Extracting More Value from Funds Transfer Pricing · Since transfer pricing systems tend to be...

Extracting More Value from Funds Transfer Pricing 1a company

Extracting More Valuefrom Funds Transfer Pricingby Kenneth M. Levey

Extracting More Value from Funds Transfer Pricing2

Funds Transfer Pricing (FTP) has been an important management accounting tool within the banking industry for several decades. FTP provides insight by dissecting net interest margin (NIM) in a manner the general ledger can’t. General ledger balances at the entity level can be structured to give a clear picture of margin on an aggregated basis. But general ledgers are unable to provide a disaggregated or more granular view of NIM. What if a retailer could only measure gross margin at a consolidated level with no visibility to stores, product lines etc.? Such is the situation with banks and credit unions apart from transfer pricing. A well-structured funds transfer pricing mechanism provides margin clarity at every level of the organization, even down to an individual customer. In so doing it enables management to:

• Understand the profit contribution of its balance sheet products such as loans and deposits more accurately

• Understand the profit contribution of a channel, an officer, an organization or individual customer

• Create internal accountabilities and associated targets for funds users (credit risk takers), funds providers (core deposit gatherers) and funding managers (interest rate risk takers)

• Price interest-earning products more knowledgeably based on market dynamics and target contribution levels.

• Reward loan officers, branch managers, and regional leaders based on margin contribution vs. mere balance growth or number of new accounts opened

These immediate benefits of FTP can be further enriched when combined with other processes such as service transfer pricing and economic capital allocation, resulting in a complete view of the risk-adjusted profit contribution of every customer relationship. But even as a stand-alone analytical tool, FTP can and does provide great benefits.

The Big PictureImagine a bank that has just two customer transactions originated on the same day - a three-year loan at 4.55% and a six-month CD at 1.15%. The combination of the two transactions results in a net interest margin for the bank of 3.40%. At first blush, it appears that the loan generates 100% of the revenue, and the deposit 100% of the expense. One might come to the conclusion “loans are great and deposits a necessary evil” – now insert FTP. By adding a funds transfer

mechanism, specifically an appropriate funds transfer pricing yield curve, a completely different picture as illustrated in the following graphic emerges.

The inclusion of a transfer pricing curve radically changes our understanding of the makeup of net interest margin and how the bank’s activities have contributed to it. For taking credit risk in the form of a three-year fixed-rate loan, the institution earns a spread of 1.25% above the match-funded point on the curve. For sourcing funds in the local market in the form of a six-month CD, the bank earns a spread of .95%. Essentially the deposit-gathering function is being rewarded for the realized cost savings compared to sourcing the same funding in the wholesale market. Finally, for taking interest rate risk in the form of a mismatched position, (in this case lending long and funding short) the institution earns 1.20%. In this example, the spread earned by taking a mismatched position is positive due to the upward sloping nature of the yield curve. As the yield curve changes shape through time, the contribution from being mismatched is uncertain. Net interest income derived from taking a mismatched position can easily go negative depending upon a variety of factors. It is the bank’s ALCO function that must take the responsibility for managing the portion of net interest income that comes from taking interest rate risk. From this simple example, it is clear how funds transfer pricing yields a completely different picture than the general ledger tally of interest income and expense. Perhaps more importantly, from this process clear new accountabilities for margin can be assigned to various functions inside the institution.

Three Year Loan 4.55%

Transfer Rate 3.30%Transfer Rate 3.30%

Spread of 1.25%

Transfer Rate of 2.10%Transfer Rate of 2.10%

Spread of .95%Spread of .95%

6 Month CD 1.15%

Spread Due toFunding Mismatch

1.20%

Spread Due toFunding Mismatch

1.20%

Net InterestMargin of

3.40%

Net InterestMargin of

3.40%

Extracting More Value from Funds Transfer Pricing 3

The above illustration is simplistic—in the real world, banks and credit unions have potentially millions of transactions in various shapes and sizes created over a long period of time. As such, transfer pricing systems are designed to price each customer record based on the expected cash flows at the time of origination, utilizing the interest rate environment in play at that time. For example, to price the entire balance sheet the institution needs to have all the historic term structures for their selected transfer pricing yield curve going back as far as their oldest asset or liability would require.

The Mechanics One of the first steps the institution must take in implementing transfer pricing is defining their transfer pricing curve. The curve should represent the institution’s own ability to source funds of various terms on the wholesale market. The LIBOR/Swap Curve is widely used as it represents the credit worthiness of non-governmental entities approximating an A+ rating, which is a good proxy for a healthy bank. Clearly each institution should adjust their transfer pricing curve to reflect their ability to source funds in the market.

