The Review of the UK listing regime has previously been commented on by Professor Paul Davies in a very interesting OBLB post, summarising the recommendations of the Review and pointing out two of the proposed reforms, those relating to SPACs and dual-class shares, as likely to be particularly controversial. In a recent article, we have analysed the latter proposal.
In the Review, it is suggested that the LSE should allow dual-class firms to list on the Premium segment, subject to a number of conditions. In short, these are the following:
- The dual-class structure may only be in place for 5 years (a time-based sunset);
- Multiple-voting shares must convert to single-vote shares on transfer (a transfer-based sunset);
- Only directors may hold the multiple-voting shares and the extra votes they confer may only be used to let the voter stay in office and to block takeovers, and finally;
- The difference in voting rights between multiple-voting and single vote share may not be greater than 20:1.
The purpose of these conditions is to ensure ‘high corporate governance standards’. In our article, we discuss these conditions based on available empirical studies on the effects of dual-class shares on firm performance. We also draw on the Swedish experience of dual-class listings and regulation since dual-class structures have been more common than not in listed Swedish companies for a very long time.
We start by addressing the main arguments that are usually presented against dual-class listings, which we also assume have influenced the conditions recommended in the Review. The arguments analysed are that:
- Dual-class structures damage company value;
- Dual-class structures undermine the market for corporate control (entrenchment);
- Dual-class structures lead to increased agency costs, and;
- Dual-class structures make it hard to hold managers accountable.
Based on a literature overview of some 90 empirical studies on the effects of dual-class shares in jurisdictions comparable to the UK (including the US and many European countries, including Sweden), we find that the empirical evidence 1) is clearly ambiguous with regards to the effects of dual-class structures on company value; 2) does not support that dual-class structures have negative effects on takeover frequency or the market for corporate control; 3) does not support higher agency costs in dual-class firms than in ‘one share, one vote’ firms, and; 4) does not support that dual-class structures make it harder to hold managers accountable.
Based on these findings, we argue that several of the listing conditions in the Review are uncalled for, and in some cases might even lead to lower corporate governance standards. The mandatory time-based sunset clause seems unjustified since there is no clear empirical evidence of governance issues that time-based sunset clauses could potentially remedy, while they would limit the possible upsides of allowing dual-class structures. The transfer-based sunset clause along with the requirement that dual-class shares may only be held by directors, and the limitation on the use of the extra votes are equally unmotivated when the empirical evidence is considered, and seem designed to lead to entrenchment of multiple-vote shareholders.
Based on the Swedish experience of regulating dual-class structures, we do however argue that the recommendation to limit the maximum difference in voting rights allowed to a ratio of 20:1 is well founded. The Swedish regulation on dual-class structure is not a matter for the stock market regulator to address, but is instead handled in the company law rules on minority shareholder protection. The Swedish Companies Act provides strong minority shareholder protection which addresses the conflict that can arise between controlling and non-controlling shareholders regardless of share structure, and therefore mitigates the negative corporate governance effects that dual-class structures could give rise to. Perhaps most importantly, the law requires that matters where agency problems can be expected to arise (such as directed share issuances, charter amendments, mergers and buy-backs) must be decided on by the general meeting by qualified majority vote, where both the votes cast, and the voting shares represented at the general meeting are to be counted. The regulation also includes a very strong principle of equal treatment and several specific minority shareholder rights. The latter include rights to demand appointment of a minority accountant, to demand minimum dividends, and to have any (legal) issue tried at the general meeting as well as to require an extraordinary general meeting to be held. Most of this regulation does however presuppose that the control of the company is not entirely concentrated in the hands of one or a few shareholders for it to be efficient. In this case, a maximum voting difference will effectively prohibit raising capital on the stock market, unless the controller can afford to hold on to a significant part of the equity of the company (in Sweden. Swedish law therefore limits the maximum voting difference between share classes to 10:1, as compared to the ratio of 20:1 suggested in the Review.
Erik Lidman is a Senior Lecturer at the Department of Law, University of Gothenburg.
Rolf Skog is Professor at the Department of Law, University of Gothenburg.