Journal of Bank Cost & Management Accounting, The: AMIfs Bank Cost Analysis Model (BCAM) implementing activity-based cost management in a financial institution

FindArticles > Journal of Bank Cost & Management Accounting, The > 2001 > Article > Print friendly

AMIfs Bank Cost Analysis Model (BCAM) implementing activity-based cost management in a financial institution

Smith, J Timothy

ABSTRACT

Over the past several years, much has been written about the application of Activity-Based Costing/Management (ABC/M) in both manufacturing and service business sectors. This paper is intended to specifically address the implementation of ABC/M in financial institutions, primarily banks. This paper provides guidelines for creating and maintaining an ABC/M solution in these environments.

In addition, AMIfs is interested in providing benchmark information to members if there is an interest and if members can reach a consensus on a method of cost development.

One of the objectives of this article, besides providing information on Activity-Based Costing, is for members to review the detail functions, activities, drivers, etc., to determine if they provide the right level of consensus amongst member banks and the basis for understanding of costs through benchmarking.

From the answers to the above question, AMIfs will determine whether it is feasible to begin a cost benchmarking exercise.

This set of guidelines is divided into the following sections:

1. Introduction - Basic information on ABC/M and uses of ABC/M to solve business problems.

2. The Case for ABC/M in Banks - What created interest in ABC/M and why should it be applied to banks?

3. The AMIfs Bank Cost Analysis Model (SCAM) - Definition of banking activities, cost drivers, and cost objects designed to cover the traditional banking costs of a community or regional bank.

4. BCAM Implementation - Basic process for the implementation of BCAM in a bank.

5. Implementing ABC/M Is More Feasible Once the Misconceptions Are Disproved - Tips for success for implementing ABC/M.

1. INTRODUCTION

ABC/M implementation is frequently an intimidating proposition, potentially in conflict with an organization's current management accounting practices. The implementation effort can be resource-intensive, fraught with data challenges, and subject to frustrating delays. A poorly or slowly executed ABC/M study can result in loss of credibility, cost factors becoming obsolete before they are finalized, and no better management decision-making ability. So why bother to take on what appears to be an awesome task with high risks and potential failure?

Simply stated, the rewards are great. Properly implemented ABC/M systems in banks provide the cornerstone for product pricing, product/ customer/organization profitability measurement, staffing models, reengineering efforts, aligning workload with the organization's strategies, and cost reduction.

Because the management information resulting from ABC/M systems can be so rich, it is imperative at the outset to determine the primary business uses) for the data prior to starting any ABC/M study. That is, begin by working backwards with the end in mind. Strong executive sponsorship is also essential, accompanied by a clear definition of the business problem to be addressed through an ABC/M implementation. The level of detail employed in an ABC/M implementation will be directly affected by the ultimate use of the data. Ideally, it is advisable to start with a more summarized ABC/M structure, and then later selectively build in more detail in places that need greater visibility. ABC/M is scalable. See the section on leveling and right-sizing an ABC/M system in the introduction of BCAM (Bank Cost Analysis Model).

How Does Traditional Accounting Distort Reality?

The ABC/M logical assignment of expenses and costs obliterates the use of simple averages as the basis for tracing costs. Figure 1 illustrates the impact of applying ABC/M versus the traditional allocation of resource expenses into costs. The diagram reveals ABC/M's "S-curve." The horizontal line represents the flawed costs that are calculated by a traditional standard costing system that forms the organization's belief system.

In practice, one discovers the under-costed products are substantially under-costed because these products may be low-volume with small batch sizes, require more technical attention, consume more or special handling, or need extra inspection. ABC/M removes the distortions from simplistic cost allocations.

ABC/M corrects for these flaws by identifying the work activities that are responsible for costs. ABC/M provides a cost flow assignment network, which allows the work activity costs and their output costs to be continuously re-assigned, or passed on only if the products, services, or customers, (or in some cases, other work activities) actually use the activity. This condition of consumption and use is what sets ABC/M apart from traditional cost allocation schemes. See also section 2 of BCAM regarding the case for ABC/M in Banks.

From General Ledger to ABC Database

How can traditional accounting that has been around for so many years all of a sudden become considered so bad? The answer is the existing data is not necessarily bad as much as it is somewhat distorted, incomplete and unprocessed. The left side of Figure 2 shows the classic responsibility-center statement report that managers receive each month. When you translate those "chart-of-accounts" expenses on the left into the "chart of activities" on the right that consume the financial general ledger's expenses, a manager's insights from viewing the activity costs begin to increase. The right side of Figure 2 is the ABC/M view that is used for analysis and as the starting point for calculating the costs for both processes and for diverse outputs such as products and customers. In effect, the right side ABC/M view resolves the deficiencies of traditional financial accounting by focusing on work activities. ABC/M is very work-centric, whereas the General Ledger is transaction-centric.