The process of deriving an accurate theoretical cost of funds for a loan or an earnings credit for a deposit has evolved over the years. Methodologies have become more complex over time and the level of refinement continues to evolve. In generalized terms, well-designed FTP systems will transfer price each customer record based on the cash-flow characteristics of the record (for example,. fixed or floating rate, term, cash flows, etc.) at the time of origination or the last reprice date. This is typically called a “match-funded” approach. Unlike the previous illustration where 100% of the loan’s principal comes due at maturity, many loans amortize over their life, paying back principal all along the way. FTP systems can compute the duration of such loans and match fund on a duration basis, but most systems today assign a match-funded transfer rate to each principal payment and then amalgamate these individual rates into a single rate based on a time/balance weighted algorithm.

The following example, shows a twelve-month amortizing loan with a rate of 4.50%. The graph on the right depicts how the outstanding balance on the loan declines through time due periodic payments.


Below, we can see each principal payment represents a partial retirement of the overall loan and is transfer priced using the appropriate maturity point on the curve. Since some payments happen early and others later, a weighting factor is applied to each principal payment to accurately take into account the length of time each portion of principal is outstanding. From this process an overall composite transfer rate for the entire loan is calculated – in this case 2.33%.

In this example the resulting spread is 2.17% (4.50% - 2.33%). One could say that 217 basis points is needed to cover all the associated costs of originating and servicing this loan as well as the target profit.

A number of refinements can be made to this process. For example, how should the institution account for unscheduled principal payments (“prepayments”)? Most transfer pricing systems support the ability to enter a prepayment assumption about a particular tranche of loans. As contractual cash flows are projected, additional principal reduction is applied which accelerates the payoff the loan, thereby shortening the life of the outstanding principal. Each cash flow strip (contractual + prepayment) is transfer priced. An option-adjusted spread (OAS) may be calculated and added to the transfer rate to account for the optionality of the loan.

There are other challenges as one seeks to transfer price the entire balance sheet. For example, some accounts have no defined cash flow structure - accounts such as money market savings. Over the years different approaches to handling these ambiguous accounts have evolved. Some have focused on distinguishing core and volatile portions of these accounts as a means to determining their value. Some have used more advanced statistical methods, studying the behavior of these accounts in different rate cycles to ascertain a mix of funding that best correlates with the account’s historic behavior. A broad body of work is available to enable the institution to become as precise as it desires in determining the theoretical matched rate for all account types on the balance sheet.


Extracting Further Value from Transfer Pricing Through AnalyticsPursuing higher levels of precision and accuracy in one’s method of transfer pricing is generally a worthy goal, but at some point it yields diminishing returns. In fact for many institutions their current methodology is sound enough. Investing further time and energy refining their approach creates relatively modest incremental value provided the institution has implemented some form of matched rate transfer pricing. Making the shift from how transfer pricing is done to how transfer pricing is used can often yield greater benefit.

Since transfer pricing systems tend to be transactional in nature, so do their reports. It’s not uncommon for FTP systems to produce lines and lines of detail which can be quite difficult to extract any meaning from. The report sample below illustrates the point. This particular query to an FTP system resulted in 1024 rows of loan detail. Although the information presented is accurate, it isn’t particularly useful except in the case where someone is researching information on an individual loan.

It is difficult to draw any conclusions from the data using the above example. If there are relationships between any of the variables presented, it isn’t decipherable. This sample institution has been interested in moving towards a more risk-adjusted approach to pricing, but from this report it is difficult to tell if new guidelines are being followed.


Using the same set of loan records, we quickly learn some key information about the portfolio: its size, weighted average rate, spread and credit score. The weighted average spread of 2.97% is strong for the portfolio but the red scatterplot trend line highlights a lower correlation between the bank’s pricing and the borrower’s credit rating. The scatterplot of detail provides a visual of the relative relationship between FTP margin and credit score. While the trend line of the data is modestly downward sloping (which one would expect), it is flatter than the risk return profile the bank expects. Recent policy guidelines stipulate a graduated scale for margin targets given the underlying credit risk as measured by the credit score. The guideline makes it acceptable to originate loans with spreads anywhere from 1.0% to 5.00% based on the credit score.

The following analytic is a simple example of how better information can be extracted from the existing data.


