9

PREDICTIVE ACCOUNTING AND BUDGETING WITH MARGINAL EXPENSE ANALYSIS¹

Managers are increasingly shifting from reacting to after-the-fact outcomes to anticipating the future with predictive analysis and proactively making adjustments with better decisions. Despite some advances in the application of new costing techniques, are management accountants adequately satisfying the need for better cost planning information? Or is the gap widening?

There is a widening gap between what management accountants report and what managers and employee teams want. This does not mean that information produced by management accountants is of little value. In the last few decades, accountants have made significant strides in improving the utility and accuracy of the historical costs they calculate and report. The gap is caused by a shift in managers’ needs, from just needing to know what things cost (such as a product cost) and why, to a need for reliable information about what their future costs will be and why.

Despite the accountants advancing a step to prioritise the increasing needs of managers, the managers have advanced two steps. In order to understand this widening gap and, more importantly, how accountants can narrow and ideally close the gap, let’s review the broad landscape of accounting as described in Chapter 4, ‘A Taxonomy of Accounting and Costing Methods.’

WHAT IS THE PURPOSE OF MANAGEMENT ACCOUNTING?

Contrary to beliefs that the only purpose of managerial accounting is to collect, transform and report data, its primary purpose is, first and foremost, to influence behaviour at all levels—from the desk of the CEO down to each employee—and it should do so by supporting decisions. A secondary purpose is to stimulate investigation and discovery by signalling relevant information (and consequently bringing focus) and generate questions.

The widening gap between what accountants report and what decision makers need involves the shift from analysing descriptive historical information to analysing predictive information, such as budgets and what-if scenarios. All decisions obviously can only affect the future because the past is already history. However much can be learned and leveraged from historical information. Although accountants are gradually improving the quality of reported history, decision makers are shifting their view toward better understanding the future.

This shift is a response to a more overarching shift in executive management styles, from a command and control emphasis that is reactive (such as scrutinising cost variance analysis of actual versus planned outcomes), to an anticipatory, proactive style in which organisational changes and adjustments, such as staffing levels, can be made before things happen and minor problems become big ones.

Figure 4-1 in Chapter 4 illustrated the large domain of accounting as a taxonomy with three components: tax accounting, financial accounting and managerial accounting. As a review, two types of data sources are displayed at the upper right in the figure. The upper source is from financial transactions and bookkeeping, such as purchases and payroll. The lower source is non-financial measures, such as payroll hours worked, retail items sold or gallons of liquid produced. These same measurements can be forecasted.

The financial accounting component in the figure is intended for external reporting, such as for regulatory agencies, banks, stockholders and the investment community. Financial accounting follows compliance rules aimed at economic valuation and, as such, is typically not adequate or sufficient for decision making, and the tax accounting component is its own world of legislated rules.

Our area of concern—the management accounting component—was segmented into three categories: cost accounting, the cost reporting and analysis and decision support with cost planning. To oversimplify a distinction between financial and managerial accounting, financial accounting is about valuation, and managerial accounting is about value creation through good decision making.

The message at the bottom of the figure is that the value, utility and usefulness of the information increases, arguably at an exponential rate, from the left side to the right side of the figure.

Figure 4-2 in Chapter 4 illustrated how a firm’s view of its profit and expense structure changes as analysis shifts from the historical cost reporting view to a predictive cost planning view. The latter is the context from which decisions are considered and evaluated.

WHAT TYPES OF DECISIONS ARE MADE WITH MANAGERIAL ACCOUNTING INFORMATION?

University textbooks contain hundreds of pages on managerial and cost accounting. Let’s try to distil all those pages into a few paragraphs. The broad decision-making categories for applying managerial accounting are as follows:

Rationalisation

Which products, stock keeping units, services, channels, routes, customers and so forth are best to retain or improve? Which are not and should potentially be abandoned or terminated?

Historical and descriptive costing (the left side of Figure 4-2) can be adequate to answer these questions. In part, this explains the growing popularity in applying activity-based cost principles to supplement traditional direct costing. Considerable diversity and variation exists in routes, channels, customers and so forth that cause a relative increase in an organisation’s indirect and shared expenses to manage the resulting complexity. IT expenses are a growing one. Having the direct and indirect costs become a relevant starting point allows you to know what the variations cost. This answers the ‘What?’ question discussed in Chapter 4. It is difficult, arguably impossible, to answer the subsequent ‘So what?’ question without having the facts. Otherwise, conclusions are based on gut feeling, intuition, misleading information or politics.

Planning and Budgeting

Based on forecasts of future demand volume and mix for types of services or products, combined with assumptions of other proposed changes, how much will it cost to match demand with supplied resources (eg, workforce staffing levels)?