Much more additional information can be gleaned from the right side ABC/M view. Look at the second activity, "process branch transactions" for $121,000 and ask what would make that cost significantly increase or decrease. The answer is the "number of branch transactions processed" That answer represents that work's activity driver. Figure 2 illustrates that each activity on a stand-alone basis has its own unique activity driver. At this stage, the costing is no longer recognizing the organization chart and its artificial boundaries. The focus is now on the work that the organization performs and what impacts the level of that workload.

ABC/M is not a replacement for the traditional general ledger accounting. It is a translator or overlay that lies between the ledger account and cost center accumulators in the general ledger and the end-users, like managers and analysts, who apply cost data in decision making. ABC/M translates expenses into a language that people can understand. ABC/M translates expenses into elements of costs, namely the work activities, which can be more flexibly linked or assigned to business processes or cost objects based on demand-driven consumption patterns, not simplistic cost allocations.

Defining Costs, Drivers, and Objects

Part of the problem defining and designing costing systems involves understanding just what exactly costs are. Costs are not cash expended in the sense you see them categorized in the general ledger. Costs themselves are abstract and intangible. One cannot see costs or hold a couple of them in one's hands. Yet we all know they are there. We know they exist whether we measure them or not.

We know that costs increase or decrease as there are changes in the workload that affect the activity costs via their cost drivers. Work activities are triggered by events, and the costs react as the effect. Costs measure effects more so than they illuminate root causes. Costs are the collective effects of work activity triggered by various demands on work. ABC/M links these effects to their usage by varying types of customers, products, service-lines, channels, or any other user object. ABC/M systems provide an enterprise-wide image of all the collective effects plus the causal relationships that result in an organization's costs. So costs give insights to root causes, but mainly through their inferences. This may sound ironic, but "cost management" is a contradictory phrase. You do not really manage costs and financial results. You understand the causes (or drivers) of costs. Then you manage the causes. In ABC/M, driver management accomplishes cost management.

A cost driver is something that can be described in words but not necessarily in numbers. For example, a storm would be a cost driver that results in much clean-up work and the resulting costs. In contrast, the activity drivers in ABC/M's cost assignments must be quantitative, using measures that apportion the costs among various products and customers. In the ABC/M vertical cost assignment view, there are two types of drivers, and both are required to be quantitative:

* Resource drivers- trace expenditures (cash outlays) to work activities

* Activity drivers - trace work activity costs to cost objects

Activity drivers will have their own higher order cost drivers. Events or other influences that are formally called cost drivers, prompt work activities. A cost driver, such as processing a new bank customer, is the trigger that causes the work activity to utilize resources to produce output or results.

Cost drivers and activity drivers serve different purposes. Activity drivers are output measures that reflect the usage of each work activity, and they must be quantitatively measurable. An activity driver, which relates a work activity to its cost objects, "meters-our" the work activity based on the unique diversity and variation of the cost objects that are consuming the activity.

Cost objects are the persons or things that benefit from incurring activity costs; examples are products, internal or external customers, stakeholders and outputs of internal processes. Cost objects can be thought of as: to what and for whom work is done.

For example, Bank X can discern through management accounting that their consumer loan portfolio is not meeting a corporate ROE hurdle rate of 18%. Product pricing is competitive and largely market driven. Costs appear to be the issue, but there is no understanding of the dynamics driving the costs.

Since Bank X has several consumer loan types (e.g., auto, home equity, credit lines), the first question is "What is the profitability of each product?" If one is simply interested in the total cost by product, as a first cut in the analysis, then process-based costing may be appropriate. In other words, define and measure processes at a high level (such as Make Auto Loan) rather than lower level work activities (e.g., interviewing customers, taking application, reviewing application, approving application, preparing loan docs, funding the loan, etc). In this case, all of the process might be linked to a single cost object: an Auto Loan Made.

Once the process cost has been determined, management may have identified a new problem: The cost must be reduced in order to price this product competitively and still earn a profit. In order to further analyze the product, the work activities contributing to each type of product will need definition and analysis. The cost drivers for each activity will require definition and data collection. How many applications are turned down before a loan is made to a customer? How many times is an application shuttled back and forth for approvals? How many times is the customer contacted for additional information? These are all potential drivers that could affect the overall product cost plus reveal substantial cost differences among products that people believe have always been comparable in cost and profit. However, to initially pursue this level of information for all products and service-lines without first understanding high level product costs might be overkill. The initial ABC/M structure and system will likely be way over-designed and engineered beyond diminishing returns for extra levels of administrative effort. These delays can risk losing the all-important buy-in from managers and employees.

