Methodological thoughts on expected loss estimates for IFRS 9 impairment: hidden reserves, cyclical loss predictions and LGD backtesting

TL;DR

This paper discusses methodological approaches for IFRS 9 impairment estimates, focusing on structural conservatism, hidden reserves, cyclical loss predictions, and LGD backtesting to improve credit risk assessment.

Contribution

It introduces a framework for quantifying conservatism, separating operational effects from credit risk, and proposes a new approach for LGD monitoring and backtesting under IFRS 9.

Findings

Quantification of conservatism using iACV(c)

Framework for cyclical loss prediction and testing

LGD backtesting approach with key risk indicators

Abstract

After the release of the final accounting standards for impairment in July 2014 by the IASB, banks will face the next significant methodological challenge after Basel 2. In this paper, first methodological thoughts are presented, and ways how to approach underlying questions are proposed. It starts with a detailed discussion of the structural conservatism in the final standard. The exposure value iACV(c) (idealized Amortized Cost Value), as originally introduced in the Exposure Draft 2009 (ED 2009), will be interpreted as economic value under amortized cost accounting and provides the valuation benchmark under IFRS 9. Consequently, iACV(c) can be used to quantify conservatism (ie potential hidden reserves) in the actual implementation of the final standard and to separate operational side-effects caused by the local implementation from actual credit risk impacts. The second part continues with a quantification of expected credit losses based on Impact of Risk(c) instead of traditional cost of risk measures. An objective framework is suggested which allows for improved testing of forward looking credit risk estimates during credit cycles. This framework will prove useful to mitigate overly pro-cyclical provisioning and to reduce earnings volatility. Finally, an LGD monitoring and backtesting approach, applicable under regulatory requirements and accounting standards as well, is proposed. On basis of the NPL Dashboard, part of the Impact of Risk(c) framework, specific key risk indicators are introduced that allow for a detailed assessment of collections performance versus LGD in in NPL portfolio (bucket 3).