When questions like these and many more like them are asked, one needs more than a crystal ball to answer them. This is where the predictive view of costing (the right side of Figure 4-2) fits in, and the predictive view arguably is the sweet spot of costing. On an annual cycle, this is the budgeting process. However, executives are increasingly demanding rolling financial forecasts at shorter intervals, partially due to the fact that the annual budget can quickly become obsolete, and future period assumptions, especially continuously revised sales forecasts, become more certain. At its core, this costing sweet spot is about resource capacity planning (the ability to convert and reflect physical operational events into the language of money) expenses and costs.

Capital Expense Justification

Is the return on investment of a proposed asset purchase, such as equipment or an information system, justified?

If we purchase equipment, technology or a system, will the financial benefits justify the investment? A question like this involves what microeconomics refers to as capital budgeting. Capital budgeting analysis typically involves comparing a baseline, reflecting business as usual, with an alternative scenario that includes spending on (ie, investing in) an asset in which the expected benefits will continue well beyond a year’s duration. An example would be investing in an automated warehouse to replace manual, pick-and-pack labour. Some refer to the associated investment justification analysis as same as, except for or comparing the as-is state with the to-be state. A distinction of capital budgeting is that it involves discounted cash flow (DCF) equations. DCF equations reflect the net present value of money, incorporating the time that it would take for that same money to earn income at some rate if it were applied elsewhere (eg, a bank certificate of deposit). The rate is often called the organisation’s cost of capital.

Make Versus Buy, and General Outsourcing Decisions

Should we continue to do work ourselves or contract with a third party?

If we choose to have a third party make our product or deliver our service—basically outsourcing (or vice versa by bringing a supplier’s work in-house)—then how much of our expenses remain, and how much will we remove (or add)? This type of decision is similar to the logic and math of capital budgeting. The same description of the capital budgeting method applies, measuring ‘same as, except for’ incremental changes. Activity-based costing techniques ideally should be applied because the primary variable is the work activities that the third party contractor will now perform, which replace the current in-house work. Because cost is not the only variable that shifts, a service-level agreement with the contactor should be a standard practice.

Process and Productivity Improvement

What can be changed? How do you identify opportunities? How do you compare and differentiate high impact opportunities from nominal ones?

Some organisation’s operations functions focus on reducing costs and future cost avoidance. (Strategic profitable revenue enhancement is addressed with managerial accounting for rationalisation.) These operational functions are tasked with productivity improvement challenges, and they are less interested in understanding strategic profitability analysis, that is, which of our priced products and services makes or loses money, and more on streamlining processes, reducing waste and low value-added work activities and increasing asset utilisation. This is the area of Six Sigma quality initiatives, lean management principles and just-in-time scheduling techniques. Examples of these types of costs are as follows:

• Unit costs of outputs and benchmarking

• Target costing

• Cost of quality

• Value-adding attributes (such as non-value added vs. value-added).

• Resource consumption accounting (RCA)

• German cost accounting (Grenzplankostenrechnung [GPK])

• Accounting for a lean management environment (also Kaizen costing)

• Theory of constraint’s throughput accounting

The term cost estimating is a general one. It applies in all the preceding decision-making categories. One might conclude that the first category, rationalisation, focuses only on historical costs and, thus, does not require cost estimates. However, the impact on resource expenses from adding or dropping various work-consuming outputs (ie, products, services and customers) also require cost estimates to validate the merit of a proposed rationalisation decision.

Activity-Based Cost Management as a Foundation for Predictive Accounting

In the late 1990s, the more mature and advanced activity-based costing users increasingly began using their activity-based cost management (ABC/M) calculated unit cost rates for intermediate work outputs and for products and services as a basis for estimating costs. As previously described, popular uses for the activity-based costing data for cost estimating have been to calculate customer order quotations, perform make versus buy analysis, and budget. The ABC/M data were being recognised as a predictive planning tool. It is now apparent that the data have a tremendous amount of utility for both examining the as-is current condition of the organisation and achieving a desired to-be state. (As mentioned earlier, a more robust version of ABC/M is RCA, which is based on the German accounting practice GPK for marginal expense analysis and flexible budgeting for operational control. RCA is a comprehensive approach that focuses on resources and capacity management logic with ABC/M principles.²)

Cost estimating is often referred to as what-if scenarios. Regardless of what the process is called, we are talking about decisions being made about the future, and managers wanting to evaluate the consequences of those decisions. In these situations, the future of the organisation is not very distant and, in some way, the quantity and mix of activity drivers will be placing demands on the work that the organisation will need to do. The resources required to do the work are the expenses. Assumptions are made about the outputs that are expected. Assumptions should also be made about the intermediate outputs and the labyrinth of inter-organisational relationships that will be called upon to generate the expected final outcomes.