In another example demonstrating the evolution of cost data, in early cost studies, Bank Y was satisfied knowing the average cost of their five types of consumer checking accounts. As they realized the wide range of transaction activity among clients, they felt it appropriate to calculate the cost of checking deposits and withdrawals. As ATM and phone delivery channels became more popular, this bank wanted to understand the relative costs of transactions through different delivery channels and be able to compare the cost of a teller withdrawal to an ATM withdrawal. This led to the ability to calculate customer profitability incorporating delivery channel usage. The results were substantially different than what the managers and employees believed. The message here is that ABC/M does not automatically fix the problem, but it stimulates discovery and discussion with fact-based data.

In summary, it is wise to carefully plan an ABC/M implementation to address the current information needs of bank management, while considering where the organization is in its evolution of costing data. It is not an "all or nothing" proposition. The design should embody a "top down" approach to initially enhance the understanding of overall cost structure of the organization and then focus efforts on the greatest needs of management. ABC/M may be implemented at varying levels of detail throughout an organization. It is an iterative process, with subsequent cost studies "peeling back additional layers of the onion" in order to better understand the cost of activities, processes and products. This paper provides banks with guidelines and information to implement ABC/M at whatever level is appropriate to the organization. More importantly, it also provides a model that encompasses this"top down" approach and provides the basis on which to expand and build not only greater levels of detailed analysis but also potentially a consistency in operational benchmarking never achieved in our industry.

2) THE CASE FOR ACTIVITY-BASED COST MANAGEMENT (ABC/M) IN BANKS

In the early 1980's many organizations began to realize their traditional accounting systems were generating inaccurate costing information. The typical organization's cost structure had been substantially changing. For most of them, overhead and indirect costs were increasing and mainly displacing the direct labor costs - the costs of front-line workers. For banks, higher and higher levels of cost are shared across products, customers, channels, etc. The three primary causes for this shift were: (1) increasing organizational complexity resulting from proliferation in the variety of product and service offerings, (2) a more diverse group of channels and customers, and (3) increased automation, new technologies, and new methodologies.

Clearly, this situation exists in banks. Today the margin for error is even slimmer. Banks cannot make as many mistakes as they could in the past and remain competitive or effective. Capital investment decisions, product mix selection, technology choices, outsourcing, and make vs. buy decisions today all require a sharper pencil. More competitors are better understanding the cause-and-effect connections that drive their costs, and they are fine-tuning their business processes and adjusting their prices accordingly. The resulting price squeeze from more intense competition is making life for banks much more difficult relative to the past. Knowing what your real costs are for outputs, product costs, and the costs-to-serve different types of channels and customers is becoming key to survival.

With ABC/M visibility, organizations can identify where to remove waste, low-value adding costs, and unused capacity plus understand what drives their costs. They can also see the degree of alignment (perhaps better thought of as misalignment) of their cost structure with their organization's mission and strategy.

Forces That Caused the Need for ABC/M

ABC/M was developed as a practical solution for problems associated with traditional cost management systems that we now realize are distorting and incomplete. Non-interest expense and overhead cost allocation practices of traditional systems can bring more damage than good to organizations. The problem with overhead cost allocations is where excessively broad-brushed average cost rates are applied to calculate the costs. Cost allocations often rely on factors such as head count or department expenses. Some factors are based on volume or balances; however, these metrics may be entirely independent of the level of effort or time duration. Correcting an error may take the same amount of time for a $300 account as it does for a $50,000 account.

The use of volume-based allocations will provide reasonably accurate calculated costs when the following conditions exist:

* Few and very similar products and service-lines

* Low overhead expenses

* Homogeneous conversion processes

* Homogeneous channels, customer demands, and customers

* Low level of expenses shared across product or customer lines

* Very high margins

How many banks possess those characteristics today? Hardly any.

The result of inaccurate cost allocations, because allocating is a zerosum error game, is that some cost objects get severely over-costed while the remainder are under-costed. In other words, as a consequence of unquestioned formula cost allocations, traditional financial accounting can grotesquely distort the true cost of products and service-lines, which in turn can wildly distort their individual profit margins. Many managers understood intuitively that their outdated accounting system was distorting the product and service-line costs, so they sometimes made informal adjustments to compensate. However, with so much complexity and broad product and channel diversity, it was nearly impossible for managers to predict the magnitude and impact of their mental adjustments needed to achieve accuracy.

A bank manager has said, "You know what we think about our cost accounting system here? It is a bunch of fictitious lies - but we all agree to it." This is a sad resignation by people to accept inaccurate information at a time when they need reliable and true managerial accounting data to make better decisions.

