Incurred Loss Model Research Articles

The Incurred Loss Model and Investor Pricing of Unrecognized Expected Credit Losses

Accounting for credit losses under U.S. GAAP is transitioning from an incurred to an expected loss model. The model change was motivated by concerns that reporting only incurred losses does not provide investors with sufficient and timely information about banks’ credit risk. In this paper, I develop a measure of lifetime expected credit losses using vintage analysis to examine whether stock prices reflect information about unrecognized expected credit losses in an incurred loss regime. Consistent with investors being able to obtain information about expected losses that are not recognized in the financial statements, I find that unrecognized expected credit losses are negatively associated with bank stock prices. The pricing of these losses is stronger for larger banks, consistent with lower costs of obtaining this information for banks with better information environments. I also find that recorded allowances were less than estimated expected losses, on average, consistent with concerns that implementing the expected loss model will adversely impact regulatory capital adequacy.

Adapting the Basel II advanced internalratings- based models for International Financial Reporting Standard 9

Banks around the globe are implementing International Financial Reporting Standard 9 (IFRS 9), which is a considerable effort. A key element of IFRS 9 is a forward-looking “expected loss” impairment model, which is a significant shift from the incurred-loss model. We examine how we may use advanced internal-ratings-based (A-IRB) models in the estimation of expected credit losses for IFRS 9 purposes. We highlight the necessary model adaptations required to satisfy the new accounting standard. By leveraging on the A-IRB models, banks can lessen their modeling efforts in fulfilling IFRS 9 and capture the synergy between different modeling endeavors within institutions. In outlining the proposed probability of default, loss given default and exposure at default models, we provide detailed examples of how they may be implemented on secured lending. Moreover, in discussing the issues related to the estimation of the expected credit loss for IFRS 9, we highlight the challenges involved and propose practical solutions to deal with them. For instance, we propose the use of a convexity adjustment approach to circumvent the need for assigning probabilities in multiple-scenario analysis.

The predictive ability of loan loss provisions in banks – Effects of accounting standards, enforcement and incentives

In this study we compare the predictive ability of loan loss provisions with respect to actual losses under IFRS and local GAAP. The ‘incurred loss model’ of IAS 39 is a model that requires a relatively low level of judgment by preparers compared to alternative models that exist under local GAAP. We find that loan loss provisions in IFRS bank years predict future credit losses to a lesser extent than in local GAAP bank years, consistent with the incurred loss model reducing the timeliness of provisions. We also examine the interaction of standards with enforcement of financial reporting and with preparer incentives. In testing the role of enforcement from, e.g., banking supervisory authorities, we find that the benefits of local GAAP are largely limited to high-enforcement settings. Local GAAP also performs relatively better than IFRS in large and in profitable banks. This has implications for the IASB and the FASB as they prescribe the adoption of the more judgment-based expected loss model in IFRS 9 and the corresponding US GAAP standard (ASC topic 326), as well as for supervisory authorities that will enforce these standards.

Adapting Basel's A-IRB Models for IFRS 9 Purposes

Banks around the globe are implementing IFRS 9 which is a considerable effort. A key element of IFRS 9 is a forward-looking “expected loss” impairment model, which is a significant shift from the current incurred loss model. In this paper, we examine how we may use A-IRB models in the estimation of expected credit losses for IFRS 9 purposes. We highlight the necessary model adaptations required to satisfy the new accounting standard. By leveraging on the A-IRB models, banks can lessen their modeling efforts in fulfilling IFRS 9 and capture the synergy among different modeling endeavors within the institutions. In outlining the proposed PD, LGD, and EAD models, we provide detailed examples of how they may be implemented on secured lending. Moreover, in discussing the issues related to the estimation of the expected credit loss for IFRS 9, we highlight the challenges involved and propose practical solutions to deal with them. For instance, we propose the use of a convexity adjustment approach to circumvent the need of assigning probabilities in multiple-scenario analysis.

Effect of the Basel Accord capital requirements on the loan-loss provisioning practices of Australian banks

There are two distinct regimes for bank provisioning in Australia: a forward-looking model for regulatory purposes and an incurred loss model for financial reporting. This study examines the former using a unique but confidential database. We find evidence that: (i) banks increase provisions in anticipation of future lending growth, (ii) banks allocate part of surplus capital above regulatory requirements to pre-fund future credit losses through provisions, and (iii) banks allocate part of higher earnings for the same purpose. These results suggest that bank managers use their discretion in setting provisions to dampen the impact of fluctuations in credit market conditions on their lending activities. For internal ratings-based banks, results suggest that the revised Basel framework allowing these banks to estimate expected losses using their own credit risk models may come at a cost of reduced general provisions.

Ireland and Iceland in Crisis B: Decreasing Loan Loss Provisions in Ireland

All public companies in the European Union, including Ireland’s major banks, were required to adopt IAS 39 for their annual accounting periods beginning on or after January 1, 2005. Under the “incurred loss” model of IAS 39, banks could set aside reserves for loan losses only when objective evidence existed that a loan was impaired, not in anticipation of future losses. As a result, Irish banks saw their aggregate reserve for bad loans drop from 1.2% of loan balances at the end of 2000 to only 0.4% by 2006-2007, just before the collapse of the banking industry caused loan losses to soar. In the aftermath of the global financial crisis, financial regulators and accounting bodies recognized the weakness of the pro-cyclical incurred loss model. As a result, they have proposed alternative “expected loss” models that allow reserves for expected losses to be built up over the life of a loan in a counter-cyclical fashion.

