Content uploaded by Hovik Tumasyan
Author content
All content in this area was uploaded by Hovik Tumasyan on Mar 17, 2018
Content may be subject to copyright.
Revisiting Funds Transfer Pricing
By Hovik Tumasyan
PricewaterhouseCoopers, LLP∗ †
PwC Tower, 18 York Street, Suite 2600,
Toronto,ON M5J 082, Canada
hovik.tumasyan@ca.pwc.com
February, 2012
Abstract
Fundamentally, the objectives of funds transfer pricing (FTP) has remained the same
as it was when it was first developed. However, in the environment of very low interest
rates that preceded the recent financial crisis FTP frameworks in many banks have
been left underfunded and underdeveloped.
The financial crisis brought FTP into the regulatory spotlight. Regulators expect banks
to be able to demonstrate how their FTP frameworks are aligned to the best practice
principles for liquidity management.
The paper reflects on the fundamental aspects of a FTP framework and its role in a
bank.
∗Views expressed are those of the author and do not reflect those of PwC or its affiliates.
†Parts of this article have appeared as a chapter in Asset-Liability Management for Financial Institutions,
Bloomsbury Information Ltd., London 2012.
1
Revisiting Funds Transfer Pricing 2
Contents
1 Background - The Changing Landscape 2
2 How Does FTP Work - The Mechanics 3
2.1 PoolApproaches .................................. 6
2.2 Matched Maturity Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3 Matched Maturity Funds Transfer Pricing 9
3.1 Transfer Pricing of Assets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2 Transfer Pricing of Liabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.3 Methodological Aspects of the Matched Maturity Approach . . . . . . . . . . 17
3.4 Transfer Pricing Equity Capital . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4 Closing Remarks 21
1 Background - The Changing Landscape
In an environment of very low interest rates that preceded the recent financial crisis many
of the funds transfer pricing (FTP) frameworks in banks have been left underfunded and
underdeveloped. Because in such environments cost of liquidity is not a constraint enough
to feature in the growth strategies at banks, infrastructures for charging for liquidity con-
sumption have remained overly simplistic, lacking scalability and responsiveness.
Not surprisingly, in the aftermath of the financial crisis FTP, as one of the fundamental build-
ing blocks of bank liquidity measurement and management, was brought into the spotlight
of regulatory scrutiny. Recent reviews of the FTP frameworks in banks by the regulators
have revealed major functional inefficiencies in FTP systems that run through practically
all the dimensions of a FTP framework (see for example, [FSA 2010], [EBA 2010] and also
[Grant 2011]).
The UK regulators, for example, articulated their requirements for FTP frameworks in a
letter to the Treasurers of the banks under their oversight ([FSA 2010]). In particular:
Pricing and P&L Attribution
•Price all the liquidity costs - funding costs, liquidity premiums, indirect/hedging costs,etc.,
•Minimize allocation of liquidity costs to the center,
•Price and allocate costs of contingency liquidity and stress test buffers.
FTP Granularity
•Apply FTP to a sufficiently granular level to affect business transaction decision makers,
•Aggregate bottom-up the cost of liquidity to performance measurement points.
FTP Consistency
•Establish an oversight and governance framework around FTP to monitor consistency of
its application,
Revisiting Funds Transfer Pricing 3
•Synchronize the funding costs calculations and methodologies across the organization.
Responsive FTP Framework
•Eliminate over-reliance on offline systems requiring manual intervention,
•Revise the simplistic assumptions in the implementation of the FTP framework.
Strategic role for FTP
•Implement forward looking liquidity cost measures,
•Embed FTP revaluation capabilities in stress test scenarios,
•Enforce the strategic roles for FTP - a signaling tool to business units and balance sheet
steering tool to senior management.
FTP has become a regulatory requirement. Regulators expect banks to be able to demon-
strate how their FTP frameworks are aligned to the best practice principles for liquidity
management ([Basel 2008], [Basel 2010], [IIF 2007]).
Given the deliberate efforts by regulators globally to achieve harmonisation in both quali-
tative requirements and quantitative calibration of liquidity risk management, it would be
na¨ıve not to expect that these requirements will soon find their ways into other jurisdictions.