Some accountants are intimidated by what they perceive might be a mudslide of data and an onerous task to build a better cost assignment system. But at some point, the accountants must realize that their role is to collect, validate, and report the data. They are not the users. Yes, they might produce various ratios and reports for the executives, but the main users are managers and employees who are all striving to make decisions to improve the organization's performance.

We discussed how traditional accounting might distort reality in the Introduction. A key difference between ABC/M relative to the General Ledger and a traditional technique of cost allocation (i.e., absorption costing) is that ABC/M describes activities using an "action-verb-adjective-noun" grammar convention, such as "render statements" or "open transaction accounts" or "process branch transactions." This gives ABC/M its flexibility. Such wording is powerful because managers and employee teams can better relate to these phrases, and the wording implies that the work activities can be favorably impacted, changed, improved, or eliminated. The General Ledger uses a chart-of-accounts whereas ABC/M uses a chart-of-activities as its language. In translating General Ledger data to activities and processes, ABC/M preserves the total reported revenues and costs, but allows the revenues, budgeted funding and the costs to be viewed differently.

Connecting Bank Customers to Resources

Figure 2 (Introduction) illustrates that each activity on a stand-alone basis has its own unique activity driver. At this stage the costing is no longer recognizing the organization chart and its artificial boundaries. The focus is now on the work that the organization performs and what impacts the level of that workload.

Let's now assume there were 100,000 branch transactions processed during that period for the department in Figure 2. Then the unit cost per each "branch transaction processed" is $1.21 per transaction. If a specific group of senior citizens over the age of sixty were responsible for half those branch transactions, then we would know more about a specific customer or beneficiary of that work. The senior citizens would have caused $60,500 of that work (i.e., 50,000 branch transactions at $1.21 per transaction). If married couples with small children required another fraction, married couples with grown children a different fraction, and so on, then ABC/M will have traced all of the $121,000. If each of the other work activities were similarly traced using the unique activity driver for each activity, then ABC/M will have piled up the entire $914,500 into each segment of customers. This re-assignment of the resource expenses will be much more accurate than any broad-brushed and volume-based cost allocation traditionally applied with traditional costing procedures and their broad averages. In summary, the General Ledger view on the left describes "What was spent" whereas the activity-based view on the right describes "What it was spent for."

ABC/M is a translator, not a replacement for the General Ledger. ABC/M translates expenses into a language that people can understand. ABC/M translates expenses into elements of costs, namely, the work activities, which can be more flexibly linked or assigned to business processes or cost objects based on demand-driven consumption patterns, not simplistic cost allocations.

Figure 3 uses the analogy of an optical lens to show how ABC/M serves as a translator of general ledger data to provide more focused data for improved decision support. The lens not only translates the ledger costs into a more useful and flexible format, but the ABC/M lens provides more sensory information. The reason ABC/M is becoming popular is because the General Ledger is now recognized as being structurally deficient for delivering good business information for decision support - the General Ledger is a sound mechanism for collecting and accumulating transaction-intensive costs, but not for converting those costs into useful managerial information.

Costs, Drivers, and Objects

To repeat the explanation and definitions from the Introduction, a cost driver is something that can be described in words but not necessarily in numbers. In contrast, the activity drivers in ABC/M's cost assignments must be quantitative, using measures that apportion the costs among various products and customers. In the ABC/M vertical cost assignment view, there are two types of drivers, and both are required to be quantitative:

* Resource drivers - trace expenditures (cash outlays) to work activities

* Activity drivers - trace work activity costs to cost objects

* Activity drivers will have their own higher order cost drivers.

* A cost driver is the trigger that causes the work activity to utilize resources to produce output or results.

* Cost objects are the persons or things that benefit from incurring activity costs.

The ABC/M Structure: Two Views

With ABC/M, the General Ledger account balances are first converted into activity costs using resource drivers. Then ABC/M: (1) assigns the activity costs to cost objects, or (2) re-assembles the activity costs across business processes. This new and transformed ABC/M cost data can be used to identify operating relationships that can be used effectively in making product, channel, market, and customer-oriented decisions. This ABC/M information can also be useful in managing processes and any quality-related issues within the processes. In all cases, BCAM transforms the General Ledger data into a different type of cost data that is more useful for decision-making. Figure 4 illustrates the ABC/M structure.

The ABC/M Cross reveals that work activities, which are located in the center intersection of the Cross, are integral to reporting both the costs of processes as well as the costs of cost objects.

Figure 5 reveals the types of questions that the ABC/M structure answers.

The vertical cost assignment view (i.e., traditional ABC/M) direction explains what specific things cost, whereas the horizontal process view, which some refer to as activity-based management (or ABM for ABC/M without the "C"), explains what causes costs to exist and to fluctuate. Let us be clear about how these two views apply the same activity cost data, but orient the activity costs differently.