Superimposing the bank’s guidelines over the portfolio profile, one can see that numerous loans are priced well below the desired threshold. This is especially true for riskier loans with lower credit scores. Bank wants to attract higher quality borrowers, this brief analysis indicates they need to enforce stricter guidelines especially in regard to pricing lower quality credits. This analysis suggests that they should be more aggressive in attracting higher quality borrowers than they are today. In summary, the use of weighted average totals combined with relatively straightforward data visualizations results in some meaningful insights buried in the original flat list of loan records.

An important analytical tool that is often not fully utilized in transfer pricing is incremental analysis. While there is value in understanding the overall contribution of a portfolio, one quickly realizes the information is not particularly actionable. A portfolio’s overall balance and spread after transfer pricing is the aggregation of all historical lending decisions summed together into one number. Nothing can be done about historic lending decisions that were made in some cases many years prior. Incremental analysis allows the institution to focus on recent activity which is more pertinent. It helps answer the question “What’s happening now?” Focusing on current activity is more actionable as it enables the institution to course correct if necessary. It also provides a

good indicator as to whether overall spreads for the portfolio will directionally widen or narrow based on how loans and deposits are being priced today.

The following example scenario illustrates this point. Meadowbrook Market is a large urban market bank with many large corporate credits. Senior management gets a report that displays the balance of the commercial portfolio as $2.8B as of 3/31. Included in the report is some transfer pricing information showing periodic interest and spread information for the portfolio which is 138 basis points. While management is pleased that both the balance and overall spread has increased from the prior quarter, there is little to draw from the information being presented – certainly nothing actionable. However, if Meadowbrook were to source the very same record-level data and segment it by officer and origination date, as well as add a few simple calculations to estimate incremental revenue and rankings, they would have a far richer picture of what’s happening with the portfolio as shown below.

While total portfolio information is still present, segmentation provides a much clearer picture regarding last quarter’s activity—including new visibility into the contribution being made by loan officers. Officer WIMK00 led the team with the highest volume of new originations, but he was thirteenth in


terms of the spread—the combination of which resulted in a #2 ranking for incremental QTD NII contribution. Further detailed or drill-down analysis (not shown) gives an indication as to whether Officer WIMK00 has overly sacrificed spread for the sake of volume. Officer SNSD00 is new to the bank, and although she is building up her pipeline it is good to see the high margin she was able to generate on the few loans she closed during the quarter. The report also highlights that the Meadowbrook Market has been able to originate and/ or re-negotiate terms on existing credits at 64 basis points over the portfolio average. This is a good indicator that the overall contribution from this market should continue to improve. Management has a much clearer picture of what is happening with the portfolio, and from this analytical view they are armed with good information to help coach and manage their loan officers in a way that’s impactful to the bank’s performance moving forward.

Beyond reporting, incremental analysis as highlighted above can be used for other purposes such as incentive compensation. Too often incentive plans are not rightly aligned with the institution’s processes for measuring contribution. Using the information above, the bank could

devise an incentive plan that rewards loan officers for the incremental NII revenue they generate after transfer pricing. This approach harmonizes management reporting with the institution’s incentive plans. Creating the linkage between “how we measure” and “how we reward” ensures that the bank is shaping the behavior of its personnel in alignment with its overall corporate objectives. Some banks have extended this approach to add economic capital and other factors to determine a risk-adjusted return on capital (RAROC) for a period’s activities. This approach takes into consideration not only the volume and spread, but the underlying risk when calculating contribution and the associated incentive compensation plan. Banks and credit unions should look for holistic performance management systems that can bring all these pieces together.

Once the institution has organized its data in a manner that easily supports incremental analysis, it will discover a host of analytic views that give management a clearer picture of what’s happening in the organization. With a few quick changes the previous loan officer report can be altered to view product contribution.


From this report we learn that the Commercial and Mortgage business is the most active, but as of late it is contributing relatively thin margins. We also see the downward pressure on Auto and Consumer Secured margins when compared to the portfolio average, which we subsequently learn is due to increased competition in the local market. Incremental analysis is becoming real time for some institutions. Rather than looking at last quarter, last month, or even last week, some institutions are enabling their customer-facing personnel to model the impact of a proposed new deposit or loan relationship in real time as the decision is being made. This instant feedback enables the loan officer or branch manager to fine tune the financial terms in order to meet predetermined hurdle rates. In the example below, a loan officer is prompted to key in a number of attributes of a proposed new loan. The system then provides immediate feedback in the form of the first year’s P&L and RAROC, including a hurdle rate comparison.