2 How Does FTP Work - The Mechanics
In its simplest form FTP is the process wherein the Treasury of a bank (its funding cen-
ter) aggregates funds centrally and then redistributes them throughout the business units,
balancing funding resource excesses and shortages, thus creating an internal market for
liquidity. If there is still a deficit for funds, Treasury raises more funds from the capital
markets, and if there is an excess of funds Treasury invests them or lends in the wholesale
markets.
FTP has been an integral part of bank management for over three decades. It traces its
origins to the 1970s and the deregulation of interest rates in the U.S., when it was developed
as a tool for managing the interest rate risk in banks1.
Fundamentally, the purpose of an FTP still remains the same as it was when it was first
developed - aggregate the interest rate exposure of the whole bank into a central location
for its effective management. However, in doing so, FTP generates a few other results that
are sometimes quoted as main purposes for FTP:
•By transferring the interest rate risk into a central location, it makes the balance sheets
of business units immune to interest rate fluctuations,
•By charging for such transfers, it effectively determines the net interest income of
business units,
1A selective history of these events can be found in [van Deventer et al. 2005].
Revisiting Funds Transfer Pricing 4
•Because banks acquire interest rate exposure in the process of funding their balance
sheets, FTP is perceived as well to be the mechanism of charging for funding costs
and as a tool to manage liquidity risk of the bank.
It has to be noted however, that equating management of interest rate risk to allocating
the costs for funding can be an oversimplification in today’s banking organizations. The
simplistic and directional view that higher interest rates increase the costs of funding and
this risk needs to be managed against seems to be a reminiscent of times where all the loans
(mortgages) in the banking books were fixed rate (as far back as the 70s and 80s). Today
banking books have almost as much of variable rate loans indexed to a variety of alternative
indices (which create basis risks) as they have fixed rate loans. Moreover, interest rate man-
agement transactions sometimes are carried out between a banking unit (like retail) and a
swap desk in the Capital Markets division, while the Treasury charges a fixed rate for an
overall use of funding resources2.
In the aftermath of the recent financial crisis, FTP has regained its prominent role as the
key tool in measuring and managing the liquidity risk in banks. We will follow the funding
cost allocation side of the FTP and will acknowledge the interest rate exposures aspect in
passing, distinguishing between the two in examples.
The mechanism of FTP is dictated by the very nature of the banking business. In the course
of their day-to-day business banks either lend or take deposits independently. As a result,
business units either end up being short of funding for lending or in excess of it and are look-
ing to invest. The Treasury of the bank owns the process of transferring the funds internally
from businesses that have the excess to the ones that need the funding. In the process Trea-
sury charges a rate for the funds provided to, and pays a rate for funds purchased from the
business units (the FTP rate). This results in a decomposition of the net interest margin
(NIM) of the bank into three components - the lending business NIM, the deposit busi-
ness NIM and the Treasury NIM (here Treasury NIM and FTP NIM are the same and will
be used interchangeably). Fig.1 illustrates the FTP process and decomposition of thee NIM.
In this example Treasury (its funding center) has acquired $1,000 of deposits at 3%, to fund
a loan of the same amount, passing the 3% rate as an FTP rate to the lending unit as an
interest expense. In this transaction neither the bank as a whole nor the units involved
(including the Treasury) have taken interest rate or liquidity risks, as reflected in the zero
FTP NIM of the Treasury. Notice that the bank NIM of 4% is composed of the three NIMs
NIMLending +N I MD eposits +N I MT reasur y = 3% + 1% + 0% = 4%.
Fig.1 is the simplest of the situations that can occur in a bank. Normally, the deposits are
of shorter terms than the loans. So there is a maturity mismatch that creates both interest
rate and liquidity risk exposures. An example of this is shown in Fig.2 below. Here the same
5-year loan of $1,000 is funded by acquiring $1,000 of 2-year deposits from depositors at
2While there can be a few things that can go wrong with such a setup, the two more significant ones seem
to be the conflict of interest in the positioning of the swap desk and the practical non-feasibility of forming
an aggregate view of the interest rate exposure for the bank as whole. Nevertheless, today such practices
are fairly wide spread to be ignored.
Revisiting Funds Transfer Pricing 5
Figure 1: The mechanics of FTP - no interest rate or liquidity risks.
1%. First, notice that the bank in this scenario generates a greater NIM of 6% −1% = 5%.
To achieve this the bank exposes itself to the risk associated with being able to roll the
2-year deposit until the loan is repaid (liquidity risk) and the risk of paying higher rates for
the 1-year deposits if the interest rates go up in the process (interest rate risk).
