The impact of guaranteed bailout assistance on bank loan overstatement

Purpose – The purpose of this paper is to highlight how the basic binomial option pricing model (BOPM) might be used by regulators to help formulate rules, prior to financial crisis, that help prevent loan overstatement by banks in emerging market economies undergoing financial crises. Design/methodology/approach – The paper draws on the theory of soft budget constraints (SBC) to construct a simple model in which banks overstate loans to minimize losses. The model is used to illustrate how guarantees of bailout assistance (BA) (to banks) by crisis stricken countries’ financial authorities may encourage banks to overstate loans and delay the implementation of IFRS for loan valuation. However, the model also illustrates how promises of BA may be depicted as binomial put options which provide banks with the option of either: reporting loan values on poor projects accurately and receiving the loans’ liquidation values; or, overstating loans and receiving the guaranteed BA. An illustration is also provided of how authorities may use this representation to help minimize bank loan overstatement in periods of financial crisis. In order to provide an illustration of how the option value of binomial assistance may evolve during a financial crisis, the model is generalized to the Mexican financial crisis of the late 1990s. During this period, Mexican authorities’ guarantees of BA to the nation’s largest banks encouraged those institutions to overstate loans and delay the implementation of (previously adopted) international “best practices” based loan valuation standards. Findings – Application of the model to the Mexican financial crisis provides evidence that, in spite of Mexico’s “official” 1997 adoption of international “best accounting practices” for banks, “iron clad” guarantees of BA by the country’s financial authorities to Mexico’s largest banks provided those institutions with an incentive to knowingly overstate loans in the late 1990s and early 2000s. Research limitations/implications – The model is compared against only one country in which the BA was directly infused into banks’ loan portfolios. Thus, as conceived, it is directly applicable to crisis countries in which the bailout took this form. However, the many quantitative variations of SBC models as well as recent studies which have applied the binomial model to other forms of bailout (e.g. direct purchases of bank shares by authorities) suggest that the model could be modified to accommodate different bailout scenarios. Practical implications – The model and application show that guaranteed BA can be viewed as a put option and that ex-ante regulatory policies based on the correct valuation of the BA as a binomial option might prevent banks from overstating loans. Social implications – Use of the binomial or similar approaches to valuing BA may help regulators to determine the level of BA that will not encourage banks to overstate the value of their loans. Originality/value – Recent research has used the BOPM to value, on an ex-post basis, the BA which appears on the balance sheet of institutions which have been rescued. However, little research has advocated the use of this type of model to help prevent, on an ex-ante basis, the overstatement of loans on poor projects.