We talked a while back about how "corporate governance" is a thing that exists more or less orthogonal to the thing that is "running your corporation as though you were a group of competent humans," as evidenced by the fact that Citi's mangled and perhaps legally problematic semi-firing of Vikram Pandit has been celebrated as a paragon of good governance. I don't really know what "corporate governance" is, if not that, but much of its semantic space is covered by:

• do your directors and CEO like each other? - [ ] Yes [ ] No
• do you have strong takeover defenses? - [ ] Yes [ ] No

Two "No" answers = good governance; two "Yes" answers = sketchy.1

You might if you wanted to attempt to quantify those things - which is more important, and how if at all does the good governance that they reflect translate into things like shareholders making money? I enjoyed this Lucian Bebchuk DealBook post on a paper he wrote about golden parachutes in part because it gets at that a bit. Golden parachutes are a weird takeover-y topic: CEO employment contracts that provide for big payouts upon acquisition look formally like takeover defenses, insofar as they cost an acquirer money, but they're actually sort of an anti-takeover-defense. They encourage takeovers since they're a sign to acquirers that the CEO is not going to make things difficult if he gets a bid.

Anyway Bebchuk and his coauthors look at some data and find:

• Golden parachutes increase a company's 1-year chance of being acquired from ~4% to ~5.3%.
• They also increase shareholders' expected gains from being acquired, with a footnote.2
• They do not, however, increase shareholders' expected wealth generally. Parachute-having stocks underperform non-parachute stocks by ~4.35% a year.3

The paper interprets the underperformance data as being "consistent with selection," i.e. you adopt a golden parachute because your company sucks and you want to be paid when you're taken over, "and managerial slack," i.e. once you have a golden parachute, you have less incentive to build a successful independent firm; just sell it and get paid. Here's Bebchuk in DealBook:

What explains this pattern? Why do companies with golden parachutes fail to deliver for their shareholders overall even though they provide them with more benefits in the form of acquisition premiums? This pattern could be at least partly a result of the adverse effect that golden parachutes have on the incentives and performance of executives not facing an acquisition offer.

The market for corporate control benefits shareholders not just by providing the prospect of pocketing an acquisition premium but also by affecting performance more generally. Executives face the possibility that they might be ousted if they underperform. By ensuring executives of a cushy landing in the event of an acquisition, golden parachutes weaken the disciplinary force exerted by the market for corporate control.

While it makes sense that the M&A market would be more active but less motivating for executives with golden parachutes, golden parachutes are not just an abstract feature of the M&A market. They also connect to the other element of corporate governance, the thing where CEO's and board members can be friends (bad!) or not friends (good! good governance!). If the CEO gets a large golden parachute it may be a purely rational board "encouraging long-term planning and investments in firm-specific human capital"4 as well as optimizing responses to takeover bid. Or it may be because the board is filled with his buddies and they don't want him to be out on the street if the company is taken over.

Reuters has a nice piece today on corporate governance, which takes the messy and inept firings of Pandit and Carol Bartz as a jumping-off point to celebrate the rise of good corporate governance, which, again, that is strange, but they make good points:

Directors say these battles, and an increasing number of boards asserting themselves, may signal the ultimate end of the corporate board as a virtual country club for the chief executive's friends.

"There is a growing sense of responsibility by the board that they are indeed in charge of the company. They are not just there as friends of the CEO giving advice. They are, in fact, responsible for making sure that the company is on the right track," said Steve Miller, chairman of AIG and director of software maker Symantec Corp. ...

The resulting change, governance experts say, is good for investors.

GMI Ratings, a corporate governance ratings agency, reported in June that companies that split the chairman and CEO roles post five-year returns that are 28 percent higher than companies where the same person holds both jobs. GMI also found that companies with a combined chairman and CEO role pay that person nearly double the average for someone who is only CEO.

The amount of coziness between board and management that is optimal is a difficult question; there are good arguments that a lot of the governance-y things to reduce that coziness - using lots of outside directors, for instance - are not good for shareholders or corporations. These arguments are based on both empirical data about returns and fundamental philosophical views about the purpose of the corporation, though I am perhaps unusually sympathetic to them as I am very lazy and the thought of boards as stern taskmasters who keep management constantly on their toes and throw them out at the first sign of incompetence makes me stressed and tired. Vikram's ouster sapped my morale.

But there's a limit! At least some of the classic signs of coziness - the CEO as chairman of the board, the golden parachute cradling the CEO in case of ignominious fire-sale ouster - are no doubt delightful for CEO morale, but apparently less so for shareholders. Who I guess - I guess? - are more important.

Analysis: Citi board fight signals rise of the activist chairman [Reuters]
For Whom Golden Parachutes Shine [DealBook]
Bebchuk, Cohen & Wang: Golden Parachutes and the Wealth of Shareholders [SSRN]
Throwing Out Insiders Won't Fix Corporate Boards [HBR]

1. So I mean this is an idiosyncratic definition? But kind of? Like "good governance" is loose code for "how easily can an activist force major strategic changes?"?

2. The footnote being that they reduce expected acquisition premium, because basically as soon as an offer comes in the CEO is like "DONE SEE YOU SUCKERS." Bebchuk et al. find that "the presence of a GP reduces 1-week premiums by 12.8% and 4-week premiums by 19.2%, an economically significant discount." But that effect is canceled out by the higher likelihood of being acquired in the first place; "the presence of GPs is associated with an average increase in unconditional acquisition premiums of 36 basis points."

3. In their sample, which goes 1990 - 2006.

4. That's from the Bebchuk paper.