As can be seen from the Fig.2, FTP process has moved both risks into Treasury, plus the
extra 1% NIM for assuming the responsibility of managing these risks.
So the basic question arises, how does the Funding Center decide how much to pay and how
much to charge for the funds it acquires and re-distributes (the 3% in Fig.1 and 2% in Fig.2?)
To develop the discussions further we will maintain the business structure described in Fig.1
and Fig.2 - a lending unit, a deposit unit and the Treasury. We will also assume that the
Treasury owns a Funding Center, which runs a FTP book that has all the transfer priced
assets and liabilities of the bank.
At a fairly high level there are two main approaches that Treasuries employ to determine
the FTP rate applied to business units and the funding center - pool approach (single and
multiple pool approaches) and matched maturity approach.
Revisiting Funds Transfer Pricing 6
Figure 2: The mechanics of FTP - with interest rate and liquidity risks aggregated to the
Treasury.
2.1 Pool Approaches
In the simplest case the Funding Center nets the excesses of some business units with the
deficits of others. A central pool lends to deficit units and purchases the excesses from
others and uses the same rate for both (see Fig.3 below). It is a simple approach, but has
drawbacks that are difficult to ignore.
First, since the assets and the liabilities are matched at the business unit level before iden-
tifying the shortfalls and excesses of funding resources, the Funding Center does not know
the maturity profile of the pools and will operate on an average rate basis. As such it will
always leave both interest rate exposures as well as funding costs unmatched.
Furthermore, since it has no control over the business unit operations the maturity mismatch
profile of the excess pool or the funding required is not known to the Funding Center a priori.
The size of the unknowns and the success of such a simplistic framework depend on how
homogeneous the products of the respective business units are. Therefore, in today’s bank-
ing environment where products are increasingly structured and with embedded contingent
cash flows, the single pool approach is increasingly looking like an oversimplification.
The next level of sophistication in FTP approaches moves in the directions of the shortfalls
of the single pool approach and is referred to as a multiple pool approach. This approach
acknowledges the maturity structure during the netting of the assets and liabilities of the
Revisiting Funds Transfer Pricing 7
Figure 3: Single pool approach to funds transfer pricing.
business units, as well as product-specific features to produce pool-specific FTP rates. Pool-
ing may be based on maturity structures of the products, re-pricing of indices, as well as
other specifics like behavioral patterns.
While a clear enhancement, the approach still suffers from dependencies on pool averages
and assumptions made about the acceptable level of granularity and the number of pools -
how many pools adequately reflect the maturity mismatch profile and the funding costs of
the bank as a whole correctly?
2.2 Matched Maturity Approaches
The approach that is recognized today as the most adequate one for achieving the goals of
an FTP framework is known as the matched maturity transfer pricing. Under this approach,
FTP rates charged for the use of funds and rates credited for providing funds are based on
matching the rate on the cost of funds curve to the maturity (or the arrival/departure time
of each principal cash in/out flows) of the asset or the liability instruments. To do so, all
the assets and the liabilities are first transferred into a central book referred to as the FTP
book Fig.4. This is structurally different from the pool approaches in Fig.3 and serves both
goals of FTP - allocation of the cost of funding and construction of the interest rate risk
exposure for structural risks hedging (both for structural interest rate and foreign exchange
risks).
Variations exist to the picture in Fig.4. For example, pools could still be formed after all
the assets and liabilities have been transferred into a FTP book transaction by transaction.
Revisiting Funds Transfer Pricing 8
Figure 4: Matched Maturity approach to funds transfer pricing.
This is however counter-productive, since it loses the level of granularity that generates the
precision in assessing the liquidity profile and interest rate risk.
Experience shows that the moment banks move to multiple pool approaches, the incremental
effort to make the switch to maturity matched transfer pricing becomes smaller and smaller,
and banks usually adopt the latter as an approach for the whole balance sheet.
For this reason in the remainder of this paper we will discuss the various aspects of funds
transfer pricing using the matched maturity approach.
