Comparing Norwegian banks' capital ratios.

• Author: Andersen, Henrik
• Position: Report

1. Introduction

   The objective of the current capital adequacy framework (Basel II) is improved risk management and more efficient use of capital than under the previous framework (Basel I). (2) Basel II is intended to ensure that the risk that banks assign to their exposures in calculating their capital ratios better reflects actual risk than under Basel I. (3) Basel II allows banks to chose among various approaches to calculate their capital ratios. (4) For that reason, comparisons of reported capital ratios may give a misleading picture of banks' relative financial strength. The largest banks (5) calculate their capital requirements using internal risk models based on data regarding their own borrowers (internal ratings-based (IRB) approach), whereas smaller banks use the simpler, more standardised approach (the standardised approach (6)). (7) Capital requirements calculated using the IRB approach are normally assumed to reflect actual risk better than capital requirements calculated using the standardised approach. The assumptions underlying capital ratios calculated using the IRB approach often vary widely.

   Banks obtain lower risk weights for most of their exposures when they change over from the standardised to the IRB approach. (8) The Basel II framework has been calibrated to create incentives to apply the IRB approach in order to improve risk management. By itself, improved risk management will increase banks' financial strength. However, smaller banks often lack resources to develop internal models. The result is higher risk-weighted assets and thus lower reported capital ratios. Consequently, the solvency of banks that apply the standardised approach may often be underestimated compared with that of IRB banks when using report capital ratio. This makes it difficult to compare banks' financial strength on the basis of reported figures.

   It is also difficult to compare the capital adequacy of different IRB banks. Most IRB banks are still working to expand the use of internal risk models. The proportion of the portfolio covered by internal risk models varies among banks. (9) It will be somewhat easier to compare IRB banks' capital adequacy once they have all finished putting in place risk models approved in accordance with the Basel II framework.

   Even when all IRB banks have risk models that cover approximately the same segments of the portfolios there will be differences in risk weights for virtually identical assets. At the end of 2008, risk weights for residential mortgages averaged between 9% and 17% for the largest Nordic banks. Banks using the standardised approach instead of internal risk models must apply a 35% risk weight to residential mortgages with loan-to-value ratios under 80%.

   An important source of differences in internal risk models may be variations in the length of the time series used to calculate risk. Some countries permit the use of substantially shorter time series than set out in the requirement that an ideal time series should cover an entire business cycle. The reason may be that longer time series are unavailable or deemed insufficiently representative of the current risk picture. Risk weights will be substantially lower if time series do not contain data from downturns, see Andersen (2010). To calculate risk, Norwegian banks are required to use data that include the banking crisis of the early 1990s. (10)

   There is an additional problem related to banks' current transition process from Basel I to the Basel II capital adequacy framework. The transitional rules currently state that capital requirements calculated under the Basel II framework cannot be lower that 80% of what they would be under Basel I. The Ministry of Finance has decided that the transitional rules shall apply to Norwegian banks until the end of 2011. This means that a number of IRB banks continue to report regulatory capital adequacy figures that in reality depend on the Basel I framework. Definitions of capital and risk-weighted assets also differ across borders. (11) In addition, the transitional arrangements are interpreted differently from country to country. (12)

   The credit rating agency Standard & Poor's has examined the problem of the lack of comparability of banks' capital ratios and has developed its own risk-adjusted measure--the Standard & Poor's risk-adjusted capital (RAC) ratio (see Standard & Poor's (2009a)). The aim of the RAC ratio is to better enable credit rating agencies to analyse and compare banks' capital adequacy. In calculating the RAC ratio, the agency uses the same approach to calculate the capital base and risk-weighted assets of various banks. (13) Both the definition of capital and risk-weighted assets used in the RAC ratio are considerably more restrictive than Basel II. (14) On 23 November 2009 Standard & Poor's published a comparison of the RAC ratios, Tier 1 capital ratios and leverage ratios of 45 large banks (see Standard & Poor's (2009b)). Standard & Poor's concluded that Tier 1 capital ratios and leverage ratios do not give a sufficiently adequate picture of banks' capital position. Banks with identical leverage ratios or Tier 1 capital ratios had very different RAC ratios. For example, Nordic banks (15) had lower leverage ratios than US banks, but higher RAC ratios. No comparable analyses comparing the capital position of Norwegian banks have been published.

   Following the same line of reasoning, I use a uniform approach to calculate comparable capital ratios in all Norwegian banks. I utilise the advanced IRB approach and more detailed data than Standard & Poor's used in calculating risk weights for banks' assets. The analysis can provide information on how important the choice of approach is for banks' reported capital adequacy--both the choice between the standardised approach and the IRB approach and the use of various risk models under the IRB approach.

   Section 2 addresses the particular portions of the Basel II framework that are relevant to my analysis. Section 3 describes developments in Norwegian banks' reported capital ratios. Section 4 provides an overview of the data used in my analysis, and section 5 describes how I approximate exposures and risk parameters included in the calculation of banks' capital adequacy ratios. The article concludes by comparing my calculated capital ratios with the banks' reported capital ratios.

2. The Basel II framework

   The Basel II framework rests on three pillars: minimum capital requirements (Pillar 1), supervisory review (Pillar 2) and market discipline (Pillar 3). (16) This article focuses only on Pillar 1. Pillar 1 allows banks to use one of three different approaches for calculating capital requirements on the basis of credit risk: the standardised approach, the foundation IRB approach and the advanced IRB approach. Basel II also requires banks to hold capital reserves to address market risk and operational risk under Pillar 1.

   Under the IRB approach, bank portfolio exposures are categorised into six broad asset classes: corporate, sovereign, bank, retail, equity exposures, as well as purchased receivables and securitisation exposures. With the exception of exposures classified as retail, risk weights shall be calculated for each exposure within the particular class. Retail covers loans to small and medium-sized entities (SMEs) and households including residential mortgages and revolving credits. Loans to larger enterprises are included in corporate. Bank covers loans and other exposures to financial institutions. Sovereign covers loans and other exposures to government authorities.

   IRB banks must use a separate formula for calculating capital requirements for credit risk (see Appendix). The formula has been calibrated to a solvency margin of 99.9%, that is, the estimated probability that a bank's regulatory capital will not cover its losses the following year is less than 0.1[degrees]%. The formula is a function of probability of default (PD), loss given default (LGD), exposure at default (EAD) and effective maturity (M). The formula also includes parameters for maturity adjustment (b) and correlation (R) between exposures, as well as a factor for systemic risk.

   Banks using the advanced IRB approach must apply their own estimates of PD, LGD, EAD and M. These estimates must be grounded in historical experience. The Basel II framework does not specify whether more recent observations should be weighted more than observations further back in time. The historical observation period used to estimate PD must be at least five years. PD for corporate, retail and banks may never be set below 0.03%. Estimates for LGD and EAD must be based on a minimum data observation period of seven years (five years for retail) that contains at least one complete economic cycle. LGD may not be lower than the long-run...