Transfer Pricing Indeterminate-Maturity Deposits

A financial institution's net interest margin is the result of product pricing and portfolio structure decisions involving multiple functional profit centers. Senior managements use profitability measurement systems to quantify the amount to which lending, treasury, deposit-gathering and asset-liability

mismatching activities contribute towards risk-adjusted earnings. A quality performance assessment framework includes tools that assist line managers to identify the costs incurred or value created in the course of running their business activities. Armed with these informational inputs, the businesses are better positioned to achieve the profit contribution goals they are being held accountable for.

Funds transfer pricing (FTP) is a management information system that establishes the value of funds gathered or deployed for the purpose of measuring net interest income contribution at the transaction level. Charges are levied against financial assets and credits allotted to funding liabilities so as to mirror a record's repricing and cash flow attributes. In the process, lending and deposit units obtain a "balanced" balance sheet and a net interest spread per record is established. The offsets to FTP charges and credits collectively represent that portion of the institution's net interest spread attributable to repricing and cash flow mismatches. The FTP system transfers mismatch earnings from the business units into the Funding Center (FC). Once isolated and quantified, the asset-liability committee (ALCO) can monitor the level, trend and volatility of earnings related to its assumption and supervision of interest rate risk.

Matched-term transfer pricing is widely regarded as the most conceptually pure approach to running FTP. A unique charge or credit rate is assigned to each record as of its origination date on the basis of four characteristics: repricing tenor, repricing spread, contractual cash flow and embedded options. The FC acts as the central clearinghouse for funds provided and used, benchmarking the all-in transfer rate against a market-derived yield curve plus other market pricing signals. Calculated charges and credits are applied to book balances until the earlier of next repricing or maturity.

An enduring enigma surrounds the question of how to formulate transfer rates for indeterminate-maturity deposits (IMDs). The matter attracts considerable attention because demand deposits present clear challenges involving the four factor elements that guide matched-term transfer pricing. This paper will survey the popular transfer pricing approaches, from simple to complex, and describe why these techniques can misrepresent margin contribution analysis. Transfer pricing at the short-end of the benchmark curve represents an improved solution, avoiding many of the pitfalls to conventional methods. Furthermore, the manner of developing liquidity adjustments in FTP is clearly explained. The paper will close with a discussion of implementation issues surrounding true four-factor IMD transfer pricing, including implications for profitability results.

ELEMENTS TO MATCHED-TERM FTP

Business units are more accepting of FTP results when management's transfer pricing policies have a transparent, rational basis and are applied consistently throughout the organization and across time. Matched-term FTP develops the assigned marginal rates for funds supplied and used based on mirroring each record's repricing and cash flow factor elements in the capital markets, as discussed below.

Repricing tenor: This factor takes into account the timing to reset for the record's offering rate. Given a fixed-rate, option-free record, the repricing tenor component is determined by observing the benchmark rate for matched-maturity funds. With IMDs, FTP must assign a repricing tenor to an instrument that has no defined maturity and appears to lack a formal, defined repricing schedule.

Repricing spread risk: This element arises when there is an absence of perfect correlation between rate indices on sources and uses to variable-rate funds. A common example is the prime rate, the interest rate at which financial institutions lend to their best and most creditworthy borrowers. In general, the prime rate has unique pricing drivers and may lack a direct link to the benchmark funding curve. The decision to lend based on prime exposes the business unit to earnings volatility if there is a change in the basis or spread between this index and PC's short-tenor funding rate.

The institution's interest rate risk measurement and control policy may elect to centralize the management of prime's repricing spread risk in ALCO. FC can remove spread risk from the line unit's performance and guarantee a stable net interest spread contribution on the prime-indexed loan. To do so, FC embeds a spread premium that "risk-adjusts" the base transfer price, thus assuming the risk/reward responsibility on its book. The appropriate charge to isolate the risk may be indicated in the swaps market. If such pricing signals are unavailable, statistical analysis can suggest the average spread most closely representing prime's repricing behavior versus the short rate.

