12 December 2016

Actually Existing Capitalism

Something that's cropped up a few times with recent discussion of neocameralism as a concept is the role of shareholders in existing firms.

Conflicts of interest between principals and agents are one of the most significant forces acting on the structure of any kind of organisation, so it is essential when discussing how to apply structures from one kind of organisation to another, to have a feel of how the conflicts are playing out in existing structures and organisations.

In particular, I have seen more than one person on twitter put forward the idea that present-day joint-stock companies totally fail to resolve the conflict of interest between shareholders and managers, with the result that shareholders are powerless and managers run companies purely in their own interest:

In discussion of this piece by Ron Carrier from November 24th the author said on twitter,

"Because they are non-contractual, shares are a useful way of financing a company without ceding control.... Contrary to shareholder theory, power in the corporation is actually located in mgmt. and the board of directors."

More recently (December 9th), Alrenous followed the same path: from the suggestion that dividend payments from public companies are in aggregate very low, he draws the conclusion that stocks are "worthless" and that those who buy them are effectively just giving their money away for managers to do what they want with.

I'm sure Alrenous understands that the theory is that a profitable company can be delivering value to shareholders by reinvesting its profits and becoming a more valuable company, capable of returning larger amounts of cash in future. And of course I understand that just because someone believes that a company has become more valuable in consequence of reinvested profits, doesn't mean it is necessarily true.

Discussions like this among people not involved with investment professionally carry a risk of being based on factoids or rumour. In particular, mainstream journalists are fantastically ignorant of the whole subject. But in the end everything to do with public companies is actually public, if you can find the information and not misunderstand it. (Note that I am not including myself among the professionals, though I've worked with them in the past in an IT role).

At any rate, here is a publication dealing with aggregate dividends across the NY stock exchange. factset.com

"Aggregate quarterly dividends for the S&P 500 amounted to $105.8 billion in the second quarter, which represented a 0.8% increase year-over-year. The dividend total in Q2 marked the second largest quarterly dividend amount in at least ten years (after Q1 2016). The total dividend payout for the trailing twelve months ending in Q2 amounted to $427.5 billion, which was a 7.1% increase from the same time period a year ago."

So, that's getting on for half a trillion dollars in dividends paid out by the S&P 500 over the last year. Throwing numbers around without any indication of scale is another media trope, but that's about 2-3% of US GDP, which seems like the right sort of scale.

As an aside, if some of these companies hold shares in others, the dividends are effectively double-counted: one company in the set is paying out to another, which may or may not then be paying out to its shareholders. I would assume this is not more than a few percent of the totaleven investment companies like Berkshire Hathaway are likely to invest more in private companies than other S&P 500 membersbut it's an indication of the pitfalls available in this sort of analysis.

In addition to dividends, as I pointed out, share buybackswhere a company purchases its own shares on the open marketare economically equivalent to dividends: the company is giving cash to its own shareholders. If every shareholder sells an equal proportion of their holdings back to the company, then the result is that each shareholder continues to hold the same fraction of the company's outstanding shares, and each has been paid cash by the company. Of course, some will sell and some not, but the aggregate effect is the same. The choice of whether to take cash by selling a proportion of one's holding, or whether to simply hold shares, thereby effectively increasing one's holding as a fraction of the company, enables shareholders to minimise their tax liability more efficiently, which is apparently why share buybacks have become more significant compared to dividends.

Alrenous found this article from Reuters, which says "In the most recent reporting year, share purchases reached a record $520 billion.". That's not the same period as the one I found for aggregate dividends, so adding them together might be a bit off, but it looks like we can roughly double that 3% of GDP. As I said on twitter, as a general rule, large companies are making profits and paying shareholders.

The reason neocameralism makes sense is that joint-stock companies basically work.

That is not to suggest that the principal-agent conflicts are insignificant. They are always significant, and managing the problem is a large part of any organisational practice. That is what the bulk of corporate law is there to deal with.

I picked up a recent article in Investor's Chronicle in which Chris Dillow suggests that management is simply overpaid:

"...bosses plunder directly from shareholders by extracting big wages for themselves. The High Pay Centre estimates that CEOs are now paid 150 times the salary of the average worker, a ratio that has tripled since the 1990s - an increase which, it says, can't be justified by increased management efficiency."

However, Dillow also links other source with other suggestions: the 1989 Harvard Business Review article by Michael Jensen is particularly fascinating.

