Financial models: the accountant's edge.
In his book Patriot Games Tom Clancy says that in almost all situations when intelligence has been gathered, the data are such a mass of information that they cannot be used. Available techniques do not allow easy analysis of large volumes of data.
CPAs in particular need techniques that produce accurate and effective data analysis and enable them to forecast expectations, evaluate alternatives and provide corporate managers with the details they will need on the most relevant aspects of a business.
Models, mathematical formulas or equations that state the relationships between important factors (variables) in a problem or system provide an efficient way to study relationships. Instead of observing every value of a variable, one simply gathers representative sample data and formulates inferences or decisions on the population as a whole and, ultimately, the degree of relationship that exists between cerlain variables. (For an example of a simple model, see the exhibit on page 114.)
Modeling techniques can be applied to every aspect of business: production, distribution, research and development, inventory control, capital investments and pricing decisions. For example, the financial analyst may develop one equation to optimize inventory levels and another to represent the relationship between a company's sales and the issues that affect them.
Models offer an extra edge by
* Providing information that is sometimes impossible to collect.
* Helping to quantify uncertainties.
* Measuring relationships between variables.
* Providing a measure of reliability.
* Enabling managers to test risky decisions or experiment with complex situations on paper before implementation.
* Considering many managers' data and experiences.
* Removing the biases that may be held by decision makers and financial personnel.
* Offering a flexible and reusable tool for similar future decision making.
As models gain greater use, corporate executives will turn to CPAs for answers to a variety of different questions. Which color, shape and size package will induce customers to purchase a product? Will lower or higher purchase payments ensure better repayment trends? Can a company reasonably estimate next year's sales based on projected gross national product and past financial statements? Does lowering or raising prices have a significant effect on total sales?
A CASE IN POINT
The answers to most of these questions are purely speculative without modeling techniques. Take for example, the pricing query. It is commonly assumed that lowering prices should increase sales, but one Houstonbased accounting firm found in an economy-and-efficiency study that it performed that the price of a drink doesn't necessarily affect total bar sales.
Objective. The study was intended to determine how to increase bar operating revenues, decrease operating expenses and improve customer service in 25 officer, enlisted and consolidated open mess bar operations in 15 continental U.S. and overseas military installations. The main areas of concern were beverage control, bar operations, club management and external considerations and regulatory controls.
Precursory study review. The first concern was to identify and evaluate all possible factors that might affect sales from data gathered at several management interviews, on-site evaluations and a thorough review of applicable financial' statements. Although many pertinent aspects of open mess operations were covered, this article addresses the use of models for managerial decision making.
Mathematical analysis and regression model. The accountants chose a regression model to accompany their report because of the need for a detailed analysis of the relationships of the variables that might affect sales.
A regression model functionally relates sales (or another dependent variable) to other economic, competitive or internal variables and forms an equation using the straight-line or least-squared method. Expanding one step further, the model then calculates measures of correlation, which show the percentage of relationship between the variables. Statistical computer software packages (starting around $45) help add and average data and the subsequent calculations of complex equations and percentages. (See sidebar at left for recommended packages.)
If a CPA believes product price directly affects total sales, he or she could use a software package to graph price against sales and analyze the results.
This management tool can be used to predict future sales or tailored to project other business variables, such as payroll, shipping or insurance costs.
IS PRICE MOST IMPORTANT?.
Finding the optimal sales level involves a careful mix of many factors. Sometimes what one perceives as an important variable may not even factor into the big picture as the Houston CPAs discovered about price. This is one reason for using regression analysis. A firm can analyze which factors are important in the decision, whether all factors have been considered and each factor's contribution.
In this study, additional statistical data were gathered and analyzed to verify the investigators' judgmental observations. Next, analyses were performed on the data. Noticing one problem area, the accountants extracted from club records cost data for a sample of widely used liquors to determine whether clubs consistently made the best purchasing decisions. If clubs as a whole obtained the best price for products, one would expect to find a tight cluster of unit costs for a given liquor. The analysts performed separate sets of calculations for both domestic and overseas installations.
The CPAs judged the variability of liquor costs to be too large. Put into perspective, costs at 5% of the clubs probably would vary by more than 20% of the average or mean cost of a particular brand. Chivas Regal, for instance, cost a club on the average $.43 per ounce. Yet some clubs paid as little as $.33 or as much as $.49 per ounce under identical purchasing conditions.
The results indicated not all clubs made the best purchasing decisions. Continuing the analysis, the investigators next employed the statistical technique of multiple regression, which uses a linear equation to predict values for a variable. Since this isn't a detailed study of statistics but, rather, the appropriate application of models, the more complicated details of deriving the equations are omitted. A good statistics book will explain each pertinent equation.
In the bar study, 25 factors were considered, including number of clubs, club type (officer, enlisted, consolidated), military command, bar and dining sales, member fees, product cost, advertising, snacks, domestic and imported beer prices, cocktail price, hours of operation, number of bars, number of beers and liquors available and liquor cost.
After collecting data on these factors, the accountants performed a variety of calculations for each variable group. They quantified relationships, which enabled them to make sound recommendations, particularly on pricing and increasing profit.
The CPAs drew some surprising conclusions from their models:
* The price charged for drinks did not affect sales significantly.
* Social hour snacks and other expenditures had no effect on sales.
* A wider drink selection did not stimulate sales noticeably.
* There was no relationship between restaurant and bar sales.
Using the information generated by the regression model, the analysts recommended a $.10 to $.25 price increase for house liquors and a $.70 increase for premium drinks. Additionally, they recommended that open messes increase drink prices another $.25 during entertainment. This would considerably increase drink profit margins, since price seemed to have no effect on purchasing. The CPAs concluded that patrons probably have fixed dollar budgets for alcohol. The budget and drink prices determine the number of drinks purchased.
The CPAs used models to deliver a comprehensive and quantifiable picture of operations that management can use to improve operations. CPAs who wish to gain more knowledge about using models might turn to courses in decision sciences or quantitative methods at a local university. In addition, the American Institute of CPAs practice aid, Preparing Financial Models, published by the management consulting services technology and industry consulting practices subcommittee, is a good resource (practice aid no. 92-6 is $13; order no. 055137).
To employ models, CPAs must gather pertinent data, correctly apply the models and astutely interpret the results. By doing so, they can provide workable solutions and procedures that ultimately increase bottom-line revenues.
|Printer friendly Cite/link Email Feedback|
|Author:||Hamblin, David E.|
|Publication:||Journal of Accountancy|
|Date:||Nov 1, 1992|
|Previous Article:||Calculating child support awards: new opportunities for CPAs.|
|Next Article:||Marketing the practitioner and the firm.|