The Cost Assignment View

The vertical axis reflects costs as they are sensitive to demands from all forms of product, channel, and customer diversity and variety. The work activities consume the resources, and the products and customer services consume the work activities. Figure 6 illustrates this cost re-assignment view.

The ABC/M cost assignment view is a cost consumption chain. When each activity cost is traced based on its unique quantity or proportion of its driver, all of the resource expenses are eventually re-aggregated into the final cost objects. Activity drivers are critical to ABC/M because not only do they segment and flow the costs to reflect the diversity of the products and customers, but they govern the accuracies as well.

The Process Assignment View

The horizontal view of the ABC/M Cross represents the business process view. A business process can be defined as two or more activities or a network of work activities with a common purpose - usually customer-focused. Activities belong to the business processes. Therefore, the activity costs comprise the costs of the business processes. Figure 7 illustrates the process view.

Across each process, the activity costs are sequential and additive. In this orientation, activity costs satisfy the requirements for popular flowcharting and process modeling techniques and software. Business process-based thinking, which can be visualized as tipping the organization chart 90 degrees, is now dominating managerial thinking. ABC/M provides the cost elements for process costing that are not available from the General Ledger. A future AMI/s project will focus on the business process view of ABC/M.

The ABC/M Cross displays in a simple fashion that the work activities at the intersection of the vertical and horizontal axes are integral to determining the cost of an organization's processes as well as the cost of its cost objects. The work activity at the intersection schematically represents an individual work activity - a very local view. But, from a global perspective, the vertical (cost assignment) and horizontal (process) views may consist of many activities that are each networked together based on their relationships to resources, cost objects, and other activities.

To be clear, both of the views are of increasing value to managers and employees because the General Ledger's chart-of-accounts is structurally deficient. It is an excellent instrument for accumulating transactions into buckets, but it fails at supporting decision making.

Expanding the Two-Stage ABC/M Cross Model

The fundamental structure of the ABC/M Cross Model appears to have withstood the test of time. However, the two-stage allocation appearance can leave an impression that tracing costs is an overly simplistic process.

The two primary elements of an ABC/M application will always be:

1. the three cost modules and two cost assignments of the cost object assignment view, and

2. the linking of sequential activities in the process view as processes and sub-processes.

The initial focus of early ABC/M applications was the determination of product costs through better segmentation of resource consumption. Subsequent applications in larger and more complex organizations revealed that ABC/M data solved broader problems. For these solutions, the ABC/M cost calculation usually required more than a simple socalled "two-stage" cost re-assignment, as indicated by the cost assignment view of the ABC/M Cross. Figure 6 illustrates the expansion of the cost assignment network from a two-stage to a multi-stage network.

The expanded ABC Model includes intermediate stages of activities i.e., activity outputs that are inputs to successive work activities. Specific usage, not time-based sequence, is the dominant factor for determining this cost assignment structure. These intermediate input/outputs cannot easily be traced directly to final cost objects (i.e., products, service lines, types of channels, or customers) since there is no causal relationship. As a result, intermediate activities are two or more stages removed from a final cost object.

The need for multi-stage cost assignments, in contrast with the earlier use of a simplistic two-stage assignment, has been due in part to assure accurate costing. A substantial and material amount of organizational work activities support the more primary activities that are in closer proximity to products and customer services. Organizations often refer to this support-related work as overhead. These support-related activity costs raise the question, "How much of this activity is consumed by specific products or service-lines?" But it is virtually impossible to answer this question because the work is simply too indirect and remote from the products to detect or sense any differences. However, these support activities can be traced in proper proportions to other activities that require their work. Such support-related activity costs are eventually burdened into the primary activity costs. These intermediary activities support the work activities that do detect the variation and diversity of the products or service line. In sum, the ABC/M uses multi-stages to trace all of the costs through a network of cost assignments into the final cost objects.

In summary, many activity cost relationships are in reality indirect with respect to the organization's resources and final cost objects. As a consequence, the network-like structure of the expanded ABC Model.

Stages of Evolution of Allocating Costs

As just described, ABC/M cost assignment structures are now recognized to have multi-stage re-assignments, not just two steps. There are activity-to-activity assignments for overhead and support costs, which cannot directly reflect variation with final cost objects, and there are also cost object-to-cost object assignments.

An example of the latter assignment type is the costs to process a special transaction (versus a standard transaction) traced to a specific customer or a group of customers consuming a unique quantity and mix of products, services, or outputs.

Figure 9 is analogous to Charles Darwin's model for the evolution of the species. The left graphic represents a single-celled paramecium. The middle graphic represents reptiles, amphibians, and snakes. The right graphic represents man beginning to walk upright.