MAJOR CLUE: CAPACITY ONLY EXISTS AS A RESOURCE

As most organisations plan for their next month, quarter or year, the level of resources supplied is routinely re-planned to roughly match the firm customer orders and expected future order demands. In reality, the level of planned resources must always exceed customer demand to allow for some protective buffer, surge and sprint capacity. This also helps improve customer service in shipping performance. However management accountants will be constantly disturbed if they cannot answer the question, ‘How much unused and spare capacity do I have?’ because in their minds, this excess capacity equates to non-value-added costs.

The broad topic of unused and idle capacity will likely be a thorny issue for absorption costing. As management accountants better understand operations, they will be constantly improving their ability to segment and isolate the unused capacity (and the nature of its cost) by individual resource. Managerial accountants will be increasingly able to measure unused capacity either empirically or by deductive logic based on projected standard cost rates. Furthermore, accountants will be able to segment and assign this unused capacity expense to various processes, owners, the sales function or senior management. This will eliminate over-charging (and over-stating) product costs resulting from including unused capacity costs that the product did not cause.

Figure 9-1 illustrates that the effort level to adjust capacity becomes easier farther out in time. It takes a while to convert in-case resources into as-needed ones. However committed expenses (in-case) today can be more easily converted into contractual (as-needed) arrangements in a shorter time period than was possible ten years ago. Fixed expenses can become variable expenses. The rapid growth in the temporary staffing industry is evidence. Organisations are replacing full-time employees who are paid regardless of the demand level with contractors who are staffed and paid at the demand level, which may be measured in hours.

Figure 9-1: Capacity Only Exists as Resources

In the very short term, you would not fire employees on Tuesday due to low work load, then hire them back on Wednesday. In the future you may replace full-time employees with contractors, or lease assets you might have purchased. In this way, so-called ‘fixed costs’ behave variably.

Source: Copyright Gary Cokins. Used with permission.

Understanding the cost of the resource workload used to make a product or deliver a service is relevant to making these resource re-allocation decisions. Ignoring incremental changes in the actual resources (ie, expense spending) when making decisions can eventually lead to a cost structure that may become inefficient and ineffective for the organisation. There will always be a need to adjust the capacity based on changes in future demand volume and mix. This, in turn, equates to raising or lowering specific expenses on resources.

PREDICTIVE ACCOUNTING INVOLVES MARGINAL EXPENSE CALCULATIONS

In forecasting, the demand volume and mix of the outputs are estimated, and the unknown level of resource expenditures that will be required to produce and deliver the volume and mix is solved for. You basically are determining the capacity requirements of the resources. Estimating future levels of resource expense cash outflows becomes complex because resources come in discrete, discontinuous amounts. For example, you cannot hire one-third of an employee. That is, resource expenses do not immediately vary with each incremental increase or decrease in end-unit volume. Traditional accountants address this with what they refer to as a step-fixed category of expenses.

The predictive accounting method involves extrapolations that use baseline physical and cost consumption rates that are calibrated from prior period ABC/M calculations. Managerial accountants relate predictive accounting to a form of flexible budgeting (which is normally applied annually to a 12-month time span).

Figure 9-2 illustrates how capacity planning is the key to the solution. Planners and budgeters initially focus on the direct and recurring resource expenses, not the indirect and overhead support expenses. They almost always begin with estimates of future demand in terms of volumes and mix. Then, by relying on standards and averages (such as the product routings and bills-of material used in manufacturing systems), planners and budgeters calculate the future required levels of manpower and spending. The predictive accounting method suggests that this same approach can be applied to the indirect and overhead areas as well or to processes where the organisation often has a wrong impression that they have no tangible outputs.

Figure 9-2: Predictive Accounting Information Flow

Source: Copyright Gary Cokins. Used with permission.

Demand volume drives activity and resource requirements. Predictive accounting is forward-focused, but it uses actual historical performance data to develop baseline consumption rates. Activity-based planning and budgeting assesses the quantities of workload demands that are ultimately placed on resources. In step 1 in Figure 9-2, predictive accounting first asks, ‘How much activity workload is required for each output of cost object?’ These are the work activity requirements. Then predictive accounting asks, ‘How many resources are needed to meet that activity workload?’ In other words, a workload can be measured as the number of units of an activity requires to produce a quantity of cost objects.

The determination of expense does not occur until after the activity volume has been translated into resource capacity using the physical resource driver rates from the direct costing and ABC/M model. These rates are regularly expressed in hours, full-time equivalents, square feet, pounds, gallons and so forth.