Revisiting Funds Transfer Pricing 9
3 Matched Maturity Funds Transfer Pricing
A matched maturity of funds approach to FTP is currently considered to be the most ade-
quate one by many practitioners. As mentioned before, under this approach rates charged
for the use of funds and rates credited for providing funds are based on matching the ma-
turity profile of principal cash flows of asset and liability instruments to the FTP rate that
corresponds to that maturity on the cost of funds curve (an example of such curve is shown in
Fig.5). After this rate is identified, shadow interest expense (credits) to assets and shadow
interest income (debits) to liabilities of the business units are created following a double
entry accounting mechanism. Simultaneously, mirror positions of these shadow entries are
created in the FTP book.
We will first follow a few key examples to illustrate the mechanics of funds transfer pricing
of assets and liabilities using the matched maturity approach. We then will explore some of
the theoretical aspects of this approach. Parallel to assigning FTP rates and the calculation
of the NIMs created, we will identify the interest rate risk exposures for each of the cases
discussed and will plot the maturity mismatches according to the liquidity risk created. We
will use the FTP curve in Fig.5 as our working example for the transactions discussed bellow.
Figure 5: An example of a FTP curve.
Revisiting Funds Transfer Pricing 10
3.1 Transfer Pricing of Assets
Consider a 5-year loan that pays 6% annual interest. The FTP book reflects the matched
maturity funding of the loan, where the FTP rate of 3% is picked from the 5-year point on
the cost of fund curve (see Fig.5) of the bank and is passed through to the lending unit as
an interest expense. Such a transaction fixes the net interest margin across the whole bank
at 3% (6% −3% = 3%).
Figure 6: Transfer pricing a fixed rate loan.
As can be seen from Fig.6, this transaction does not create either an interest rate exposure
or liquidity risk, since both interest rates are fixed and principal cash-flows are matched in
time. The only risk that the loan carries for the bank at this point is the default risk of the
borrower, management of which is the main competency of the business unit.
Fig.7 illustrates how interest rate exposure and liquidity risk are acquired for the same
transaction due to variations to this simple picture resulting from taking a view on either
liquidity or the interest rates.
The Funding Center (which runs the FTP book), is of the view that wholesale markets are
liquid enough and it will be able to roll the wholesale funding position at the end of the
first year either for another year or borrow at more beneficial 4-year rates at that time. The
Funding Center has created an extra net income margin of 1% in the FTP book and has
Revisiting Funds Transfer Pricing 11
Figure 7: Transfer pricing a fixed rate loan with views.
enhanced the bank NIM overall, albeit by taking liquidity and interest rate risks. The in-
terest rate risk taken by the Funding Center needs to be also allocated an adequate amount
of risk capital.
Consider now the same 5-year loan that instead pays an interest indexed to 1M LIBOR and
equal to LIBOR + 350bps. The Funding Center funds the loan with a 5-year floating rate
note indexed to 1M LIBOR, for which it pays 1M LIBOR + 50bps. The 50bps spread is
referred to sometimes as the liquidity premium. The funds transfer pricing of a floating rate
loan creates the balance sheets and net interest margins shown in Fig.8.
Again, in this example we have matched the interest rate index (and its tenor) explicitly,
which generates a zero NIM for the FTP Book and locks a 300bps NIM for the business
unit, which in its turn rolls up to the NIM for the bank as a whole. No interest rate or
liquidity risk has been acquired.
In this situation, however, there are a few things that the Funding Center could do to take
a view - create exposures to liquidity risk (maturity mismatch), take interest rate risk (by
funding at a fixed rate or at different terms) and generate a basis risk (by mismatching the
index or its tenor). All of these actions create the same situation where the funding center
is taking risks (different risks and different mechanisms, but the same outcome) and has to
Revisiting Funds Transfer Pricing 12
Figure 8: Transfer pricing a floating rate loan.
provide return on the capital allocated for covering those risks3.
We illustrate in Fig.9 only one of these cases, where liquidity risk exposure gets created by a
maturity mismatch. In this case the Funding Center takes a view that the bank-specific asset
swap spreads will tighten (and hence the liquidity premiums will decrease). Accordingly the
funding center is funding the 5-year floating rate loan with a 1year floating rate note, indexed
to the same interest rate. Fig.9 shows the balance sheets and the net interest incomes of the
business unit, funding center and the bank as a whole. This of course creates the liquidity
risk of not being able to roll the 1-year note in one years time or issuing a 4-year note on
the same terms. It also exposes the bank to the volatility in asset swap markets, with the
risk of spreads widening at the short end of the term structure.
3Allocating risk capital to liquidity risk exposures usually does not help mitigate liquidity risks, so we
are talking about allocating capital to interest rate and basis risk.