The evolution of the vertical ABC/M cost re-assignment network starts with the simplistic allocations of the traditional accounting system, and it ends with a multi-stage network of costs flowing through ABC/M's three cost modules. This multi-stage arterial costing network is capable of detecting greater diversity and variation in not only product costs, but also all final cost objects, including different types of customers.

In Figure 9 the left graphic is primitive. It represents traditional accounting's "cost allocation" method that simply re-distributes the source costs into destinations, such as product costs, without regard to logical causality. Ashamedly, many financial controllers still allocate their organization's costs this way. This method of allocating costs does not question if any cause-and-effect correlation exists - it uses an allocation factor or basis that is often convenient, like the amount of dollars, regardless of physical effort. But without any causal relationship, there is an expected, undesirable error in calculating the costs of cost objects. The results are inaccurate costs. Sadly, the financial controllers who continue to allocate costs this way are misleading their end-users with flawed data. Some people view this as an irresponsible act on their part.

The middle graphic in Figure 9 represents the ABC/M Cross model, where the expenditures for resources are assigned at the work activity level, not at a department level (i.e., using verb-adjective-noun grammar to define activities). At the work activity level, the amount of activity costs varies linearly with changes in the quantity of their activity driver. In the ABC/M Cross model, the use of multiple resource and activity drivers reflects the unique consumption relationship between resources, activities, and cost objects. As a result, with ABC the calculated costs of the cost objects are more accurate. This is the minimum entry into ABC/M, but it is too simplistic to be adequate.

The right-hand side of Figure 9 symbolically shows the ABC vertical cost assignment as a multi-stage cost assignment network with an expanded structure that allows for: (1) intermediate activities and activity drivers, and (2) cost objects being traced into other cost objects. This graphic is labeled as the expanded ABC/M model in order to distinguish it from the obsolete two-stage ABC/M Cross model. Thus, the three modules of the ABC/M Cross model have now matured to become a multistage network of activities and objects. This cost assignment network has the flexibility to link resources to their cost objects - and the tracing relies on cause-and-effect relationships. Hence, the complete cost assignment network leads to much greater accuracy of cost object costs.

The BCAM provides a tremendously rich source of activity definitions and their location with respect to business' processes. Using these work activities, the multi-stage ABC/M structure can be designed and populated with data for calculating costs.

3) THE AMirs BANK COST ANALYSIS MODEL (BCAM)

AM Ifs has several objectives in creating this model.

1) To provide an example for its members to unleash the power of ABC/M in their institutions in a manner that is straightforward and replicable

2) To provide a basis on which to build additional guidelines on some of the more complex issues and difficult decisions associated with ABC/M

3) To provide a basis for the development of meaningful and comparable operational benchmark information for its members and the industry

To achieve these objectives, our approach is to start simple and move gradually to the complex. This top-down approach is intended to constantly re-focus the work on a practical cost vs. benefit basis. A good guideline to prevent constructing an overly complicated ABC/M system is to constantly ask, "Is the view worth the climb?" That is, are we building excessive detail and precision into our ABC/M system, well beyond what is necessary?

Benefits from Comparable Benchmarking Data

If successful AM I fs members will have reached a consensus on a level of operational activities (level 2 and 3) that allows for not only a powerful decision-making model but also a tremendous resource for benchmark operations for all its members. A shortcoming of today's benchmarking is the "apples-to-Oreos" problem resulting from inconsistencies. Organizations that measure benchmark data often unknowingly do not include or exclude the same parts of their organization. ABC/M brings the rigor of definition without a great effort of work.

The initial BCAM is intended to cover all processes of traditional banking. It will be expanded in the future to include elements such as trust, brokerage, insurance, and other fee businesses.

Error-Dampening: The Key to Right-Sizing BCAM for Quick Results

ABC/M has an extremely useful property of error-dampening that means the BCAM model does not need to be very granular and consequently very large in size. Unfortunately, because this is counter-intuitive, many ABC/M systems have been over-engineered and therefore become weighed down with undue complexity. Many ABC/M models go well beyond diminishing returns in extra accuracy for extra levels of effort of work. We refer to this as the "leveling problem." No one knows in advance how detailed to build the first model. This error-dampening property of ABC/M allows the AMIfs project team to choose not to err on the side of excess detail. In the end, the level of accuracy and detail depends on what decisions are made with the data. And usually the accuracy requirements are not unreasonably harsh.