Jensen claims that regulation brought in after the Great Depression had the effect of limiting the control of shareholders over management:

"These laws and regulations—including the Glass-Steagall Banking Act of 1933, the Securities Act of 1933, the Securities Exchange Act of 1934, the Chandler Bankruptcy Revision Act of 1938, and the Investment Company Act of 1940—may have once had their place. But they also created an intricate web of restrictions on company 'insiders' (corporate officers, directors, or investors with more than a 10% ownership interest), restrictions on bank involvement in corporate reorganizations, court precedents, and business practices that raised the cost of being an active investor. Their long-term effect has been to insulate management from effective monitoring and to set the stage for the eclipse of the public corporation.

"...The absence of effective monitoring led to such large inefficiencies that the new generation of active investors arose to recapture the lost value. These investors overcome the costs of the outmoded legal constraints by purchasing entire companies—and using debt and high equity ownership to force effective self-monitoring."

A quarter of a century on from Jensen's paper, the leveraged buyout looks not so much like an alternative form of organisation for a business, but rather an extra control mechanism available to shareholders of a public joint-stock company. The aim of of a buyout today is, as Jensen describes, to replace inefficient management and change the firm's strategy, but today there is normally an exit strategy: the plan is that having done those things the company will be refloated with new management and a new strategy.

The "Leveraged" of LBO obviously refers to debt: that takes us to the question of debt-to-equity ratio. A firm needs capital: it can raise that from shareholders or from lenders. If all its capital is shareholders', that limits the rate of profit it can offer them: the shares become less volatile. If the firm raises some of its capital needs from lenders, the shares become riskier but potentially more profitable.

Under the theory of the Capital Asset Pricing Model (CAPM), the choice is arbitrary: leverage can be applied by the shareholders just as by the company itself. Buying shares on margin of a company without debt is equivalent to buying shares of a leveraged company for cash. However, this equivalency is disrupted by transaction costs, and also by tax law.

There is considerable demand in the market for safe fixed-income investments. A large profitable company is exceptionally well-placed to meet that demand by issuing bonds or borrowing from banks, and therefore can probably do so much more efficiently than its shareholders would be able to individually, were it to hold its cash and leave shareholders to borrow against the more expensive shares.

The transaction costs the other way, the ones caused by corporate indebtedness, come through bankruptcy. Bankruptcy is essential to capitalism, but it involves a lot of expensive lawyers, and can be disruptive. For an extreme example, see the Hanjin Shipping case in September. It's clearly in the interest of the owners of the cargo to get the cargo unloaded, but the international complications of the bankruptcy of the shipping line means that it's unclear who is going to end up paying for the docking and unloading. If Hanjin had a capital structure that gave it spare cash instead of debt, all this expensive inconvenience would be avoided.

Aside from transaction costs, the argument in Jensen's paper is that the management of a company with spare cash is better able to conceal the company's activities from shareholders. In his account, once the company has been bought out and restructured with debt, any expansion in the cost base has to be directly justified to shareholders and creditors, since capital will have to be raised to pay for it. This improvement in the monitoring of the management is part of what produces the increased value (in his 1980s figures, the average LBO price was 50% above the previous market value).

A quarter of a century later, we frequently read the opposite criticism, that pressure from investors makes management too focused on short-term share prices, which is a bad thing. I linked this article by Lynn Stout, and while I think the argument is very badly stated, it is not entirely wrong. The problem in my opinion is not with the idea of managing in order to maximise shareholder value: that is absolutely how a company should be managed. The problem is with equating shareholder value to the price at which a share of the company was most recently traded. Though that is most probably the best measure we have of the value of the company to its shareholders, it is, nonetheless, not a very accurate measure. Given that the markets have a relatively restricted view of the state of the company, maximising the short-term share price relies on optimising those variables which are exposed to view: chiefly the quarterly earnings.

If outside shareholders had perfect knowledge of the state of the company, then maximising the share price would be the same as maximising shareholder value. Because of the information assymetry, they are not the same. Value added to the company will not increase the share price unless it is visible to investors, and some forms of value are more visible than others. Management are certainly very concerned by the share price. As I mentioned on twitter, "in any company I worked for, management were (very properly) terrified of shareholders"

But this is a well-known problem. There are various approaches that have been tried to improve the situation. Where a company has a long-established leadership that has the confidence of investors, shareholding can be divided between classes of shares with different voting rights, so that the trusted, established leadership have control over the company without owning a majority of the equity. This is the situation with Facebook, for instance, where Mark Zuckerberg owns a majority of the voting shares, and most other shareholders hold class B or C shares with reduced or zero voting rights. Buying such shares is an act of faith in Mr Zuckerberg, more than owning shares in a more conventionally structured business. The justification is that it allows him to pursue long-term strategy without the risk of being interrupted by a takeover or by activist investors.