Revisiting Funds Transfer Pricing 13
Figure 9: Funds transfer pricing a floating rate loan with a view.
In the examples considered so far principal cash flows of the loans were arriving at maturity.
This is usually not the case, and when principal cash flows come with a certain schedule
they need to be matched to their time of arrival, not the maturity of the transaction. This
creates a principal cash flow weighted FTP rate to be assigned to the whole transaction. An
example is shown in Fig.10. Here the FTP rate is 2.20%, not 3.00% as it would have been
in the case of a single principal cash flow at the maturity point (5 years).
Revisiting Funds Transfer Pricing 14
Figure 10: Principal cash flow weighted FTP rate calculation.
3.2 Transfer Pricing of Liabilities
As the banking industry was reminded during the recent financial crisis, deposit collecting
businesses are providers of very valuable funding sources. Because banks offer deposits at
much lower rates than the rates on available wholesale funding, FTP allocates this oppor-
tunity cost benefits to the deposit collecting units. Consider, for example, a certificate of
deposit of $1,000 offered by a bank that pays 1% to the holder in 1 year time. The deposit
unit sells the funds raised to the Treasury at the price of the analogous 1year funding avail-
able from the wholesale markets (Fig.11), which is 2% at that point in time (cfr. Fig.5).
This transaction does not create any interest rate or liquidity risk, because the terms of the
both sides of that particular transaction are matched.
In the examples above the loans were matched with a deposit product of an equal maturity
for the sake of simplicity of the exposition of maturity matched approach to FTP. In real-
ity, however, deposits fund banking assets with maturities much longer than the deposits.
So inevitably, funding loans with deposits creates a negative maturity mismatch for banks.
Such a direct interpretation (or use) of the deposit contracts would infer that banks have to
refinance some portion of their deposits almost daily, to cover these negative maturity gaps.
Herein, however, hides the real value of the deposit businesses for a bank that many deposit
businesses feel that FTP frameworks do not give enough credit for, when the matched ma-
turity approach is implemented as mechanically as above.
Deposit products at a high level can be divided into two main categories term and non-
Revisiting Funds Transfer Pricing 15
Figure 11: Funds transfer pricing deposits.
term, and both can be interest bearing or not. Truth of the matter is, however, that deposits
hardly ever behave as they are contracted to.
For example, customers with term deposits may decide to roll forward (sometimes even
with increased balances), rather than withdraw the expired deposit contracts. On the other
hand, a deposit instrument with no stated maturity contains no penalties for withdrawing
the whole balance on the account with no or little notice (a no-cost put option in the con-
tract design). So to understand the behavior of the deposits from the standpoint of the
behavior of principal cash flows and the associated ”liquidity value” one has to view them
in bulk and with respect to the macro-environment (interest rates, economics and other
collective behavioral specifics). In bulk deposit balances demonstrate stable patterns or
”stickiness” with liquidity life characteristics resembling those of a term debt. This is of
course of fundamental economic benefit to the banks, and as such should be reflected in
the FTP methodologies to create the right incentives in the growth of deposit business lines
[McGuire 2004] and [Turner 2008].
Consider, for example, a situation where a large portion of 1-year deposits rolls into another
1 year holding period at least twice. This creates a cash outflow with a 3-year tenor, with
the cost of a 1-year deposit. The bank will be able to fund a 3-year loan from the pool of
such sticky deposits (see Fig.12).
Bank’s NIM is again a combination of the NIMs for the deposit unit - 1.70%, the lending
unit - 1.30% and the FTP book - 0%, equal to 3% = 4% −1%.
Revisiting Funds Transfer Pricing 16
Figure 12: Funds transfer pricing premium for ”sticky” deposits.
The 1.70% margin for the deposit unit can be decomposed into two components - whole-
sale market component - 100 bps and a stickiness component - 70 bps (Fig.13). While the
Figure 13: Decomposition of the FTP rate for sticky deposits.
wholesale component is an external to the deposit unit (and the bank) factor, the stickiness
component is driven by the behavioral characteristics of the bank deposits. Notice as well,
that the success of funding the bank assets with sticky deposits in a liquidity risk neutral
manner (neither the bank, nor the deposit and lending business units incurred maturity
mismatches, in Fig.12) rests entirely on how well the deposit unit knows its customers and
how well it tailors the deposits products to create more of the type of deposits that exhibit
reliable sticky behaviors.