The error-dampening property is the result of work activities being consumed by the work upstream from the ultimate products and customer services, which initially trigger the demands on work. The reassignment of cost based on activity consumption is responsible for the majority of ABC/M's superior costing accuracy. Therefore, an ABC/M model can tolerate reasonable resource driver estimates as proxies for precise detail drivers because the error from estimates there does not compound - it dampens out on its way to the final cost objects. That is, the over- and under-costed activity costs offset when they pile-up into the product or customer cost object. Admittedly, this is somewhat counter-intuitive, but with ABC/M, precision inputs are not synonymous with accurate outputs. Consumption of activities is a much greater determinant of accuracy. This property significantly lightens the load for data collection or, at a minimum, allows much greater ability to determine the cost and potential benefit of additional detail in data collection.

Basic Structure of Bank Cost Analysis Model

Traditionally, banks considered interest to be the only direct cost of the bank, leaving all non-interest costs classified as indirect or overhead costs. Certainly interest represents the largest expense of most banks. What if your bank was only funded by non-interest-bearing demand deposits? Would you price loans the same? What about deposits? Interest expense would no longer be the largest expense of the bank, and we would likely earn far less from our lending because the lack of an interest expense would allow for lower pricing on loans and maybe even greater margins. But would the bank be more profitable? The cost structure and the economics in this situation would be very different than a bank that was purely funded by wholesale funding sources. Both banks could be very profitable but for completely different reasons.

The fact is every bank is made of both of these extremes and every imaginable combination in between, and it is growing more complex every day. By breaking the cost structure down to its simplest terms and then rebuilding it, we can better understand the true economics of our business lines, products, customers, channels, etc. Furthermore, we can design products and services to maximize profitability and improve value propositions with our stakeholders. That is what activity-based cost management in banks is all about.

In its simplest terms, a bank as a financial intermediary buys money from suppliers through various means and sells that same money to its customers. The true economics of each of these funding sources is very different. The funds are the raw materials of the bank that are repackaged into the products sold by the bank. These repackaged raw materials sold to various customers also provide very different revenue streams, risks, and cost structures. This intermediation function of banking can most simply be expressed as the buying and selling of money through the development of relationships with customers. The customer relationships are maintained and deepened by the level of service provided to them.

The level and complexity of customer service that is provided drives the economics behind all revenues and expenses of the bank. By level of service, we mean the products offered, the convenience provided, the people and the technology providing these services, etc. Why can some banks pay lower rates on deposits and charge higher rates on loans? The answer lies in the level of service. The elements of a bank's service and their associated costs determine the character of the bank and its profitability. Yet banks know infinitely more about the cost of interest than they do about the cost of service.

Given this view of banking, a robust methodology for understanding the changing cost structure of the bank and its true underlying economics becomes imperative. That is why an advanced methodology of costing is necessary in banks. To be sure, ABC/M was not recently invented. It is simply a refined form of absorption accounting. The increases in diverse types of products, channels, and customers that have caused complexity, and thus greater indirect costs to manage the complexity, require this refined application of assigning expenses into costs.

As service organizations facing the continued commoditization of products and increasing levels of competition in the buying and selling of money, our industry must face this harsh reality. Although activity-based cost management was not developed in the banking industry, it is the first place it should be used if you believe that service and its related cost determines profitability in the industry.

4) SCAM IMPLEMENTATION

The first step in the implementation process is to categorize expenses into BCAM's levels one, two, and three. Because of the structure of the General Ledger, there will initially be many categories of expenses that will be included in more than one level of BCAM's expenses. In some cases you will need to go to an additional source of data to decompose the General Ledger accounts down to the level needed for ABC/M. For example, payroll ledgers for personnel expenses, contracts or invoices for technology expenses, space utilization plans or lease agreements for facilities cost, etc. In other cases, the General Ledger accounts may be overly detailed, and account balances may require aggregating into a summary.

Regardless of the level of detail in the general ledger, the key concept at this stage is to identify all expenses that can be directly or indirectly traced to a product, customer, channels, etc. From BCAM's perspective, the General Ledger is an external source of data that will be imported into BCAM as well as driver quantity data used for tracing and accumulating the costs. Effective decision-making should be done at the margin, and the nature of the decision should govern how detailed and precise the information needs to be. BCAM provides flexibility to use summarized information and selectively drills down into more detail only if needed.

Once expenses are categorized, your resource drivers can be applied to assign cost to level three activities, level two processes and level one functions. Each bank is different. Banks have different cost structures, products, service-lines, strategies, policies, and procedures. However, the generic processes and activities for all banks should be the same, and different banks simply do more or less or none of various activities. The modeling of a bank's work activities allows all banks to utilize the SCAM model and ultimately to benefit from a set of benchmark data that is consistent and comparable. As we discussed earlier, each institution may want or need to further break down and decompose any of these activities to a lower level. When doing so, it is important to be careful to evaluate the cost and benefit of the additional detail and maintain the hierarchy of BCAM so data will be comparable across the industry.

