the choice has significant consequences.

Matched Maturity FTP

Under matched maturity FTP, each transaction is assigned a rate based on its own tenor and cash flow profile. A five year loan gets a five year FTP charge. A 95 day fixed deposit gets a 95 day FTP credit. The rate is locked at origination and does not change over the life of the transaction.

This is more complex to operate. You need to track each transaction's FTP rate, maturity, and optionality (prepayment, extension, early withdrawal). Behavioural assumptions are needed for products without contractual maturity, such as current accounts or revolving facilities. But the result is accurate. Each business line's margin reflects the actual liquidity cost of what it has booked.

Pooled or Blended FTP

Pooled approaches apply a single blended rate across a book of assets or liabilities, regardless of individual tenor. Simpler to operate, but the blended rate is a weighted average that will overcharge short tenor assets and undercharge long tenor ones.

In practice this creates a subsidy. Long tenor lending looks cheaper than it is, which means the bank may originate more of it than is economically rational. Short tenor lending looks more expensive, which may push the business away from what is actually attractive on a risk adjusted basis.

Most banks with mature FTP frameworks have moved toward matched maturity approaches for significant asset and liability categories, even if they retain pooled rates for smaller or more complex portfolios.

FTP and Product Profitability

FTP is the tool that makes net interest margin attribution honest.

Without FTP, Treasury appears to generate the bulk of the bank's NIM because it sits in the middle of the balance sheet. Business lines look less profitable because their gross margin is visible but the cost of the funding they consume is not charged to them.

With FTP, the NIM is decomposed properly. Treasury's contribution is the residual from balance sheet management. Each business line's contribution is its product spread after the internal funding charge. That decomposition enables real comparisons: between products, between portfolios, between vintages, and between origination channels.

This matters for capital allocation decisions, product pricing reviews, and performance management. If a mortgage book looks profitable before FTP and marginal after it, that is information the business and the board need to act on.

What Good FTP Design Actually Achieves

A well designed FTP framework does more than allocate costs. It shapes behaviour across the organisation.

When deposit gathering desks are paid a genuine credit for stable retail deposits, they have a financial incentive to grow those books. When lending desks are charged a real five year rate for five year assets, they price products to recover that cost or do not originate them. The FTP framework becomes a bank wide signal about what the bank values.

This alignment is particularly important for LCR and NSFR. If the FTP framework uses different run off assumptions or maturity assumptions than the regulatory liquidity models, there is a disconnect. A product that helps the LCR should generate a better FTP credit than one that does not. A product that stretches the NSFR by consuming required stable funding should face a higher FTP charge. When FTP and regulatory reporting are calibrated consistently, the business lines are making decisions that improve the regulatory position as a natural consequence of pursuing their own P&L.

Common failure modes worth looking out for:

• Blended rates that undercharge long tenor assets and subsidise duration risk at the business line level
• Stale curves that lag market conditions by weeks or months, producing FTP rates that no longer reflect the real cost of wholesale funding
• Absent regulatory cost component, meaning the cost of the LCR buffer is absorbed by Treasury and invisible to the businesses that drive the buffer requirement
• Optimistic behavioural assumptions for non maturity deposits, resulting in FTP credits that overvalue what is actually flighty funding

FTP design is ultimately a judgement call on several dimensions. The right level of complexity, the calibration of behavioural assumptions, the frequency of curve updates, and the treatment of optionality are all choices a bank's ALCO and Treasury function need to make deliberately rather than by default.

Related Reading

If you want to go deeper on the liquidity risk context that FTP sits inside, these posts from the series are worth reading alongside this one:

• The five types of liquidity risk a bank actually has to manage: the risk taxonomy that FTP is designed to price and allocate
• Why liquidity fails fast: why pricing internal liquidity correctly matters so much under stress
• Liquidity risk through history: the historical context for why regulators care about this at all

Practical Takeaway

If you work in Treasury, finance, or a business line role at a bank, look at how your FTP framework is calibrated and whether the curve is current. An FTP framework is only as useful as its inputs. A stale or blended rate is not a neutral choice: it actively misprices risk and distorts the decisions that flow from it.