At many institutions, a deposit pricing committee or ALCO assumes administration of IMD offering rates, with guidance from the deposit-gathering units. Adherence to the core philosophy behind matched-term FTP benchmarks the IMD transfer credit based on the institution's marginal funding rates. Under normal market conditions, non-maturity deposits price at favorable spreads under the cost of short-term wholesale debt. Changes in the offering rate usually lag movements in market yields. Moreover, when ALCO does change deposit pricing, the decision may manifest as a partial adjustment to the market rate shift.

Added together, these deposit repricing elements bring significant margin contribution value to the institution. Some of the approaches to transferpricing IMDs described in this paper intend to price and then concentrate within Funding Center an expected value for the repricing basis. However, it must be mentioned that unlike for the prime index, today's capital markets do not trade an instrument that values and hedges the repricing spread of a non-maturity deposit. This fact will have important bearing on the FTP manager's decision as to which business unit, Funding Center or the deposit gatherers, ultimately bears uncertainty inherent in the IMD repricing spread.

Contractual cash flow: Variable-rate instruments often have a mismatch between their repricing tenor and future principal cash flows. If a business unit originates a five-year loan repricing every 90 days, matched-term FTP should recognize funding is required to support the loan until maturity. Proper isolation of this record's rate sensitivity and term attributes bases the transfer rate on the institution's marginal cost to issue five-year term debt, repricing quarterly. The loan's liquidity "life" can be acknowledged by adding a liquidity commitment spread adjustment to the repricing tenor-matched cost of funds.

An indeterminate-maturity deposit has no stated maturity but contains a no-cost put option in its contract design. The customer may withdraw deposit balances at book value from the institution at any time, generally without limitation. Some customers exercise, while others allow the option to expire and roll balances into the following day, possibly even adding balances. A strict interpretation of the deposit contract infers the bank refinances some portion of its non-maturity funding every day.

Overall, deposit balances may demonstrate stable retention patterns under a variety of interest rate conditions. Deposit units that cultivate and retain "sticky" customer balances generate enduring economic benefits. Opportunity costs to holding liquid assets in support of volatile balances are avoided, while liquidity life characteristics resemble term debt. Several approaches exist to valuing the liquidity commitment spread, the most popular being estimation of a behavioral repayment schedule.

Embedded options risk: FTP processes consider transfer rate break charges or option cost adjustments, as appropriate, in instances where explicit options may alter the instrument's contractual cash flow. For example, certificates of deposit usually allow for early withdrawal (from the institution's perspective, a short put option), and may be issued as cancelable at the bank's election (a long call option). If the economic costs are deemed meaningful, FTP can be the conduit to provide business managers indication on the value of these embedded options. Offering rates may then be structured to fully consider the current market's marginal cost for comparable funding. Asset-liability managers closely monitor customer behaviors for signs of IMD option exercise. Product disintermediation due to rate- or service-related factors can have major consequences on the portfolio's overall funding costs and ALCO objectives. However, no term risk opportunity costs are created when an institution replaces withdrawn balances with identical funds. The IMD offering rate is variable, and all records in a product group accrue interest expense at the same managed rate. The option's impact on liquidity can be addressed via FTP's cash flow factor element.

CURRENT RATE METHOD

Now, let's turn attention to several prominent methods for determining indeterminate-maturity deposit transfer rates. Given ambiguous rate-setting as well as term characteristics, most IMD transfer pricing techniques involve developing mechanical representations of the repricing and cash flow characteristics. Certain product types like money market deposit accounts (MMDAs) adjust either directly or vary closely with a market index. The current rate method transfer prices the deposit at rates that are in place for the current processing period. Changes in market rates are immediately and fully reflected in the assigned transfer rate.