The Emergence of the ‘Incurred-Loss’ Model for Credit Losses in IAS 39

Following the financial crisis, the view became widespread that International Financial Reporting Standards (IFRS), because it is based on a so-called incurred-loss approach, led to significant overstatements of financial assets by placing tight restrictions on the recognition of loan losses. As a result, the International Accounting Standards Board (IASB) undertook a project to introduce an alternative expected-loss model in its standards, which would allow earlier recognition of loan losses. This paper is a historical study of the introduction of the incurred-loss model in International Accounting Standard (IAS) 39 between 1998 and 2003. With respect to the topic of loan losses, it argues that, especially at the beginning of that period, it was not yet common to view the issue in terms of a clear-cut dichotomy of incurred-loss versus expected-loss models, and that this had a significant complicating influence on the course of the debate. More generally, the paper illustrates some of the pressures on the quality of the Board's due process during its early years, when it attempted to complete an ambitious agenda in time for the first mass adoption of IFRS in 2005. While this paper takes no position on the correctness of the IASB's decisions as embodied in IAS 39 (2003), it does suggest that the episode covered provides justification for the considerable enhancements of its due process effected by the IASB over recent years.

Loan Loss Provisioning – A Case Study of Romanian System Smoothing Pro-cyclicality

Abstract Loan loss provisioning is an objective of both accounting standard and prudential bank regulation. Still, depending on the perspective, accounting or prudential wise, this objective can be debated between the two, at least from the pro-cyclical effect that may encumber. Local regulators, along with policy makers, argued that the incurred loss model as base of loan loss accounting should be changed in order to incorporate a more forward looking approach in setting loan loss provisions and therefore reduce pro-cyclicality. In this paper we explore the perspective of the Romanian Regulator in order to smoothen the effect of the loan loss provision changes following the transit to the full IFRS provisioning policy on the Romanian banking system.

Loan Loss Accounting Rules across the Globe: What do we Know?

A number of empirical studies in recent years have used cross-country data to analyze topics related to the procyclicality of loan loss provisions. However, surprisingly little is known about countries' underlying local GAAP provisioning rules. In this study, we extend the knowledge on loan loss accounting across the globe in three different ways: First, we shed light on the distribution of different characteristics of more than 150 local GAAP loan loss accounting regimes across the globe using data from the World Bank's Bank Regulation and Supervision Survey. Second, we use these data to develop indices that reflect if a provisioning regime follows an incurred loss model or allows banks to take account of expected losses. Third, we group countries, e.g. by development status or continental region, and examine patterns in provisioning rules across these groups. Fourth, we conduct our own survey among eleven central banks which is intended to provide a more comprehensive understanding of provisioning regimes. Generally, we find considerable heterogeneity in the structure of provisioning rules across countries as well as groups of countries. Whilst most countries require banks to classify a loan as nonperforming if payments are in arrears, almost 35% of the countries in our sample allow the recognition of nonperforming loans based on a forward-looking estimate of the PD. This finding is by far more pronounced in highly developed countries. Our findings can have important implications for the analysis of loan loss provisions in cross-country settings.

The Emergence of the 'Incurred Loss' Model for Credit Losses in IAS 39

Following the financial crisis, the view became widespread that IFRS, because it is based on a so-called incurred-loss approach, led to significant overstatements of financial assets by placing tight restrictions on the recognition of loan losses. As a result, the IASB undertook a project to introduce an alternative expected-loss model in its standards which would allow earlier recognition of loan losses. This paper is a historical study of the introduction of the incurred loss model in IAS 39 between 1998 and 2003. With respect to the topic of loan losses, it argues that, especially at the beginning of that period, it was not yet common to view the issue in terms of a clear-cut dichotomy of incurred-loss versus expected-loss models, and that this had a significant complicating influence on the course of the debate. More generally, the paper illustrates some of the pressures on the quality of the Board’s due process during its early years, when it attempted to complete an ambitious agenda in time for the first mass adoption of IFRS in 2005. While this paper takes no position on the correctness of the IASB’s decisions as embodied in IAS 39 (2003), it does suggest that the episode covered provides justification for the considerable enhancements of its due process effected by the IASB over recent years.

Judgment and Enforcement in Financial Reporting: The Case of Credit Losses in Banks

In recent years, the benefits of using professional judgment in the preparation of financial statements has been debated, for example in relation to fair value measurement. An accounting area where estimates are of particular significance is that of credit losses in the banking sector. In this paper, we evaluate the 'incurred loss model' under IFRS - a model characterized by relatively few estimates compared to alternative models. We find that only recognizing incurred losses decreases the validity of loan loss provisions and thus has a negative effect on the quality of accounting for credit losses in banks. This has implications for the IASB as it deliberates the adoption of the more principles-based 'expected loss model'. Furthermore, we test the role of enforcement. We hypothesize, and find, that enforcement increases accounting quality in the high-judgment setting, while quality is decreased by strong enforcement in the low-judgment setting. Thus, stronger enforcement can have both positive and negative effects, depending on the accounting standard being enforced.

Do delays in expected loss recognition affect banks' willingness to lend?

Banks can decrease their future capital inadequacy concerns by reducing lending. The capital crunch theory predicts that lending is particularly sensitive to regulatory capital constraints during recessions, when regulatory capital declines and external-financing frictions increase. Regulators and policy makers argue that the current loan loss provisioning rules magnify this pro-cyclicality. Exploiting variation in the delay in expected loss recognition under the current incurred loss model, we find that reductions in lending during recessionary relative to expansionary periods are lower for banks that delay less. We also find that smaller delays reduce the recessionary capital crunch effect. These results hold across management quality partitions.