LOAN LOSS ESTIMATION MANAGEMENT BY FINANCIAL INSTITUTION MANAGERS AND COMMERCIAL LOAN OFFICERS1

Journal of Performance Management, 2008 by Deschênes, Sébastien

This paper presents a literature review of loan loss estimation management by bank managers. First, potential opportunistic behaviours by commercial loan officers at the time of loan granting and during file follow-up will be addressed. Then, in connection with the literature on loan loss estimation management by bank managers, the role of the commercial loan officers regarding loan loss estimation management in files involving businesses in financial difficulty will be explored.

I. Introduction

The agency theory is based on the existence of information asymmetries and conflicts of interest in organizations, between the agents and the principal or between agents (Levinthal, 1988). The more difficult it is for the principal to monitor the efforts made by the agents as they carry out their specific tasks, the more organizations are exposed to agency costs. The characteristics of the tasks (e.g. intellectual work vs. manual work) or the size of the organization can make the monitoring of agents more difficult. Since information plays an essential role both for granting loans and file follow-up, the agency theory applies especially to financial institutions.

In fact, because of information asymmetry, financial institution managers have, to some extent, the opportunity to over- or understate their loan loss estimation (LLE) depending on what they think maximizes their utility. Commercial Loan Officers (CLO) are also in a position of information asymmetry vis-à-vis the financial institution with respect to their lending and follow-up activities. This is particularly true for the identification of businesses in financial difficulty (BFD), and the estimation of the loan loss that could be incurred. Consequently, they can also engage in the management of accounting information that is relevant to various aspects of their work.

The paper will develop as follows. First, the motivating forces behind LLE management by managers will be examined. The focus will then shift to the work of the CLOs. To do so, the aspects of the effort allocation between the activities related to business development and file follow-up will be explored, as well as specific considerations relating to their horizon, and the management of information dealing with files involving BFDs.

II. LLE managment by financial institution managers

One factor that can motivate managers to engage in accounting information management is their compensation system, which is incentive-based. That system depends on accounting information, which is compiled according to recognized standards and is auditable by independent third parties (Coulombe and Tondeur, 2001). Despite these characteristics that increase reliability, some discretionary leeway still exists, which leaves room for management of accounting information. The literature on accounting information management in financial institutions highlights two main schemes used by managers to "window-dress" the financial statements to fit the image they wish to project. The first scheme involves the management of loan loss estimation, the second, the materialization of gains or losses on securities (Beaver and Engel, 1996). From the perspective of accounting information management, Scott (2006) identifies four strategies that can be adopted by a manager according to the motivations that drive him: 1) taking a bath (strong income minimization), 2) income minimization, 3) income maximization and 4) income smoothing.

One reference paper dealing with managers' incentive pay calculated directly on the basis of accounting income that is often quoted is that of Healy (1985). His results directly corroborate the earnings minimization hypothesis when managers are unable to meet the threshold performance level. By doing so, they create provisions that they will be able to reverse in subsequent fiscal years. Also, Healy's (1985) results back up the earnings maximization hypothesis when managers are in the increasing premium zone. Applied in a banking context, these results would indicate that managers have a tendency to understate their loan loss estimation when they fall into their increasing premium zone and to overstate it otherwise. A limit to this behaviour would be the manager's fear of losing his job. In fact, Defond and Park (1997) found that managers who fell below their performance threshold level did not minimize the current year's earnings. Instead they chose to adopt a maximization strategy more in line with income smoothing and keeping their position in the organization.

In accordance with the big bath hypothesis, new management teams tend to significantly reduce earnings and then blame the situation on decisions made by the previous administration (DeAngelo (1988). By that logic, a new manager of a financial institution could overstate the LLE in order to free his administration of any responsibility pertaining to loans that were granted before he was hired.

On the other hand, the income smoothing hypothesis aims at maximizing the long-term value of a company. Indeed, reducing the volatility of earnings decreases the risk perceived by investors. As a result, investors would discount anticipated cash flows at a lesser rate in order to establish the stocks' value (Fields and al., 2001). McNichols and Wilson (1988) have demonstrated that managers use, among other things, loan loss estimation to implement an income smoothing strategy.