As a result of step 1, there will be a difference between the existing resources available and the resources that will be required to satisfy the plan—the resource requirements . That is, at this stage, organisations usually discover they may have too much of what they do not need and not enough of what they do need to meet the customers’ expected service levels (eg, to deliver on time). The consequence of having too much implies a cost of unused capacity. The consequence of having too little is a limiting constraint that, if not addressed, implies there will be a decline in customer service levels.

In step 2, a reasonable balance must be achieved between the operational and financial measures. Now capacity must be analysed. One option is for the budgeters, planners or management accountants to evaluate how much to adjust the shortage and excess of actual resources to respond to the future demand load. Senior management may or may not allow the changes. A maximum expense impact exists that near-term financial targets (and executive compensation plan bonuses) will tolerate. These capacity adjustments represent real resources with real changes in cash outlay expenses, if they were to be enacted.

Assume that management agrees to the new level of resources without further analysis or debate. In step 3 of the flow in Figure 9-2, the new level of resource expenditures can be determined and then translated into the expenses of the work centres and, eventually, into the costs of the products, service lines, channels and customers. Step 3 is classic cost accounting—but for a future period. Some call this a pro forma calculation. The quantities of the projected resource and activity drivers are applied, and new budgeted or planned costs can be calculated for products, service lines, outputs, customers and service recipients.

At this point, however, the financial impact may not be acceptable. It may show too small a financial return. When the financial result is unacceptable, management has options other than to continue to keep re-adjusting resource capacity levels. For example, they can limit the amount of customer orders they accept. These other options may not have much effect on expenses.

DECOMPOSING THE INFORMATION FLOWS FIGURE

Figure 9-3 decomposes Figure 9-2. It reveals five types of adjustments that planners and budgeters can consider to align their expected demand with resource expenditures to achieve desired financial results. This approach has been called a ‘closed loop activity-based planning and budgeting’ framework.

Figure 9-3: Resource Capacity Planning and Costing

Source: Adapted from cam-i.org/docs/Methods_and_Applications.pdf.

Each of the five numbered options is intended to improve results, however, the relative impact of each adjustment will be unique to each organisation and its situation. As previously described, the predictive accounting model uses the forecasted demand quantities as its source input to determine if there is a degree of imbalance between the required and current existing resources.

Assuming that the result will be shortages and excesses of capacity, management can physically do the following:

• Adjust capacity. Additional manpower, supplies, overtime, equipment and the like can be purchased for shortages. There can be scale-backs and removals of people and machines for excesses.

• Adjust consumption rates. If possible, the speed and efficiency of the existing resources can be cranked up or down. If, for example, the increase in manpower makes a decision wasteful, fewer people can be hired, with an assumed productivity rate increase assumed.

• Adjust demand. If resources remain constrained, demand can be governed or rationed.

The latter two options are operational but also affect the level of resource expenses required. After this cycle of adjustments balances capacity of supply with demand, if the financial results are still unsatisfactory, management can make two incremental financial changes:

• Adjust pricing. In commercial for-profit enterprises or full cost recovery operations, pricing can be raised or lowered. This directly affects the ‘top line’ revenues. Of course, care is required because the price elasticity could cause changes in volume that more than offset the price changes.

• Adjust resource cost. If possible, wage levels or purchase prices of materials can be re-negotiated.

Predictive accounting acknowledges that there is a substantial amount of expenses that do not vary with a unit volume of final output. These expenses are not variable costs as determined by an incremental new expense for each incremental unit of output. As mentioned, resources often come in discontinuous amounts. Economists refer to these as step-fixed costs. An organisation cannot purchase one-third of a machine or hire half of an employee. Predictive accounting recognises the step-fixed costs in step 2, where resource capacity is adjusted. It recognises that as external unit volumes fluctuate, then

• some workload costs do eventually vary based on a batch-size of output or on some other discretionary factor.

• some resource expenditures will be acquired or retired as a whole and indivisible resource, thus, creating step-fixed expenses (ie, adding or removing used and unused capacity expenses).

Absorption costing is descriptive, and the only economic property for costing you need to deal with is traceability of cost objects back to the resources they consume. However, the descriptive ABC/M data is used for predictive purposes. The data provide inferences. In contrast, predictive accounting is forward-looking. Predictive accounting strives to monitor the impact of decisions or plans in terms of the external cash funds flow of an organisation. In the predictive view, determining the level of resource expenses gets trickier because you now have to consider an additional economic property—variability. The variability of resources is affected by two factors: (1) the step-fixed function because resources come in discontinuous amounts, and (2) adjustability of resources because the time delay to add or remove resource capacity can range from short to long.