A frequently-voiced complaint regarding the current rate approach is the resulting transfer price may not provide period-to-period profit lock-in to the deposit gatherer. As the argument goes, deposit units are not immunized from changes in market rates, the end result being assumption of interest rate risk. The grievance is misguided, since deposit pricing committee's administration of the offering rate is in itself an explicit act of fixing a basis versus money market rates. Simply stated, the deposit gatherers "create" the repricing spread "problem" in issuing managed-rate liabilities, which of course is a worry any financial institution would enjoy having to wrestle with. Variable-rate savings technically sees its rate reset daily based on an institution-specific determination of the appropriate repricing spread. Any lag in adjusting the offering rate actually represents an overt decision to the leave that rate unchanged. The oft-made suggestion that the deposit coupon assumes fixed-rate features because of committee inaction is inherently flawed.

Furthermore, one role of FTP is to value a record's repricing elements and assign accountability for spread management, not to insulate or remove interest rate risk. The latter task is accomplished via hedging, not transfer pricing. For example, mortgage loan prepayments expose the business unit to reinvestment risk despite implementation of option-adjusted matched-term transfer pricing. By its nature, an IMD has a variable month-to-month profit contribution that is deliberated upon by ALCO in the context of the present and expected future rate environment, deposit supply trends, profitability goals and other factors. The current rate approach is a solid conceptual start to transfer pricing IMDs, although additional refinement is needed in order to address the cash flow factor elements.

Case Study in Current Rate Method

An institution seeks to develop a transfer rate for their interest-bearing checking product. To determine if a suitable repricing benchmark can be found, log linear regression of the historical offering rate against various market indices is performed.1 Exhibit 1 examines two rate environments: 134 data points for the 1998 - 2000 rising rate scenario, and 105 observations during a declining rate trend between 2001 and 2002. Regression results detect the strongest statistical association (highest r-squared) exists between historical deposit offering rates and three-month LIBOR, which is then selected as the transfer pricing benchmark for the IMD.

AVERAGE RATE METHODS

A well-known feature of the IMD is the magnitude and timing of offering rate adjustments are administrated by committee. Numerical representation of this process is complicated by the fact management may set rates asymmetrically in rising versus falling market rates environments. This idiosyncratic repricing behavior is most frequently observed in retail checking and statement savings deposit products. Transaction demand accounts represent a special case: in exchange for full access to the payments system, the deposit's stated interest cost is oftentimes zero.

The moving average method, a form of average rate transfer pricing, intends to describe the sluggish pricing observed for these types of liabilities. A time period for the moving average is chosen based on the observed or estimated repricing responsiveness for the deposit category. The selected averaging period may also encompass an attempt to capture the deposit's simulated effective "term" or liquidity life. The current period transfer rate is then read from a moving average of historical data gathered for the selected term point on the pricing curve. The transfer rate will move up or down over time given a trend within the shifting window of past data points.

As demonstrated previously, regression tools may help decide on an appropriate time index for transfer pricing. Exhibit 2 shows log linear regression statistics for monthly savings rates as reported by CUNA for a sample of credit unions.2 When market rates declined in the years 2001 - 2003, the average savings rate fell from 3.1% to 1%: regression reports a strong statistical relation between the savings rate and three-month LIBOR (r-squared of 86.7%). An even stronger association is found with the 60-month moving average of five-year U.S. Dollar swap rates, as the exhibit shows.3

However, when market rates are rising, checking and savings account rates may exhibit long repricing delays, then low (or nonexistent) repricing response. Such a "dual beta" to the offering rate was demonstrated in the December 2003 to September 2005 time period. As average short-term market rates jumped 291 basis points, the average savings rate for the CUNA sample of credit unions increased a mere 10 basis points. Like most financial institutions, credit unions successfully exerted savings rate rigidity for reasons including limited competitive pressure for market share (ample funding availability), strong internal liquidity (loan growth funded by liquidating securities) and segmented pricing strategies that encouraged rate-sensitive balances into higher-rate service lines, like certificates.

Standard regression analysis can have difficultly explaining managed rate-setting behavior when checking and savings accounts show sluggish repricing. Quantitative techniques may bias towards longer-term market rates on account of the relatively lower yield volatility generally observed the further out one moves along a market curve. Moreover, moving averages smooth data deviations. For example, rates paid on credit union savings accounts during 2004 - 2005 show poor statistical fit against market rates, the best regression occurring with the 60-month moving average of the five-year U.S. Dollar swap rate (r-squared: 28 percent).

