Models, Supervision Determine Banks’ Risk Weights – Report
Investors have a hard time comparing the riskiness of the major global banks because there are differences in how each bank calculates the potential danger of their assets, according to a report by the Basel Committee on Banking Supervision.
Based on tests of how 15 major banks assign risks to a simple, hypothetical portfolio of financial instruments, the Basel Committee found differences, either due to supervisory decisions or due to the in-house models that banks use to calculate risk.
“While some variation in risk weightings should be expected, excessive variation arising from bank modelling choices is undesirable when it does not reflect actual risk-taking,” said Stefan Ingves, Chairman of the Basel Committee and governor of Sveriges Riksbank.
The Swiss-based Basel Committee, which includes banking supervisors from almost 30 countries, sets global standards and has been tightening its rules in recent years in an effort to prevent another global financial crises.
The Committee’s analysis of how banks assess the risks from financial instruments is important because the global financial crises in 2007-2009 was largely triggered by major losses on banks’ investments in housing related securities that were held in their trading books.
Banks assign risks to their investments, known as risk weightings, and the riskier the investment, the more capital banks have to set aside – the capital ratio – in case the investment turns sour.
But the financial crises showed that banks completely underestimated the risk of the mortgage-backed securities on their books and banking supervisors have now tightened their rules, telling banks to set aside more and higher quality capital.
The review of banks’ risk weighting is part of the Basel Committee’s fundamental review of banks’ trading book with the aim to ensure that the new Basel III rules, which are currently being phased in, are applied consistently across the world.
A similar study of risk weightings in banks’ banking book by the Basel Committee is also underway.
Banks differentiate between trading and banking books with their trading books holding securities and instruments that are used in trading, either for the bank’s own profit or on behalf of its customers. Banks value those securities based on market prices.
The banking book typically comprises securities that are not actively traded but held to maturity and therefore accounted for differently.
“The analysis used to compile this report provides national supervisors with a much clearer understanding of how the risk models of their banks compare with those of international peers. This will allow national supervisors to take action where needed,” the Basel Committee’s Ingves said.
In their review, the banking supervisors found that differences in banks’ trading positions were reflected in risk weightings but it was difficult for investors “to assess how much of the variation reflects differing levels of actual risk and how much is a result of other factors.”
One reason for the variation in risk weightings was due to decisions by local banking supervisors that were applied to all banks in one country or to individual banks.
An example of such a difference is a restriction by supervisors on banks assigning varying risks to different types of assets, a factor that accounted for around one-fourth of the total variation in the hypothetical portfolio used in the review.
These local decisions typically resulted in higher capital requirements and differences in risk weightings across jurisdictions, the Basel Committee said.
Another important reason for different risk weightings was the models used by banks.
“The exercise found that a small number of key modelling choices are the main drivers of the remaining model-driven variability,” the review said.
Banking rules allow for some flexibility in how banks measure risks so the Committee was not surprised to find some variation, and the aim of the study was not determine optimal variation.
Using the hypothetical portfolio of securities in a test was a way for supervisors get a better understanding of what elements in banks’ models lead to different risk weightings.