Financial analysts simplistically classify costs as being either fixed or variable within the so-called ‘relevant range’ of volume. In reality, the classification of expenses as sunk, fixed, semi-fixed, semi-variable or variable depends on the decision being made. In the short term, many expenses will not and cannot be changed. In the long-term, many of the expenses (ie, capacity) can be adjusted.

FRAMEWORK TO COMPARE AND CONTRAST EXPENSE ESTIMATING METHODS

Figure 9-4 presents a framework that describes various methods of predictive cost estimating. The horizontal axis is the planning time horizon, short term to long term, right to left. The vertical axis describes the types and magnitudes of change in demands of the future relative to the recent past.

Figure 9-4: Methods of Forecasting Results

Source: Copyright Gary Cokins. Used with permission.

Examine the lower part of the figure, which illustrates the level of effort to adjust capacity across the planning time horizon. It describes expenses as becoming more variable and less committed as the planning time horizon lengthens. Historical cost rates can be more easily applied for longer time frame decisions, and there are fewer step-fixed expense issues. No defined boundary lines exist between the various zones, and there is overlap as one estimating method gives way to another as being superior.

Figure 9-4 illustrates in the upper right corner that as the time period to adjust capacity shortens and simultaneously the number of changes in conditions from the past substantially increase, it becomes risky to rely exclusively on rate-based extrapolation methods for cost forecasting. Discrete event simulation tools may provide superior and more reliable answers in this zone relative to the other methods. It can evaluate and validate decisions in any zone but, in particular, the upper right corner of Figure 9-4.

PREDICTIVE COSTING IS MODELLING

A commercial organisation ultimately manages itself by understanding where it makes and loses money or whether the impact of a decision produces incremental revenues superior to incremental expenses. Organisations are increasingly achieving a much better understanding of their contribution profit margins using ABC/M data. By leveraging ABC/M with predictive accounting and discrete event process simulation tools, an organisation can produce a fully integrated plan, including budgets and rolling financial forecasts. It can be assured that its plan is more feasible, determine the level of resources and expenditures to implement that plan, then view and compare the projected results of that plan against its current performance to manage its various profit margins.

The combination of these tools allows boardroom-level thinking to begin with the company’s complete income statement, generate a feasible operating plan and restate the results of that plan with an income statement—again, for boardroom reporting. Advocates of simulation planning software believe that the computing power of personal computers or seamless integration with servers, or both, now adequately provides simulation information that is comprehensive, finite-scheduled and rule-based and allows for various assumptions about uncertainty. Others argue that this is a last resort and that good modelling provides sufficiently accurate results.

All of this may sound like material from an introductory Economics textbook. In some ways it is, but there is a difference. In the textbooks, marginal expense analysis was something easily described but extremely difficult to compute due to all the complexities and interdependencies of resources and their costs. In the past, computing technology was the impediment. Now things have reversed. Technology is no longer the impediment—the thinking is. How you configure the predictive accounting model and what assumptions you make becomes critical to calculating the appropriate required expenses and their pro forma calculated costs.

DEBATES ABOUT COSTING METHODS

Confusion can arise because some of the costing methods calculate and report different costs that are not just variations in cost accuracy but are also different cost amounts. For example, should there be two or more different, co-existing cost reporting methods that report dissimilar numbers? One tactical costing method might be used for operations and making short-term decisions, whereas another strategic costing method (for planning, marketing, pricing and sales analysts to evaluate profit margins) is used for longer-term decisions.

There will be debates, but, eventually, some form of consensus will triumph within an organisation. The underlying arguments may be due to the inappropriate usage of standard costing information—and the potential inappropriate actions that may result. Therefore, key factor for deciding which costing method to use should be how does it handle economic projections? Can it accommodate classifying resource expenses as variable, semi-variable, fixed, sunk or as unavoidable or avoidable (ie, allowing for capacity adjustment decisions)? Does it isolate unused or idle capacity expenses?

The good news is that organisations are challenging traditional accounting. So in the end, any accounting treatments that yield better decision making should prevail. The co-existence of two or more costing approaches may cause confusion over which one reports the correct cost, but that is a different problem. What matters is that organisations are seeking better ways to apply managerial accounting techniques to make better decisions.

Endnotes

1 The ideas in this chapter on activity-based resource planning are based on research from a professional society. The researchers are from the Activity Based Budgeting Project Team of the Consortium of Advanced—International (CAM-I)Cost Management Systems (CMS) group. More information can be found at www.cam-i.org.

2 You can learn more about resource consumption accounting at www.rcainfo.com.