Can moving average regression statistics substantiate the use of long-term transfer rates? The answer is no. Some FTP advisors counsel selection of longer maturity term points for the moving average if the IMD shows "sticky" balance and rate behaviors. Such recommendations overlook the concept that is matched-term FTP: the marginal at-market funding credit should replicate the liability's current repricing and cash flow characteristics. Being historical in construction, moving average transfer rates do not adhere to this view.

Let's examine the case of management electing to transfer price savings deposits in 2004 and 2005 using a 60-month moving average of past swap rates. The approach presumes monthly tranches of five-year bonds are a matched-term analog to the savings product. In reality, the replicated cash flows in the synthetic debt portfolio could have a very different repricing risk profile through time. In fact, the moving average yield declined 92 basis points during this rising rate period. The result is sizeable tracking error in the Funding Center and a profitability signal that is opposite to logic: lagged repricing as rates rise should make IMDs appear more valuable. The use of long-term transfer rates can cause retail deposits to offer lower profit contribution when rates increase. In cases of severe yield curve shape change, the averaged transfer rate could slip below the deposit offering rate.

There are other weaknesses to the moving average method that far outweigh the single advantage of ease to computation. As rates decline, the intermediate to long transfer rate may be sluggish to respond, particularly if the yield curve steepens as is typical when central banks maneuver short rates lower. Furthermore, the moving average transfer rate fails to properly quantify the value of stable "term" funding, even though that is its intent. Correct estimation of the liquidity premium is based on the marginal cost to raise term funds that match the record's repricing characteristics. Fixed-rate points off the benchmark borrowing curve commingle the price of term risk with the cost to produce term liquidity. In summary, the fundamental difficulty with moving averages is transfer rates are not based on the institution's opportunity gain to raising stable, low-cost variable-rate funding in today's rate environment.

BLENDED TERM APPROACHES

There is near universal agreement amongst asset/liability and profitability managers that IMDs convey liquidity value to an institution. Ample empirical evidence suggests the effective "maturity structure" is usually much longer than the contractual demand attribute would indicate. Blended term methods are sometimes presented as a more refined way to integrate repricing basis (mismatch risk) and balance stability (supplied liquidity) assumptions within a transfer rate.

Blended term analysis initiates by assuming a portion or "tranche" to a product line as "core" or long-term stable based on historical analysis of past cash flow patterns in balances. Non-volatile balance profiles secure longer moving averages. Balance/time segmentation continues until some amount is apportioned to a "non-core" or transitory tranche, which receives a short transfer rate. The segmentation process for the "blended term" can be made quite elaborate, as in the case of behaviorally-weighted cash flow transfer pricing. Such approaches compute the blended rate based on a forecast of the pattern of expected remaining deposit balances over time.

Proponents suggest blended term transfer pricing allows for more granular identification of balance volatility, customer age or depositor geography parameters. Indeed, careful consideration of these repricing components is essential to prudent and profitable deposit rate-making. However, blended-term methods translate a balance attrition forecast into an expression of the FTP repricing term, producing results that are afflicted by the same drawbacks seen in historically-based average rate methods. Even when current benchmark rates are used, blended term methods only consider the path of expected future market rates (where the fixed swap rate is a geometric average of the implied forward rates). An economic estimation of the effective yield spread to add to the offering rate as an expression of funding contribution value would need to consider the deposit's future repricing as it is exposed to time evolution of those forward rates.

The second element to a typical blended-term analysis is to increment the base blended transfer rate by a liquidity commitment adjustment. The "liquidity premium" can be estimated by observing rate differentials between the organization's wholesale funding curve and the swap curve. Swap rates quote the cost to transfer interest rate risk, so differences between actual funding rates and swaps represent the cost to raise physical liquidity, less term repricing risk. For example, the transfer rate on a five-year loan that reprices to LIBOR quarterly would equal the sum of three-month LIBOR plus the basis point spread between the institution's wholesale funding rate and the swap rate at the five-year tenor.

