Nearly a third of the tightening by regulated banks was offset by the increased lending of foreign-regulated banks.
Under proposed global banking rules known as Basel III, regulators will vary banks' capital requirements over time to try to smooth the credit cycle. When lending grows too sharply, regulators can raise the amount of capital banks must hold in reserve. When it's too anemic, they can lower the capital requirements. But to be effective, such macro-prudential capital regulation must affect the loan supply of regulated banks - and not be completely offset by unregulated sources of credit.
Using bank-specific data from the United Kingdom, Shekhar Aiyar, Charles Calomiris, and Tomasz Wieladek find that capital requirements can be a powerful tool for regulating credit, but that unregulated sources of offset significantly complicate the implementation of macro-prudential policy. In Does Macro-Pru Leak? Evidence from a U.K. Policy Experiment (NBER Working Paper No. 17822), they find that UK-regulated banks did indeed lend less when regulators increased their capital requirements, but foreign-regulated institutions operating in the United Kingdom boosted lending during such times, a "leakage" that substantially reduced the impact of the policy.
"The problem of 'leakages' involving local intermediaries is particularly acute for an economy like the [United Kingdom], which is a global financial centre," write the authors. British capital requirement rules don't apply to resident foreign bank branches located in-country but with headquarters abroad. The result: nearly a third of the tightening by regulated banks was offset by the increased lending of foreign-regulated banks.
From 1997 to 2007, the UK's Financial Services Authority (FSA) used "trigger ratios" to vary a bank's minimum requirements for risk-based capital. Depending on regulators' perceptions of operational and interest rate risk (which are correlated with bank size, reliance on retail deposits, and sector concentration of its loans), the capital requirement could vary anywhere from a minimum 8 percent to 23 percent.
By looking at 104 regulated banks and 172 foreign-regulated branches in the United Kingdom, the authors find that when UK regulators increased the capital requirements for regulated banks, the growth in their lending decreased. For each single percentage point increase in a bank's minimum capital requirements, the growth in its lending fell an average 6.5 to 7.2 percentage points.
Among foreign-regulated branches, however, the story was different. For every one percent reduction in lending by UK-regulated lenders, the foreign-regulated entities increased their lending by 2.67 percent. Although these foreign-regulated entities, which are branches of foreign banks, on average are about 1/15th the size of a UK-regulated bank, they are more numerous than the UK-regulated banks. Thus, the cumulative impact of the increased lending by these entities was substantial, amounting to roughly 30 percent of the reduction in lending by UK-regulated banks. The authors view this estimate as a lower bound for the actual leakage, since more expensive methods of loaning money, through cross-border lending or through the capital markets, could also fill the lending gaps.
Basel III could well eliminate much of this leakage, since it includes reciprocal requirements between financial regulators aimed at eliminating such distortions between banks regulated by different authorities. "[T]he effect of capital requirements on aggregate lending may become stronger once the reciprocity agreement embedded in Basel III becomes enforced and the branch leakage documented in this paper is eliminated," the authors conclude.