Revisiting Funds Transfer Pricing 17
3.3 Methodological Aspects of the Matched Maturity Approach
Firstly, let’s acknowledge that in all the examples considered above the matched maturity
FTP mechanism has achieved the following:
- the liquidity and interest rate risks where aggregated into a central point in the bank,
- business unit P&L’s did not carry interest rate exposures, neither were they exposed
to liquidity risk,
- the liquidity risk and interest rate exposure of the FTP book, as well as its P&Lfor
those exposures matched that of the bank as a whole,
- the views on liquidity risk and expectations about the interest rates were taken at the
appropriate point in the organization, where it can be measured and managed,
- the point where risk capital for (structural) interest rate risk exposure of the bank
should be allocated was made easily identifiable.
The last bullet point has implications for the question whether the Funding Center should
be a profit or a cost center. In the most general case it can be a profit center in terms of
the NIM, but a cost center with respect to economic profit or on a required return on equity
basis. The size of this cost center (together with other contributing factors) could help deter-
mine the level of the risk tolerance for interest rate and liquidity risk for the bank as a whole.
Because banks lend and borrow through hundreds of transactions a day, a functioning and
self-consistent FTP framework, built on solid theoretical fundamentals is required to enable
revelation of the implied views/bets on liquidity and structural risks embedded in a banks
balance sheet that form over time, one transaction at a time.
One of the most central elements that form the methodological foundations for calculating
the funds transfer pricing rate in the matched maturity FTP approach is the choice of the
benchmark cost of fund curve (like the one in Fig.5). The choice of such a curve reflects (or
implies) both the philosophy and the approach to managing liquidity and structural interest
rate risk on the balance sheet of a bank. Usually, once this curve has been chosen, banks
have a few add-on spreads to it that range from bank-specific direct and indirect operating
costs to spreads for optionalities embedded in the transfer priced instruments (e.g., prepay-
ments in mortgages or putable deposits, costs of holding a portfolio of liquid assets, ect.).
From microeconomics standpoint, it can be shown that the optimal transfer price for an
intermediate good between two business units is its opportunity cost [Hirshleifer 1956]. Fur-
thermore, if the units are free to determine their own outputs4and the market for the
intermediate good is competitive then the optimal transfer price for it is its marginal cost in
the market. In the case of funds transfer pricing, where the intermediate good is the funds
from depositors, the choice for the benchmark cost of funds curve becomes the curve for
marginal cost of funding for the bank in the wholesale markets[Ford 2009]. For banks that
4For example, how much the lending business lends does not affect how much deposit is taken in and the
imbalances in divisional outputs can be met in the external market.
Revisiting Funds Transfer Pricing 18
have active access to funding in the wholesale markets, this curve has been the AA curve
for a while, which roughly reflects the ratings of almost all wholesale market participants.
For more than a decade before the crisis this curve has been practically replaced by the swap
curves, reflecting (among other things) the short-term nature of funding of bank balance
sheets. the spread between the two curves hovered around 10 bps for a very long time.
The recent financial crisis showed that this spread is as volatile as the curves themselves
and can have a non-trivial term structure. FTP methodologies should capture both the
size of this spread as well as its term structure. This spread appears in the FTP par-
lance as the liquidity premium. In some organizations there is a cultural misconception
that being charged this premium is a punitive measure. This sometimes can be traced to
a methodological misconception that matched term means matching the tenor on the in-
dex of a variable rate loan (for example, 1 month LIBOR), rather than its maturity term,
overlaid with the expectation that the loan can be funded by rolling this short term funding.
Another methodological misconception is around the choice of a single or multiple benchmark
cost of funds curves. For example, banks may select a special mortgage curve to transfer
price mortgages while using the institutions wholesale funding curve for all remaining assets
and liabilities. Mortgages are especially appealing candidates for separate benchmarks since
these assets have counterparts, such as mortgage portfolios and mortgage-backed securities,
which are traded in financial markets. While this may make the multiple curves option seem
reasonable at first glance, the actual consequences turn to be contrary to what an FTP is
trying to achieve, i.e. inaccurate hedging, incorrect measurement of treasury/funding center
performance and misguided product pricing [Shih et al. 2000].
The other essential component in building the FTP framework is the development of the
maturity profile for the deposit products of the bank.