Exhibit 3 illustrates this spread by plotting market rate data for swaps (solid line) versus Federal Home Loan Bank fixed advance rates (dashed line) as of December 31,2003. With 3-month LIBOR at 1.15 percent and the five-year FHLB - swap spread equal to three basis points, the liquidity commitment-adjusted transfer rate on the five-year floating-rate loan is 1.18 percent.4

Local capital markets may permit direct observance of the institution's actual marginal rate to issue floating-rate term debt, as shown in Exhibit 4 (following page).

With market data in hand, problems associated with blended-term methods rise rapidly to the surface. Say an institution desires to develop a transfer rate for its 0.50 percent savings deposits. Explicit assumptions about deposit balance behavior are prepared. The projection of balance supply combines historical retention data with a forecast of likely stability given the institution's existing pricing policies and expected market competitive pressures. Deposit balances are grouped into three tranches: 50 percent are allocated to a five-year segment, 30 percent assigned to a two-year tranche and remaining balances are considered three-month funding. Under this approach, the deposit would be assigned a 3.57 percent transfer rate credit as of year-end 2003.5

When subject to rigorous examination, the rationale to blended-term transfer rates becomes questionable. Comprehensive methods to transfer pricing will assign contribution value under the bright light of market-supplied pricing signals.6 Exhibit 4 informs us Funding Center's marginal cost to source five-year term, floating-rate debt is 1.18 percent. An option to prepay the borrowing facility can be included for a mere additional 25 basis points. Viewed another way, Bankrate.com reports the 5-year jumbo CD rate was 3.69 percent in December 2003. As an alternative source of funding, the CD locks in substantial liquidity, and when swapped to floating, has a repricing term aligned with the savings account. This case example illustrates how blended-term transfer rates can be difficult to justify on economic grounds.

ALM VERSUS FTP7

At this juncture, it is helpful to distinguish the analytic goals of ALM and FTP as they relate to indeterminate-maturity deposits. Some ALM managers find historic retention data and regression-based repricing drivers bring useful information to deposit simulation in asset-liability models. In some institutions, personnel responsible for forecasting earnings and measuring liquidity risk can be the same individuals running FTP. It would seem natural to apply similar IMD term and repricing logic in both analytic applications. Such inclinations can be a source of material error.

For example, ALCO might determine that a deposit product's cost of funds over the next 12 months is best predicted using a blended index of one-month LIBOR and 10-year swap rates. The objective of matched-term FTP is to cast the alternative opportunity rate of this funding source considering the record's repricing attributes.8 Funds contribution value for the IMD is defined by the tenor to repricing and not by the path of offering rates expected beyond the next repricing. This month's transfer rate on 0.50 percent savings deposits should not be influenced by ALCO's projection that balances are likely to cost some higher or lower rate the following month. If the deposit truly reprices according to the index blend, the correct transfer rate would not be this index, but the rate on market-based alternative funding with monthly repricing. Similarly, ALM and FTP need not always agree on the liquidity life of non-maturity funding. FTP could use historical retention analysis to assign a five-year liquidity premium during the current budget cycle. ALCO will perceive the overall desired liquidity position in the context of behavioral cash flows and liquidity needs estimated for contingent scenarios.

PROPER SPECIFICATION OF IMD VALUE IN FTP

If average rate and blended term methods are deficient for transfer pricing IMDs, how should FTP properly integrate the repricing and cash flow components? The distinctive rate and balance behaviors of demand and savings deposits have been well studied in the finance literature from the perspective of valuation and interest rate risk. A brief review of deposit modeling concepts will provide important insight into "better practice" FTP treatment of these liabilities.9

While varying slightly in terms of technical specification, the foundation of deposit valuation models rests on estimating and discounting the following cash flow elements (ignoring reserve requirements):

* The rate differential between the short-term market rate and the deposit's offering rate;

* Non-interest expenses to service the deposit, net of fee income; and,

* Changes in deposit balances.