An important part of ABC/M and the use of the BCAM is in the development of activity drivers and their relationship to cost objects. To realize the value of ABC/M in banking through the use of BCAM, reasonably accurate activity drivers should be collected and applied to the costs of each activity and process. This modeling process provides rich information for the evaluation of profitability of cost objects such as customer, products, channels, etc. When ABC/M costs are used with a robust transfer pricing tool, a complete picture of the profit margin layers and the economics of various objects becomes very clear.

But this is only the tip of the iceberg for the value of ABC/M because the really tough decisions about new product offerings, facilities, and technology expansions/upgrades can now be modeled to provide a better basis for decision-making.

Primary Objective of Costing in Banks

The primary objective of BCAM data is to simply make visible the economics of the organization and its consumption of resource expenses. You want to understand the connection between types of customers and the resource expenses they uniquely consume. A bank is continuously burning up its resources through its work activities and into its product and service-line outputs whether BCAM is monitoring these events or not. When one realizes that BCAM is fundamentally good data to be used for understanding, discovery, and decision-making, then BCAM is better positioned for longer-term use and wider acceptance.

5) IMPLEMENTING ABC/M IS MORE FEASIBLE ONCE THE MISCONCEPTIONS ARE DISPROVED

For years, activity-based cost management (ABC/M) was considered an expensive project that only large organizations with extensive resources could undertake. But today, with the proliferation of computers for gathering and computing, the cost of data collection and measurement has fallen at the same time as information processing has improved. Not too long ago, it was cost-prohibitive to accumulate, process and analyze the data necessary to run an ABC/M system. Cost accounting was restricted to a big box mainframe computer and data stored as flat-files.

Today, such activity measurement systems are not only affordable, but much of the data already exists in some form within the organization. For example, quality management systems of ISO 9000 registered organizations have an abundance of data - usually disconnected from the accounting system. Also, a few knowledgeable employees can usually estimate a large portion of any remaining data that may not be available. Estimates such as these will have a minimal adverse impact on accuracy. Hence, all organizations can easily produce ABC/M data. Better yet, information technology has dramatically improved the deployment of ABC/M data for viewing, planning, and decision-making. Powerful database management systems and computer engines means that data processing is no longer the impediment to understanding costs.

Figure 32 illustrates that the ABC/M system is not a transaction-based system, but rather it sits on top of transaction-based systems as an analytical tool. That is, ABC/M is now considered to be an analytical application by information technologists. By definition, analytical applications transform or draw on summary data from transaction-intensive operational systems. Analytical applications, such as ABC/M, are separate and apart from the transactional systems.

Also note from Figure 32 that ABC/M integrates with a variety of other analytical applications such as performance measurement scorecards and process simulation. This suite of analytical tools will become increasingly important as banks even further leverage their information technologies.

What makes ABC/M even more realizable are the lessons learned that most data for decision-making need not be accurate to several decimal places. Also, a technique for implementing ABC/M based on rapid prototyping scale models is assuring implementation success. In contrast to the long multi-month, one-chance, single-design approach, the ABC/M rapid prototyping technique follows the quick build of the initial model, built roughly in two days, by iterative re-modeling of increasingly larger-scale ABC/M models. Eventually, the larger-scale ABC/M model becomes the organization's repeatable and reliable production system.

ABC/M, BCAM, and the Future

An overarching issue with ABC/M involves its perception as just another way to spin financial data rather than its use as mission-critical managerial information. The Information Age we are entering can be mind-boggling. In our future, as technology advances, so will the demand to access massive amounts of relevant information. The companies that survive will be those that can answer these questions:

* "How do we access all this data?"

* "What do we do with it?"

* "How do we shape the data and put it in a form with which we can work?"

* "What will happen when we apply technologies developed during the Information Age forthe Information Age?"

Clearly, as information technology evolves, organizations will increase their effectiveness. Further, as markets change, banks will run into global competitors that increasingly look to information and information technology for competitive advantage. ABC/M is involved in this broad arena of "outsmartmanship."

ABC/M puts the "management" back into management reporting. For those people who are involved with ABC/M projects, the key is to create and orchestrate change rather than merely react to it and attempt to make the best of a poor situation.

It will be fun watching banks and other financial institutions move from their learning stages into mastery of building and using ABC/M systems. It is hoped that SCAM will be accepted as a foundational structure from which managers and employees in financial institutions can better improve their operational performance and align their cost structure with their organization's strategy

Next Steps:

* Give AMI fs feedback on BCAM - we cannot have consensus without participation.

* Get involved in Pilot projects or realign existing models to fit BCAM.

* Give us your ideas for expansion of BCAM.

* Volunteer for the phase of project.

By J. Timothy Smith* and Charlene Harper*

* From the AMIfs Research Committee