Basel III, SME lending and the grave errors of risk weighting - insightful report by global accountancy group

ACCAReport by ACCA (the Association of Chartered Certified Accountants), the global body for professional accountants with over  140,000 members, 404,000 students and more than 8,000 Approved Employers in 170 countries - published in 2012 but worth re-reading as the report correctly foreshadowed the impact on SME finance worldwide

The focus of Basel III, and indeed of all capital regulation so far, is not primarily a bank’s balance sheet but the sum of its risk-weighted assets. Capital adequacy rules assign a risk weighting to each of a bank’s assets that is meant to be proportionate to the credit and market risk that the asset in question represents.

Under Basel III, loans to SMEs are assigned a relatively high risk weighting, inherited from Basel II, which should, when combined with rising capital requirements and turbulent capital markets, result in a disproportionately high cost of capital for banks when lending to such businesses and a gradual shift of their entire business models away from SME lending. Moreover, the evidence reviewed by ACCA (2011b) suggests that capital set aside against SME loans has, in the past, significantly exceeded any losses from defaults.

Basel’s approach to risk weighting is based on a significant misconception regarding the purpose of capital regulation. Risk weightings incorporate rough estimates of credit, market and operational risk (Blundell Wignall and Atkinson 2010). It is not, however, the purpose of capital regulation to protect individual banks from bad debt, poor investments or flawed internal controls. Financial institutions have substantial risk management functions and governance arrangements charged with this task and can generally monitor their exposure better and more regularly than regulators can. Regulators should, of course, review these functions and arrangements and will no doubt find much in need of improvement. But the purpose of capital regulation is to protect the wider financial system and the taxpayer from the banks’ flawed incentives.

Capital regulation, like all regulation, is only justified insofar as it addresses market failure. In the case of financial intermediaries it has been amply demonstrated by the financial crisis of 2008–9 that negative externalities do exist and that avoiding these does justify some kind of capital regulation. Through their actions, financial intermediaries expose their counterparties to systemic risk without bearing the social cost of this by-product of their activities. The result is an excessive accumulation of systemic risk, which is the proper province of capital regulation.

This distinction means that inferring ‘optimal’ risk weights from past or forecast default rates is a deeply flawed methodology because it conflates risks that should be the target of regulation with risks that should not. Both internal risk models (whereby the banks assign their own risk weightings according to internal risk calculations) and standardised risk coefficients (provided by Basel and the implementing regulators) are based on this principle and are thus equally problematic.

To illustrate: for a bank, giving a three-year €1m loan to an SME is doubtless much riskier than buying €1m worth of newly issued three-year AAA-rated government bonds. Statistically, some of the SME creditor population must default in a given year, while the AAA-rated sovereign is certain, for all intents and purposes, to survive and pay off its creditors. Hence a narrow view of risk will rightly consider the SME to warrant a much greater capital allocation. In fact, there is more to this story than this narrow view.

Unlike the SME loan, the bond can be posted as collateral many times over in the wider financial system (2.4 times on average, based on the estimated global velocity of collateral), enabling a disproportionate volume of transactions (Singh 2011). Its price will correlate strongly with those of a host of other assets to which the bank is exposed, and can be subject to unstable feedback loops involving these (BIS 2011); and because bonds are publicly traded, its value will change constantly and may do so at any instant.

To be fair, SME loans are not free from systemic influences (Direr 2002), particularly in sectors that rely heavily on trade credit. But the two assets’ systemic footprints (to borrow from the literature on another negative externality) hardly compare.

Put another way, nearly all the risk involved in an SME loan is borne by the lender as straightforward credit risk, and most of the rest is borne by the business owners as straightforward financial risk. In the case of the bond, however, the risks are more diverse and diffuse, and in buying, holding or trading the bond the bank has internalised only a tiny amount of these. This begs the question of how risk weightings can be corrected to reflect the kinds of risk that Basel should be targeting.

Given the sheer amount of political will invested in it, a wholesale review of Basel III is most unlikely, but a recalibration within the current framework may be possible. In fact, the Basel Committee has already demonstrated one approach to this in its search for ‘global systemically important financial institutions’ (BCBS 2011). That assessment set out to identify institutions that are Too Big or Too Interconnected To Fail through a set of criteria: cross-jurisdictional activity, size, substitutability and complexity.

A similar approach can be used to calibrate the existing risk weights, by asking the following questions.

• To what extent are assets traded between financial institutions, especially internationally?

• What share of the financial institution’s assets and revenues does the asset class represent?

• What share of total activity in the asset class does this institution represent?

• How complex is the asset in question?

This process could be used to derive systemic risk weights complementary to the credit, operational and market-risk weights already employed by Basel III. This would result in weights that are not only more conducive to SME lending but also more consistent with the purpose of capital regulation.