Jumat, 23 Juni 2017

Index funds and voting shares

Todd Henderson and Dorothy Shapiro Lund have an interesting OpEd in the Wall Street Journal, "Index funds are great for investors, risky for corporate governance." In brief, index funds don't participate heavily in monitoring companies, finding information about companies, or corporate control contests.

This point echoes larger complaints that with the spread of index funds there won't be enough active money to make markets efficient, and especially to make efficient the market for corporate control. One of the most important functions of a public market is, if you think that a company is mismanaged, you can buy up a lot of shares, vote out the management, and run it better. This is an imperfect system, to be sure, but note how many nonprofits (universities) and privately held companies, immune from this pressure, are run even more inefficiently than public companies.

Todd and Dorothy, law professors, after very nicely reviewing how funds currently deal with voting issues, seem to favor more law.
So how can the law ensure that these institutions make informed decisions about corporate governance? ... The first is to encourage them to rely on third-party corporate governance experts. It may be necessary... for the law to create incentives for institutional investors ...option three: encouraging passive institutional investors to abstain from voting altogether.   
Hmmm. When "the law," not a person or people, is the subject of a sentence, I get cautious. When the law wants "to encourage" people, my hackles rise.  The law "encourages" and "creates incentives" pretty bluntly. One example, though discarded, is a bit chilling,
 This could be accomplished by providing a legal cause of action to shareholders that are harmed by uninformed or conflicted voting decisions. But this would be a blunt tool for curbing abuse. 
Indeed it would.

But this is forgiveable. They are lawyers, so more law is the answer. We are economists, and law a necessary evil when contracts and markets fail. Is there not an economic solution, a Coasean way to slice the knot?

I think so. Companies should issue, and index funds should want to buy, non-voting shares.  Non-voting shares seem to be regarded as a little infamy of internet companies, used to keep control in the hands of founders. But a split between voting and non-voting shares seems ideally suited to a mass of indexing investors, and a few active, information-based traders and active corporate control investors. In this vision, most of those voting shares are in public hands, unlike the internet companies.  In fact, most corporate stock grants and options to insiders should be in the form of non-voting shares.

Non-voting shares are treated exactly the same for all cash flow purposes. They receive the same dividends, same rights in repurchase, same treatment in any reorganization. They just do not allow the right to vote.

Since index funds don't value the option to vote, they should want non-voting shares.

Would such shares trade at a discount? Yes, likely so. And that's a benefit, not a cost, a feature not a bug. Index funds could buy the same cash flow, which is what they want, cheaper, by giving up the value of their votes, which they're not interested in. Buying the same cashflow cheaper gives you a better return.

Todd  and Dorothy actually advocate a form of this idea, that the index funds voluntarily refuse to vote.  But then the index funds pay the cost of an option they do not use. By purchasing non-voting shares they do the same thing, and reap a financial reward.

This separation between voting and non-voting shares would make the market for corporate control more efficient. It is easier for someone who wants to buy the voting rights to buy them from other active investors than from passive mutual funds. It also separates the stock market price into guesses about cash flows and guesses about corporate control events. As a long-term investor I'm interested in the former and less in the latter.

Non-voting shares become a sort of state-contingent long term debt. Rather than guarantee payment by its fixed value, as in debt, payment is guaranteed by its equality with whatever other shareholders are paid, and similar rights in court as debt-holders have to enforce their right to be paid ahead of voting shareholders.

I've asked a few of my corporate finance colleagues about this idea, and their general reaction is that it won't work, because sooner or later the investors with voting shares find ways to screw the non-voting shares out of money, not just out of votes. The ability to vote is the ultimate guarantor of payment.

I'm still not totally convinced.  (I admit I did not follow all of the shenanigans they suggested to accomplish stealing money from voting shares.) If we're bringing in law here, and if we're designing a security, it seems not impossible to create a class of non-voting shares whose equal treatment in all cashflow related events is the same as those of voting shares, and who have strong rights to sue to guarantee those rights. Long-term debt works, after all, as the voting shares don't find a way to escape interest and principal payments. The contract design for that right seems easier than the legal means to "encourage" behavior that Todd and Dorothy imagine.

Update: See Todd and Dororthy's response and more discussion.