Global banks have improved their capital ratios in part by understating the riskiness of their assets, not by raising their ability to stem losses, the Bank for International Settlements said.
Regulators need to monitor the use of internal risk models in determining the capital lenders hold against losses and complement them with gauges that don’t use risk weightings, the BIS said in its annual report released today. The BIS, based in Basel, Switzerland and owned by 60 central banks, hosts the Basel Committee on Banking Supervision, a group of regulators and central bankers that sets global capital standards.
“Market commentary has suggested that much of this trend reflects banks’ optimization of risk-weighted assets -- the redesign of transactions in order to lower capital requirements -- rather than a genuine increase in loss absorption capacity,” the BIS said. “Such window-dressing raises questions about the use of internal risk assessments for the determination of regulatory capital requirements.”
The BIS adds to a chorus of international regulators that are increasingly looking at leverage, in addition to capital measures based on risk weightings assigned to different assets, to gauge banks’ financial strength. Their focus intensified as some banks improved capital ratios following the financial crisis by altering internal models or cutting risk-weighted assets without correspondingly shrinking their balance sheets.
UBS AG and Credit Suisse Group AG, the largest Swiss banks, were told last week to increase their ratios of capital to total assets. Danske Bank A/S, Denmark’s biggest lender, was told by the country’s financial watchdog it had underestimated risky assets. U.S. regulators are considering doubling the ratio of capital to 6 percent of total assets, according to four people with knowledge of the talks who asked not to be named.
Lenders in the U.S. and in more than 100 other countries already comply with Basel’s risk-based capital rules, which assign weightings to assets based on their riskiness. Corporate debt is given a heavier weight than government bonds, requiring that more capital be held against it. The weighting affects the profitability of trading and investing in those assets.
While firms submit their models to national regulators for validation, they don’t have to disclose them publicly. Surveys by the Basel Committee have shown that risk-weightings for the same assets vary among banks, undermining their credibility.
“The range of variation indicates that the interaction of risk-sensitive rules with the complexity of risk modeling has created a wide scope for inconsistency, which can seriously weaken both the credibility and the effectiveness of the framework,” the BIS said.
The BIS’s analysis based on samples of large globally active banks shows that European lenders’ risk-weighted assets were at 28 percent of total assets at the end of last year, a value that has changed little over the past eight years. The ratio is at 58 percent for U.S. banks, which have reduced it from a peak of 77 percent in 2008. Japanese and U.K. banks have also reduced assets by lowering risk weights over the last five years.
Part of the difference is due to accounting methods. U.S. rules allow the netting out of derivative positions, while the international standard used by European banks doesn’t and hence inflates total assets of banks such as Deutsche Bank AG or BNP Paribas SA compared to U.S. peers like JPMorgan Chase & Co. Different business models also matter.
Still, even within the same jurisdiction, risk weights can vary widely. There are several factors driving the variation, some of which can be desirable as it allows banks to compete and the market to “balance the views of bulls and bears,” the BIS said. Distortions come from different parameters used by the lenders, from statistical “noise” and crucially, from bankers’ optimistic bias.
“Differences due to strategic choices by banks are unwelcome because they undermine regulatory efforts,” the BIS said. “Management interventions in the models skew risk assessments downwards by understating potential loss.”
To be sure, the BIS advises against throwing out the risk weighting entirely in favor of relying solely on the leverage ratio, which compares capital against total, unweighted assets. Having only that gauge to measure a bank’s resilience would also create false incentives for banks that could diminish, rather than strengthen stability.
“The leverage ratio does not address the problem of incentives,” it said. “For a given ratio, banks seeking to minimize regulatory capital can simply reallocate their portfolios toward riskier activities, or shrink their balance sheet without necessarily reducing their potential losses.”
Instead of just relying on one of the two ratios, regulators should take advantage of the fact that “it is difficult to manipulate one without affecting the other, typically in the opposite direction,” the BIS said. The right response for regulators is “harnessing the complementary strengths” while improving the risk weights by more oversight.
Regulators should address the unwanted variation of weightings by more transparency into the models’ inner workings, as well as by standardization and less discretion for banks in setting their parameters, which would also help to prevent “statistical noise,” the BIS said.