Despite the fact that banks have ample amount of historical data, the development of a sin-
gle well-understood approach to constructing the maturity profile of the deposit portfolios
remains a little more than an artisan craft in many banks today.
One of the two widely used approaches is the tranching approach whereby the analysis
determines a portion or ”tranche” of a pool of appropriately chosen deposit products as
”core” or long-term stable tranche, based on historical analysis of past cash flow patterns
in balances. This tranching continues until some amount is apportioned to a ”non-core” or
volatile tranche. Accordingly, the stable or core tranche of the deposit pool gets a long-term
FTP rate matched to it, while the non-core tranche gets assigned to it a short-term FTP.
The practice of what’s a short-term and what’s a long-term maturity appropriate for these
tranches has a very dispersed range of values between banks.
The two together determine a blended FTP rate for the pool of deposits under the analysis.
The second, simulation based, approach models both the balances and the interest rates on
deposit products through factor models with stochastic terms. The types of factors used
are usually split into two groups the interest rate and macro-environment (current rates
Revisiting Funds Transfer Pricing 19
offered, competitor rates and fees, etc.) and the pool specific factors (levels of service and
convenience, customized products, customer mobility, local demographics, etc.). The simu-
lation allows forecasting maturity profiles of deposit pools which can be used for assigning
maturity matched FTP rates. The resulting maturity profile in this approach is considerably
more granular and lends itself to more flexible and proactive deposit pricing and/or liquidity
management.
A third methodology uses replicating portfolio approach, whereby the historical cash flow
behaviors of the deposit pool are replicated by a portfolio of risk-free instruments (some-
times both cash and derivatives). Unfortunately, as with any replicating portfolio approach,
the resulting liquidity profiles and FTP rates tend to be extremely sample dependent and
unstable in time. Consequently, this approach has struggled to achieve as wider acceptance
as the first two.
In the aftermath of the recent financial crisis and the acute dependence of banks on more
stable and diversified funding sources, banks have found themselves in need of more com-
prehensive study and model building for their deposit pools.
The difference between getting the FTP fundamentals right or wrong when building a FTP
framework can be the difference between a functioning and disfunctional FTP framework.
3.4 Transfer Pricing Equity Capital
Equity capital plays a special role in a bank. The presence of the adequate to the asset risks
amount of equity capital allows banks to borrow at favorable rates in capital markets. As a
result, all (risky) assets on a balance sheet of a business unit are funded with debt capital
with only few exceptions.
More typically, equity capital gets allocated to the risky assets of the business units as a
risk capital - to cover potential losses, and is consequently used for performing risk/return
and portfolio analysis.
The cost of equity capital that should be charged for the two cases - equity funding and risk
capital allocation, are different. In the first case it is the full cost of equity capital equal to
the required return on equity5charged to the business unit as a transfer price for the equity
capital used for funding. In the second case, it is the full cost of equity capital less the
return on a risk-free investment (it is the same risk-free rate that goes into the estimation
of the cost of equity capital) applied to the amount of risk capital allocated.
This netting of the cost of equity capital with a risk-free rate happens in the form of an
equity credit rate, credited to the business unit P&L.
5The required return on equity capital is usually measured as risk-free rate + beta of the bank ×equity
market premium.
Revisiting Funds Transfer Pricing 20
When and how much of such a credit should be applied, has a wide variety of interpreta-
tions and calculation methodologies throughout the banking industry. The industry practice
seems to revolve around two main candidates for the equity credit rate - a long-term FTP
or a risk-free rate, both calculated top-down in most cases and often with some misguided
reference to the duration of the equity.
The correct equity credit rate is different for the two cases of capital consumption.
Fig.14 illustrates the equity credit rate calculation for the case where equity capital is allo-
cated as a risk capital to the business unit.
Figure 14: Transfer pricing equity capital - risk capital allocation.