The example above is high-value add process, and is an important extension to the typical funds transfer pricing process that has traditionally focused solely on booked loans vs. potential loans. One of the goals of transfer pricing is to bring more discipline into how loans and deposits are being priced. By deploying FTP in real time, the organization has equipped its customerfacing teams with a powerful tool that helps ensure the bank is making sound individual pricing decisions each and every day. One might say this is the height of incremental analysis as it focuses on each decision as it happens.


The technology implication behind this example assumes a real time FTP process, and perhaps a cost allocation process that is performing calculations based on inputs and today’s transfer pricing curve. Real-time product pricing often takes into account the entire customer relationship, such that the individual(s) pricing the opportunity can view the existing balances and their contribution along with the proposed transaction with a before and after picture.

Another analytic tool that organizations can deploy to extend the value of their transfer pricing function is a rate volume analysis. Rate volume reporting has been a staple of financial institutions for many years. It helps disaggregate net interest income variance into two causal factors: 1) variance due to changes in rates, and 2) variance due to changes in balances (or volumes).

There are a few variations of this report that can be created inclusive of transfer pricing. One variation replaces interest income with spread income (interest income – transfer expense). With this replacement, one can further explain the change in spread income due to changes in the customer rate and changes in the transfer rate since spread is a function of both. Another variation to the rate volume analytic report is to replace interest income directly with transfer income and expense. This creates a view that is

particularly interesting to the funding center. As most know, the funding center is the location where all offsets for FTP charges and credits reside. After funding all asset generators and crediting all liability generators, the funding center has a net P&L which essentially captures the net interest income/expense position from taking interest rate risk (for an example, refer to the first diagram of this whitepaper). Many have struggled to interpret the funding center’s results. A rate volume report can help explain why the P&L in the funding center is changing over time. Below is an abbreviated sample of this type of analysis.

In this example, the funding center’s net interest income dropped month over month. The rate the funding center received from asset generators dropped by 2 basis points, while the rate the funding center paid liability generators increased by 3 basis points, thereby resulting in a compressed level of income for taking interest rate risk. As mentioned earlier, changes in the shape of the transfer pricing curve will impact the portion of net interest income that is due to mismatched position on the balance sheet. Further investigation in this example revealed an increase in short-term floating rate assets and an increase in longer term CDs. These changes essentially reduced the funding center’s income in November when compared to October.


About AuthorKenneth M. LeveyVice President, Financial Institutions, Axiom EPM

Ken Levey is Vice President of Financial Institutions at Axiom EPM. In this capacity, Ken has responsibility for product strategy and design for banking solutions that reside on the Axiom EPM platform. Additionally, Ken leads Axiom’s efforts to foster and incorporate Customers’ needs into the platform’s Banking solution set. Prior to joining Axiom EPM, Ken was the Senior Director for Banking Analytics at SAP BusinessObjects where he was responsible for developing performance management and analytic applications for the Banking Industry. Ken started his career at Bear Stearns in New York and then Seamen’s Bank for Savings, before moving to a senior position at IPS-Sendero (Fiserv) where he led the development of the Asset/Liability Management system and managed the ALM and Funds Transfer Pricing implementation and consulting departments.

About Axiom EPMAxiom EPM provides sophisticated, flexible performance management solutions for banks and credit unions that empower finance professionals to analyze results, model the future and optimize organizational decision making. Solutions for budgeting & forecasting, strategy management, incentive compensation management, profitability management & FTP, reporting & analytics, and loan credit quality analysis are delivered on a single unified platform. Axiom EPM embraces and extends Microsoft Excel® functionality, allowing finance professionals to manage data in a familiar environment – while providing unmatched modeling flexibility and enterprise performance. Axiom EPM is a wholly-owned subsidiary of Kaufman Hall and Associates.

Conclusion Funds transfer pricing has been an important tool for financial institutions for several decades. Over the years there has been considerable refinement with regard to how one should transfer price various balance sheet accounts like mortgage loans and money market savings. Despite the refinement to methodology, many organizations have under-leveraged the analytical value that funds transfer content can yield. Through the use of data visualization, incremental analysis, segmentation, and other analytical methods, management can extract considerably more value from their transfer pricing investment including real-time product pricing assistance and more effective incentive compensation programs. The focus of Axiom EPM for banks is to ensure that these processes come together holistically so that the institution can maximize its return.

Extracting More Value from Funds Transfer Pricing · Since transfer pricing systems tend to be...

Documents

Transcript of Extracting More Value from Funds Transfer Pricing · Since transfer pricing systems tend to be...