The valuation model tells us today's deposits are worth their current balances, less the present value of an economic profit margin."1 The economic profit or "deposit premium" earned by the institution may be influenced by the customer's decision to hold balances or to withdraw them. For example, consider a money market deposit where the offering rate plus net servicing costs equal the market rate in each future time period. No matter the simulated "maturity" of balances, the deposit premium is zero, meaning the liability's value to the institution is its book value. Unless market rates are very low, the profit margin is typically positive, returning an economic value for the IMD that is less than book value. Deposit pricing committees well appreciate the trade-offs to increasing deposit pricing spreads versus the costs associated with slower deposit growth or outright loss of customer balances.

If we wanted to design a hypothetical portfolio that replicates the attributes of an indeterminate-maturity deposit, the hedging structure would have two components:

* Invest deposit balances in a short-term bond, reinvesting at maturity; and,

* Assume a swap obligation that requires payments at the short rate in exchange for receiving the deposit offering rate, plus net servicing costs.

Deposit valuation concepts have powerful ramifications when ported into the world of transfer pricing. The replicating hedge solution indicates the replacing tenor for all types of demand deposits is that associated with short-term, floating rate debt. A pure indexed MMDA accrues no funding spread, whereas variable-rate deposits produce economic rents when sourced at rates below alternative short-term borrowing costs. Each reporting month in FTP, the branch network earns a "deposit premium" given actual retained balances. This transfer pricing approach is consistent with the conclusions of various researchers who view demand deposits as floating rate instruments paying the offering rate i^sub t^ 11

Concise treatment of the withdrawal option is explained in the valuation model. The present value mathematics simulate balances D^sub t^ as fully withdrawn at the end of a period, followed by reinvestment at the beginning of the next period. This reflects previous mention that deposit customers hold a one-period embedded option to withdraw or add balances.

IMPLICATIONS FOR PROFITABILITY MEASUREMENT

Even though the recommended short-rate transfer pricing method rests on solid economic footing, an institution introducing this technique may encounter political headwind. First, resistance might be felt on grounds of credibility: why are we changing approaches now, are other institutions measuring contribution value similarly, and what about the integrity of previous profitability reporting? High-performance organizations seek continuous improvement in all management systems, so a design change should be embraced when based on intellectually rigorous business judgment. In truth, the technical processes to FTP should not be conditional on what management wants the tool to tell them.

The most strident resistance might come from the branch network. A likely outcome of abandoning an average rate or blended term method in favor of the short market rate is a lower transfer credit on all demand deposit balances and reduced perceived profitability. Recall that a crucial role of FTP is to convey accrual accounting pricing signals that are in step with the mark-to-market world. Indeed, the discipline of FTP was forged from the need for a management information system that would correctly allocate profitability and benchmark ALCO's mismatch decisions. Valuation concepts tell us that any transfer rate greater than the short rate front-loads future repricing spread earnings into the current period, likely distorting performance results for the Funding Center.

A simple case study will help crystallize the short rate approach to transfer pricing IMDs. Consider a basic balance sheet containing three-year fixed rate loans plus fixed assets funded by savings deposits and capital. Funding Center quotes a matched-term transfer rate on the loan, thereby transferring repricing mismatch earnings into its book. To hedge the term exposure to this earnings stream, FC executes a three-year, pay fixed swap, receiving LIBOR in exchange. The savings deposits receive a transfer credit equal to LIBOR. Assuming no volume mismatch between earning assets and funding, the FC book is flat (no earnings) and capital sees nil interest rate risk in the hedged balance sheet.

Let's say the short rate transfer credit solution leaves the branch network dissatisfied. In desiring a higher transfer credit, the managers launch a debate over the funding "term" of customer demand deposits. Statistical evidence is presented to demonstrate IMDs have significant balance stability over an economic cycle, "proof that transfer rates corresponding to intermediate- and long-term points on the benchmark curve are appropriate. With historical data suggesting deposit relationships are, say, threeyear money from a funding liquidity view, branches claim a corresponding transfer credit.

