Pre-planning analysis for effective loan decisions

By Waymond Rodgers* and John Thomas**

The process of choosing among competing alternatives is decision making. This important managerial function is intertwined with planning and control. A bank manager or loan officer cannot plan without making decisions. These individuals need a model that helps them to select among competing objectives and methods. In this paper, a model is proposed which has significant implications for the development and implementation of a bank cost and management accounting system. Such a system can improve decisions by depicting and ordering information in several processing stages.

One of the major roles of a management accounting system is to supply information that facilitates decision making. For example, loan officers can use information more effectively when they consider and weigh different sources of information. Not only is economic, management, and financial information necessary for effective loan analysis, but also knowing at what stage to use this information and what weights to apply on this information is paramount.

A decision-making model can provide a framework of standards and points of reference within which individual banking personnel can operate with confidence, relative uniformity, and flexibility. Bank employees will then be able to make their own decisions within delegated authority, without the necessity for constant referral to higher management.

The decision-making processes of loan officers can be represented in an organized manner. In order to study these decision processes, it is important to show the relationship among perception, information, judgment, and decision choice as reported in Figure 1.

These four properties greatly contribute to loan officers' decision processes in the following ways. First, information and perception are interdependent because information has no meaning without individuals interpreting it (Rodgers 1991). Information affects judgments since it is stored in memory and affects summary inferences of loan officers (Rodgers 1992). Perception entails a phase of framing and editing, followed by judgment, which is a phase of evaluation. The first phase consists of a preliminary analysis of the decision problem, which frames the outcomes. Framing is controlled by the manner in which the loan problem is presented, as well as by expectancies of the loan officer. In the second phase or judgment stage, the framed thoughts are evaluated, and the decision to grant or refuse loans is made.

The function of the editing phase is to organize and reformulate the options for loan decisions so as to simplify subsequent evaluation (i.e., judgments) and decision choice. Editing consists of the application of several operations that transform the loan outcomes and probabilities associated with the loan choices of loan/do not loan. For example, Rodgers (1989) demonstrated that decision makers normally perceive outcomes as gains and losses (or loan/do not loan), rather than as final states of wealth or welfare. The loan/do not loan outcomes are defined by the loan officer to some neutral reference point. The reference point could be, for example, a company's current liquidity position, and the loan/do not loan outcomes coincide with the actual amounts that are granted. This reference point location, as well as the consequent coding of loan decisions as loan or do not loan, can be affected by the formulation of the loan situation and by the biases of the loan officer.

A number of experiments have been aimed at an explanation of decision-making biases (e.g., Rodgers, 1989). These results indicate that a major cause of decision makers' biases is due to misaggregation of the data. For example, decision makers perceive each datum accurately but are unable to combine its diagnostic meaning well with the diagnostic meaning of other data when revising their opinion (Edwards, 1982). Some researchers have indicated that decision maker's pattern recognition method of processing information can have a biasing effect on their framing and editing of information. Bottom-up processing, in which perceptual units combine to form larger units, is sometimes referred to as data-driven processing. In this kind of processing, presented information controls or directs an individual's processing. For example, net income from a set of accounting information would greatly influence how an individual processes and analyzes larger components of the information set. Top-down information, such as context and general knowledge, is sometimes referred to as conceptually-driven processing. In such processing, the perceiver imposes his/her knowledge and conceptual structures on the financial statement information set to help decide what type of organization is being represented.

Dividing loan officers into data-driven and conceptually-driven types can provide an anchor as a first approximation of their decision choice. This anchor, or bias, is then adjusted to accommodate the implication of additional information to be processed by decision makers. Typically, the adjustment is imprecise and insufficient.

Since loan officers come to loan making processes with certain decision biases for financial information, in general terms, they can be divided into two groups: those who prefer detailed information (or datadriven types) and those who prefer to look at the broader indicators (or conceptually-driven types). The data-driven types tend to anchor on large amounts of financial information, such as profitability and liquid assets, before they can begin the decision making process. These types also tend to place less emphasis on leverage (or risk) information, which is a global or conceptual indicator. Conceptually thinking loan officers anchor on overall financial indicators such as leverage information. They also are more apt to place less emphasis on specific detailed information such as liquid assets or profitability.

The advantage for the data-driven types' style of processing is to add more structure and certainty in analyzing financial statement information. Their disadvantages are related to less priority given to management risk factors and forecasting. To give priority to forecasting the future results of financial statements is an advantage of the conceptuallydriven types' style of processing, whereas their disadvantages relate to the lesser emphasis placed on the details in analyzing financial statements, which might be crucial for a good loan decision.

Knowing a loan officer's decision bias is not enough. These decision biases feed into their judgments (evaluations) of the loan approval process. During this process, loan officers assimilate certain types of information to reach general conclusions about the loan. From these general conclusions or inferences, the loan officer makes the critical decision to approve or disapprove the loan. Finally, loan officers may short-circuit their judgment process by relying too much on "blind faith" of a company performance (or perceptual impression) for loan decisions. The pathway from perception to decision choice in Figure 1 illustrates this point.

CONCLUSIONS

New approaches to loan making can aid managerial decisions. Existing approaches can be supplemented with analyses of the loan officers' decision biases and the effects they have on their actions. The model described in this paper may help assist loan officers in their credit analysis tasks. Our approach is one of many possible models. However, any new approach which takes into account decision processes should include the following important components:

(1) an analysis of loan officers' decision biases;

(2) an analysis of the effects these biases have on their judgments;

(3) the decisions that the loan officers make; and

(4) feedback designed to help officers understand the effects of their cognitive processes on their decisions.

In the future, banks should test and validate loan officers' decision processes with decision models. Assisting loan officers to make better decisions is one of the most important components of any new approach to reduce the likelihood of a bad loan decision. The approach suggested here may help solve incorrect loan decisions, and thereby improve the profitability of a bank.

REFERENCES

Edwards, W., 1982, "Conservatism in human information processing," In D. Kahneman, P. Slovic and A. Tversky (Eds.), Judgment Under Uncertainty: Heuristics and Biases, NY: Can-bridge University Press, pp. 359-369.

Rodgers, Waymond and Lester W. Johnson, 1988, "Integrating Credit Models Using Accounting Information With Loan Officers' Decision Processes, Accounting and Finance, Vol. 28, pp. 4-18. Rodgers, Waymond, 1989, "Decision makers' forecasting ability analyzed in a covariance structural model," International Journal of Management, Vol. 6, pp. 5-12.

Rodgers, Waymond, 1991, "Evaluating Accounting Information with Casual Models: Classification of Methods and Implications for Accounting Research," Journal ofAccounting Literature, Vol. 10, pp. 151-180. Rodgers, Waymond, 1992, `The Effects of Accounting Information on Individuals' Perceptual Processes," Journal of Accounting, Auditing & Finance, Vol. 7, pp. 67-93.

* Graduate School of Manager, University of Colombla, Riverside, CA 92521. Tel: (909) 787-4786;

Fax: (909) 787-3970.

**La Sierra University

Copyright National Association for Bank Cost & Management Accounting 1998
Provided by ProQuest Information and Learning Company. All rights Reserved