IFRS 9 per le compagnie di assicurazione. Regolamentazione, modelli di credito e impatto sul bilancio

The aim of my thesis is to examine the effects of the new IFRS 9 impairment requirements on the insurance companies from different perspective: compliance, modeling and impact on the balance sheet. Therefore the main focus is the impact on debt instruments i.e. the great part of an insurance company asset side balance sheet. During the financial crisis, the recognition in the balance sheet of credit losses that are associated with loans and other financial assets recorded at amortized cost or fair value through other comprehensive income (OCI) was identified as a weakness in existing standards. Indeed credit losses in the IAS 39 are recognized only if the credit events have already occurred. Therefore ¿IFRS 9 Financial Instruments¿ has been developed in order to replace from 2018 the IAS 39 - that is currently in force - in order to track changes in credit risk exposure in balance sheet. The new IFRS 9 imposes a different classification and measurement of the financial assets based on either the characteristics of the product i.e. the cash flows and the business model used to manage the instrument, rather than a more rule based and elections as the IAS 39. In addition the impairment requirements switch from an incurred model to an expected loss model, as well as new hedge accounting rules are provided. Consequently, the application of the IFRS 9 will impact the way entities look to the issue of credit risk and not simply the accounting rules. Firstly, the firms have to evaluate risk embedded in their asset and then they have to measure and recognize expected credit losses in a probability-weighted and time value of the money framework, using historical and best forward-looking data. The loss allowance recognized could be the lifetime expected loss or a fraction of it i.e. the loss allowance weighted by the probability of default in the year following the evaluation date, depending on whether there has been a significant increase in credit risk on the financial instrument since initial recognition. As a result, the whole corporate system will be affected by this new view. It has to take into account best forward-looking information and expert judgment on how the changes in macroeconomic and microeconomic factors will affect their credit risk expected credit losses. The need of an implementation and management of new data processes and models is evident in order to fulfill these requirements, as the need of a stronger connection between accounting systems and risk management systems. The model chosen to fulfil the requirements is the Jarrow-Lando-Turnbull credit model that can be used- and extended to take care of economic factors- in order to comply with respect to regulation. A completely automatized R code, implemented by my own, is able to provide probabilities of default, compare the evolution in credit risk of instruments through time and deliver the loss allowance. In addition the model can separate the effect of time passing and a real movement in risk, as the regulation imposes. An empirical example is then provided for simple instruments, applying the model on historical data and showing how a company would have been affected by the regulatory innovation for the period from 2005 to 2012: the model built is able to capture market information and to track changes of credit risk through years. Finally, further extensions of the model are provided, as well as a discussion about possible matters of the standard, such as cyclicality.