The case for which the FTP rate enters the equity credit rate calculation arises when an
asset is partially funded with equity capital, but the FTP system has booked it as a fully
debt-funded asset (see also [Shih et al. 1997]. In such cases the equity credit rate becomes a
”blended rate”. This is illustrated in Fig.15. The FTP rate is credited back to the business
unit P&L, applied to the equity-funded portion of the asset and simultaneously the cost of
equity capital is charged to the P&Las the cost to the portion funded with equity. And
since the whole transaction needs to be charged the transfer price for the allocated risk
capital in the form of the netted cost of equity capital as described above, the equity credit
rate becomes a weighted average rate between the FTP rate and the risk-free rate. Notice
Revisiting Funds Transfer Pricing 21
Figure 15: Transfer pricing equity capital - risk capital allocation and equity funding of
assets.
that now the economic profit for the business unit is lesser by $3, which is the difference
between the equity and debt funding (9% −6%) applied to the amount funded by equity
capital ($1,000 −$900), the more expensive form of capital.
An explicit derivation of a formulaic expression for the equity credit rate can be found in
[Tumasyan 2010].
4 Closing Remarks
A faulty FTP methodology or a biased FTP framework will send wrong signals one trans-
action at a time, over time creating an unintended balance sheet structure in the form of
disproportionate mix of asset portfolios on the asset side (e.g., the size of subprime loans
and trading portfolios), and/or liquidity holes on the liability side (e.g., Northern Rock).
In many respects an effective FTP function is the first line of defense for a bank’s balance
sheet and its business model.
Revisiting Funds Transfer Pricing 22
References
[van Deventer et al. 2005] Donald R. Van Deventer, Kenji Imai and Mark Mesler Advanced
Financial Risk Management: Tools and Techniques for Integrated Credit Risk and In-
terest Rate Risk Management. Singapore: John Wiley and Sons (Asia) Pte Ltd.(2005).
[McGuire 2004] McGuire, W. J. (May, 2004). Indeterminate Term Deposits in FTP: Quan-
tified Solutions at Last! Journal of Performance Management,pp. 14-26
[Turner 2008] Turner, S. (2008). Funds Transfer Pricing:Cracking the Code on Deposit Val-
uation. Novantas White Paper Series, 1-4 and references there in.
[Cornyn et al. 1997] Cronyn, A. G., Klein, R. A., and Lederman, J. (1997). Controlling and
Managing Interest Rate Risk. Prentice Hall Pr. for a review of such models,
[Dewachter et al. 2006] Dewachter, H., Lyrio, M., and Maes, K. (2006), A multi-factor
model for the valuation and risk management of demand deposits. National Bank
of Belgium
[Kalkbrener et al. 2004] See for an example Kalkbrener, M., and Willing, J. W. (2004).
Risk management of non-maturing liabilities. Journal of Banking and Finance, vol.
28, 15471568.
[Brickely et al. 1995] See for example, Brickley, J., Smith, C., and Zimmerman, J. (Summer,
1995). Transfer Pricing and the Control of Internal Corporate Transactions. Journal
of Applied Corporate Finance , 60-67,
[Hirshleifer 1956] Hirshleifer, J. (1956). On the Economics of Transfer Pricing. Journal of
Business, 29 , 172-174
[Ford 2009] Ford, G. (2009). Internal Pricing in Financial Institutions: Issues. Macquarie
Graduate School of Management.
[Shih et al. 2000] See Shih, A., Wofford, S., and Crandon, D. (2000). Transfer Pricing: Pit-
falls in Using Multiple Benchmark Yield Curves. Journal of Performance Management
, 33-46.
[Tumasyan 2010] Tumasyan, H. Credit Where Equity is Due. Available at www.ssrn.com
[Kipkalov 2009] Kipkalov, A. (2009). Transfer Pricing Capital. Journal Of Performance
Management , 38-46
[Shih et al. 1997] Shih, A., and Tavakol, A. (1997).Making Sense of the Transfer Pricing of
Equity. Bank Accounting and Finance, 47-52.
[IIF 2007] See also Institute of International Finance. (March, 2007). Principles of Liquidity
Risk Management.
[Basel 2010] See Basel III: International framework for liquidity risk measurement, stan-
dards and monitoring. Basel Committee on Banking Supervision, December 2010.
Revisiting Funds Transfer Pricing 23
[Basel 2008] Principles for Sound Liquidity Risk Management and Supervision, Basel Com-
mittee on Banking Supervision. September, 2008
[EBA 2010] European Banking Authority Guidelines on Liquidity Cost Benefit Allocation,
October, 2010
[FSA 2010] Financial Services Authority, Dear Treasurer Letter, July, 2010.
[Grant 2011] Grant J., Liquidity transfer pricing: A guide to better practice, March, 2011
on behalf of Australian Prudential Regulation Authority.