In fact, this year Zuckerberg increased the relative voting power of his holding, by introducing the non-voting class C shares. That has been challenged in court, and is the subject of ongoing litigation.

In summary, the arrangements of public companies consist of a set of complex compromises. There are many criticisms, but they tend to come in opposing pairs. For everyone who, like Alrenous, claims that shares are worthless because companies do not pay dividends, there are some like the Reuters article he found which complain that companies pay out all their profits and do not invest enough in growth. For everyone who, like Chris Dillow, complains that managements are undersupervised and extract funds for self-aggrandizement and private gain, there are others like Lynn Stout who complain that managements are over-constrained by short-term share price moves and unable to plan strategically.

The arrangements which implement the compromises between these failings are flexible: they change over time and adapt to circumstances. A hundred-year-old resource extraction business like Rio Tinto is not structured in exactly the same way as a web business like Facebook. The point of Chris Dillow's article is that fewer businesses are publicly traded today than in the past (though even that is difficult to measure meaningfully).

The joint-stock company is not a magic bullet, it is a range of institutional forms, evolved over time, and part of a large range of institutiontal forms that make up Actually Existing Capitalism. They are ways of coping with, rather than solving, the basic conflict-of-interest and asymmetric-information issues that are fundamental to everything from a board of directors appointing a CEO to a coder-turned-rancher hiring a farm hand.

My worry is that Moldbug's form of Neocameralism is an inflexible snapshot of one particular corporate arrangement, which only works as well as it does because it can be adapted to meet changing demands. That's why I tend to think of it as one item on a menu of management options (including hereditary monarchy!)


Alrenous said...

Representation my views not representative of my views.
Not a complaint: this is common.

Uncle Saturday said...

The human brain is an inflexible snapshot of one particular cellular arrangement, which only works as well as it does because it can be adapted to meet changing demands.

Ronald Carrier said...

Excellent essay. Here's how it seems to me.

Establishing a publicly traded corporation has two purposes. The first has to do with governance: issuing stock in a corporation is a way of resolving principal-agent problems. The second has to do with finance: issuing stock in a corporation is a way of raising capital. As stock is currently issued for financial purposes, there is a primary market, in which stock is sold to investment banks and other big investors, and a secondary market, in which stock is traded among investors.

I would say a majority of those participating in the secondary market for stocks don't much care about stocks as an instrument of governance. They tend to care more about fundamentals, or finding value, or technical performance. In the secondary market, stocks are a quasi-derivative. Those in the secondary market also tend not to have holdings large enough to make a difference at the AGM, so there's also an element of rational indifference. Those participating in the primary market tend to care more about governance, both because they want to get their investment back and because they tend to own enough of the issued stock for them to have a say that makes a difference when the AGM rolls around. But governance and finance are two different things, and concern for the one doesn't entail concern for the other.

Neocameralist discussion of the governance of government service providers should focus on the relation between the GSP and its investors in the primary market, because that's where the action is. I think preference should be given to models that don't involve a secondary market for the shares of the GSP, because secondary markets tend not to contribute to governance, good or bad. That means focusing on privately held corporations or various forms of what the late Larry Ribstein called uncorporations. And the focus should really be on governance and not finance, which means the law of business associations should be the starting point.

I also think it's worth keeping in mind that in a well-functioning market, if your business model sucks, no amount of good governance will save it. Neocameralism should focus as much on the services offered by the GSP as on the governance of the GSP.

Patri Friedman said...

A key aspect of this is not to compare the principal-agent issues of a corporation with an imagined selfless democratic bureaucrat (progs) or monarch (NRX). No such utopias exist. We should compare the performance of each entity at pleasing their stockholders and customers. And corporations do a much, much better job than governments at that. So, sure, I would prefer to invest in a corporation with many provisions in its charter to prevent manager fraud, and to live in a corporate state with strong customer prevention. But the existence of manager fraud and poor customer service don't mean corporations are not our current best, only that they are not perfect. So they will have to do.

Michael Rothblatt said...

Anomaly, would you care to comment on this essay?