Regulation and systematic risk: the case of the water industry in England and Wales

Abstract

This article examines whether regulatory announcements relating to competition, pricing policy and the quality of service in the water industry in England and Wales have an impact on the industry's systematic risk. The results indicate that, although the long-run overall effect of regulation on systematic risk appears to be negligible, competition and quality announcements can have a significant impact upon the industry's systematic risk. Further analysis of individual announcements and companies suggests that price announcements also significantly affect individual companies’ systematic risk, although not in a uniform manner. This indicates that regulators need to be aware that, while some of their policy decisions tend to affect the industry as a whole, others have a diverse impact on individual companies.

Acknowledgements

I am grateful to Tony Antoniou, Phil Holmes, Gregory Koutmos, Simon Parker, Krishna Paudyal, Georage Philippatos and an anonymous referee for their helpful and constructive comments on earlier drafts of this article.

Notes

¹ For the interest of brevity, henceforth this article will refer to the water industry, although at the time of privatization each constituent company operated in both the water and sewerage sectors.

² For example, the capital expenditure to sales ratio for the different UK utility industries in 1995 was estimated to be: water industry 36%, telecommunications 17%, gas 15% and electricity distribution 8% (Warburg (1996) ‘Utilities Group Modeller’).

³ In order to examine whether the systematic risk of the 10 water utilities and the industry portfolio is stable over time, their βs were estimated recursively. Results confirmed that the systematic risk was not constant for the period under investigation. Over the sample period, the β coefficient of individual companies fluctuates on average between 0.4 and 1. Detailed results are available on request.

⁴ For a detailed discussion of the price-cap regulation used in the UK, see Beesley and Littlechild (1989).

⁵ The Government wrote off the industry's debts of £4.95 billion before privatization and injected an additional £1.5 billion cash towards the costs of the investment program.

⁶ At the time of privatization, K factors ranged between 5.5 for Anglian Water to 0 for Southern Water, with an average of 3.8.

⁷ See Brennan and Schwartz (1982) for a theoretical treatment of this problem, which leads to the dynamics of explicit versus implicit functions. From an empirical point of view, it should be recognized that the existence of this implicit relationship in regulation, together with the relatively complex regulatory framework and the youth of the industry, suggests that the findings from any investigation of this nature should always be interpreted with caution.

⁸ For a more detailed discussion of the effects of regulation on the firm's cash flow and risk see Joskow and MacAvoy (1975), Keran (1976), Peltzman (1976) and Shaffer (1984). More specifically, Binder and Norton (1999) show that the systematic risk of a firm is inversely related to profits, and thus regulation can affect β by altering: (i) the covariability of the returns on equity of the regulated firm with the market; (ii) firm or industry specific factors and (iii) the buffering effects of regulation.

⁹ This is not uncommon in regulatory proceedings, as for example was the case with the US S&L industry in 1987.

¹⁰ The sample used ends in August 1995, because after this period mergers between water companies started to take place.

¹¹ The Appendix describes samples of announcements in each group. A full list of announcements is available on request from the author.

¹² Given the differences in size of the 10 WaSCs, an equally weighted, rather than a value weighted, portfolio is chosen to gauge the effect of announcements on the industry as a whole, avoiding the problem of a small number of larger companies driving the results. Particularly since these companies were local monopolies, and the regulator was mainly concerned with the ability of each firm to finance the much needed investment, it is important to gauge the effects of regulation across the industry, without bias towards larger companies.

¹³ Note that and E(Cov(R_it , R_mt )) = E(ε_it , ε_mt ).

¹⁴ Specification tests and results from the multivariate GARCH system are available on request.

¹⁵ Since groups of announcements may result as having no significant impact on β as a consequence of an averaging out effect within the announcement group, the impact of individual regulatory announcements was also investigated. These results will be reported later wherever relevant.

¹⁶ It is desirable that the window should be small to avoid contamination from other events (Binder, 1985a). This is particularly acceptable in a developed and active stock market such as the London market. A 5-day window was however tested, but the results were found to be similar.

¹⁷ These hypotheses were formulated on the assumption that the main concern of OFWAT is to protect consumers and that there is no regulatory capture.

¹⁸ However, when estimating the effects of individual announcements on industry's systematic risk, a large proportion of announcements within the insignificant price groups were found to be statistically significant. As anticipated, the overall insignificant effect resulted from the averaging effect of individual announcements within the same group being of opposite sign. These results are available on request from the author.

¹⁹ According to the Averch–Johnson effect (Averch and Johnson, 1962), regulation that limits the rate of return on capital creates an incentives for regulated firms to over-invest in tangible assets, as capital becomes relatively ‘cheaper’ than labour.

²⁰ The detailed results of the pair-wise Wald tests for equality of coefficients are available on request.

²¹ The methodology adopted in this article cannot test for the source of these cross-sectional differences.

²² Although the 10 privatized water and sewerage companies there appears to be homogeneous by function and regulatory edict. It should be recalled here that some of these companies have stand out characteristics, such as different K factors.