The branches should not succeed in their argument. Were asset and liability to be priced off the same term point, all transfer income from the loan is passed through to the deposit unit. Funding Center earnings aim to represent the market-benchmarkcd run rate for holding mismatched positions. With nil income in the rate risk book, ALCO is not being compensated for assuming obvious mismatch exposure. Furthermore, the deposit unit's claim to a piece of the mismatch earnings on the basis of "liquidity" can be rejected because ALCO may decide to physically invest all IMDs short - then enter into synthetic asset positions to assume the desired degree of mismatch risk/reward. Funding Center cannot pay more than LIBOR for IMD funds under normal markets without injecting error into its own FTP-measured margin contribution.

Let's investigate a second, more extreme example. Consider a balance sheet where equity and zero-rate deposits are funding 10-year fixed rate loans. For the moment, assume the liquidity in the deposits is 10 years. Under no circumstances should deposits be transfer-priced at the 10-year rate: rewards for taking duration risk bets accrue to ALCO, not deposit branches. Moreover, the determination of a transfer rate in no way should be influenced by the manner in which funds are actually deployed.

With equity tied up in loans, FTP is our vehicle to properly attribute mismatch earnings. Funding Center should credit deposits at the short rate and charge the loans according to the 10-year alternative opportunity rate for funds. Embedded in this charge is the cost to relieve FC of repricing risk, which it may do by selling the fixed coupon stream into a 10-year swap for floating coupon income. Having hedged the loan's mismatch, deposit gatherers earn the short rate and equity earns a floating return. If Funding Center is left unhedged and market rates skyrocket, the duration problem is squarely in ALCO's corner, not the branches. In the event liquidity break down to the point branches have to pay to retain deposits, again, ALCO bears responsibility for financial consequences in the margin.

By definition, the short-term transfer rate method ignores any potential funding contribution value forecasted beyond the next repricing date. Recall the approach for transfer-pricing prime loans: so long as Funding Center is held accountable for managing the risk, then it should be paid for assuming the exposure. In the case of IMDs, branches may or may not set the offering rates. However, there is wide acceptance of the notion that deposit gatherers reap the accounting rewards for raising sub-market liabilities, not Funding Center. If Funding Center gets no rewards, the unit should assume no risk. Long-term transfer rates are schemes that force FC to assume the risk of forward rates diverging from that priced in the benchmark fixed rate. Should rates rise slower than the forwards indicate, or even not at all, Funding Center is crediting for contribution value that will not be realized. On the other hand, rates rising faster than the implied forwards translate into insufficient value being recognized for holding savings balances: the branches are short-changed.

SUMMARY

When organizations are in the initial stages of establishing an FTP process, it is not uncommon to observe lending units laying claim on the cheapest funds in the balance sheet. Not to be outmaneuvered, depositgatherers assert a good portion of margin creation is owing to the institution's stable, low-cost funding sources. Moreover, a portion of net interest income arises from the assumption of repricing mismatches, a risk/reward responsibility generally attributed to asset-liability management. A prime goal to FTP is accounting for mismatch contribution earnings, but attaining this objective can be mired by debate over assigning credit rates to savings and checking deposits.

Proper integration of strategic mismatch risk with performance measurement of line units starts with sound transfer pricing methodologies. An investigation of the economic underpinnings to indeterminate-maturity deposits suggests the appropriate replicating portfolio bears the short rate of interest and permits variable balance principal. Average rate and blended term transfer pricing methods do not clearly isolate the repricing basis, while mischaracterizing the depositor's decision to hold balances. Adherence to true four-factor transfer pricing provides improved understanding of the period-to-period value in this funding source; at the same time, distortions in the Funding Center are minimized. The result is a transfer-pricing approach delivering objective performance evaluation, transparent accounting of mismatch margin contribution and pricing signals that will motivate more informed decision-making.