THE DIRECT AND COMPLIANCE COSTS OF FINANCIAL REGULATION

This paper attempts to estimate both the direct and indirect costs of regulation for major sectors of the UK financial services industry. We also compare UK direct costs with those for the US and France and thi s provides a benchmark for assessing the effect of regulation on the c ompetitive position of the UK financial services industry. We believe that this is the first attempt to compare regulatory costs in the UK w ith those of its major competitors. For indirect costs, in the absence of an international benchmark we compare our results with the predict ions made at the time of the introduction of the Financial Services Ac t, by Lomax (Lomax, D., 1987. London Markets After the Financial Servi ces Act, Butterworths, London) and Goodhart (Goodhart, C., 1988. The c osts of regulation. In: Seldon, A. (Ed.), Financial Regulation or Over -regulation. Institute of Economic Affairs, London, p. 31). They estim ated that indirect costs would be pound 4 for every pound 1 of direct costs and that annual aggregate costs would be pound 100 million. Our results suggest that, so far as direct costs are concerned, the costs of regulation for the securities and derivatives trading and broking s ector are substantially lower for the UK than for the US and France. I n contrast, for the investment management and unit trust industry UK c osts are significantly higher than those for the other two countries. For the life insurance industry, UK costs are similar to those in Fran ce but markedly lower than those for the US. We also find for the secu rities industry around pound 4.1 of indirect costs per pound 1 of dire ct costs. For the investment management industry the corresponding fig ure is pound 3.2. However there is substantial variation across firms and, although our sample is too small to be definitive, the ratio appe ars to be related to firm size. Although these results are broadly in line with the predictions of Lomax and Goodhart it should be borne in mind that both numerator and denominator are substantially higher in r eal terms than those used by Lomax and Goodhart. (C) 1998 Elsevier Sci ence B.V. All rights reserved.