Ideoblog

Shareholders, bank pay and systemic risk

From Bloomberg via Dealbook,

The Obama administration intends to seek new powers for the Securities and Exchange Commission to force financial firms to give shareholders votes on executive pay packages, people familiar with the matter told Bloomberg News. * * * The changes aim to ensure that even financial companies that free themselves of government stakes will be subject to universal guidelines aimed at reducing systemic risks, the news service reported.

Now we see that bank pay regulation is politically motivated populist-feeding, and not really about systemic risk. The systemic risk argument for regulation assumes that this risk is an externality not taken into account by banks’ owners. If the owners care about the risks their banks pose to the financial system enough to micromanage pay to reduce these risks, then why can’t we trust the shareholders to manage the banks in other ways to contain these risks? Conversely, if we need systemic risk regulation because we can’t trust the shareholders, then why give the shareholders the power to control pay?

Another thing:  why should the SEC be the agency to dictate shareholder control over bank pay? Isn’t that agency supposed to be about disclosure?

Dismantling capitalism: executive compensation

The government thinks it ought to micromanage the compensation of firms getting government money. Here’s links to my previous thoughts on these pay restrictions.

Of course a big problem with these restrictions is that they defeat the purpose of the bailouts by discouraging financial institutions from taking or keeping the money the government thinks they need to boost the recovery. So the banks are rushing to repay.

How to solve that? Of course: regulate non-recipients too. Here’s today’s NYT on these plans.

“This is the government trying to tell the TARP banks not to worry, because everyone else’s compensation will be monitored, too,” Gustavo Dolfino, president of the WhiteRock Group, a financial recruiter, said of the industrywide principles. “We’re in a world of TARP and non-TARP.”

The endgame, I fear, is to use the financial crisis as an excuse to put into effect all of those wacky and ill-supported ideas about pay that have been kicking around for years. 

Remember the good old days when it was supposedly just all about disclosure? 

Sauce for the goose?

Executive compensation! Internal controls! Tell the shareholders everything!

On the other hand:

The House’s quarterly [expense] reports – which run over 3.000 pages apiece, across multiple volumes – are stored in a cupboard in a windowless office near a shoeshine stand. The Senate’s semiannual reports, which use type about half the size of the print in a daily newspaper, are in a building nearby. Expense entries for both chambers can be difficult to decipher, with entries and explanations sometimes cutting off mid-word.

Morgenson still silent on the NYT

As I have pointed out before (e.g.), the NYT’s Gretchen Morgenson, while loudly complaining about undemocratic corporate governance throughout corporate America (see my archive), hasn’t said anything “about the Stalinist corporate structure in her own backyard.”

And of course she’s railed even more loudly about executive pay. Nevertheless, as reported recently (HTBainbridge), her silence about her own employer continues even as the NYT pays its executives significant bonuses in the face of dismal financial results while asking employees to take pay cuts.

Surely Morgenson could provide special insight on this story. Could it be that the Times doesn’t want this story told about itself, even as it sold the story on others? Just asking.

Bebchuk vs. blindingly stupid pay rules

Harvard’s Lucian Bebchuk, perhaps the leading academic critic of executive pay, has found a regulation of executive pay he didn’t like – the stimulus bill. Here’s what he says in today’s WSJ:

Mandating that at least two-thirds of an executive's total pay be decoupled from performance, as the stimulus bill does, is a step in the wrong direction. * * *

[T]he value of some banks' common shares might largely represent an "out-of-the-money option," expected to deliver value only if things considerably improve. In such circumstances, restricted stock may provide incentives for executives to take excessive risks with the bank's survival. * * *

The bill provides executives with counterproductive and unnecessary private incentives to terminate or avoid TARP funding, even when doing so would not be in the bank's best interest.* * *

Public officials should be wary of introducing new distortions and perverse incentives. With so much hanging in the balance, ensuring that those running the country's banks have the right incentives is as important as ever.

Sounds familiar. As I noted (discussing a column by Jason Zweig) with respect to the intially proposed paycaps:

paycaps let covered executives get preferred stock that they can cash in the minute taxpayers get their money back. This encourages the managers to bet the bank – heads I win (and get paid), tails you lose (bank goes bust). Sounds like the sort of perverse, short-term incentives that got us into this mess. 

See also my other criticism of the paycaps here and here.

Academics often do not seem to understand when they propose regulatory fixes that they do not control the lawmaking process. I and many others have pointed out that whatever problems there are with executive pay are best fixed by the market than by turning regulators loose amid populist angst over high-paid executives. Professor Bebchuk, at least, is now learning about the dark side of regulating governance. I fear he may have his eyes opened further over the next few years.

Even more blindingly stupid paycaps

When the initial Treasury paycaps were announced I said "[t]his is so blindingly stupid that it would barely deserve comment but for the fact that it threatens to help further impede a desperately needed recovery." I added that "this could be a model for the future."  Here's more. 

Alas, the "future" is now.  From the WSJ:

The giant stimulus package includes a last-minute, little-noticed addition that restricts bonuses for top earners at firms receiving federal cash -- including those that already received it -- more severely than the Obama administration's previous pay limits. The most stringent pay restriction would bar any company receiving funds from paying top earners bonuses equal to more than one-third of their total annual compensation.

That bit is courtesy of Chris Dodd.  "Top earners" goes way beyond "top executives." Even worse, the banks can repay the government without raising new money -- which of course they now have a big compensation incentive to do.

"I'll bet you will see in the next month or so, banks paying back the government," said Alan M. Levine, an executive-pay attorney at Morrison Cohen LLP in New York.

As for my "model for the future" point, the WSJ notes: "The limits build on executive-pay rules announced earlier this month by the Obama administration."

There's more.  Treasury must "examine Wall Street and bank bonuses paid last year and early in 2009 to determine if they were in the public interest. The government could then try to claw back any bonuses deemed excessive."

It will be signed, and it's probably legal:  "Language in the contracts banks signed when banks took money from the Troubled Asset Relief Program allowed the government to change terms retroactively."

Some "stimulus." 

I don't want to, but I am beginning to believe that we may be in the hands of morons.

More on those stupid paycaps

I proclaimed the bailout pay caps “stupid” when they were announced a couple of days ago. Since then there’s been some revisionist thinking about how they're narrow, can be avoided, and even a joke.

I beg to differ, for several reasons. 

First, as I said in my initial post, "work-arounds" "will not necessarily be incentive equivalents." Indeed, I observed almost three years ago that  "second-guessing executive compensation is a tricky business." 

Jason Zweig notes in today’s WSJ that paycaps let covered executives get preferred stock that they can cash in the minute taxpayers get their money back. This encourages the managers to bet the bank – heads I win (and get paid), tails you lose (bank goes bust). Sounds like the sort of perverse, short-term incentives that got us into this mess. Except that now the government is encouraging companies to adopt them as the only kind of incentive compensation for some execs.

My colleague Vic Fleischer says the executives can hedge the risk on the preferred stock. Well, given the point above, let's hope they do.  Vic notes that Treasury could even restrict hedging. I wouldn't put it past them.

Actually, we've been down this road before.  I noted in my post from three years ago that there was solid analysis and data suggesting that Congress's limitation on tax deductibility of then-criticized cash compensation (IRC 162(m)) brought us the now-criticized stock options.  A few months before that I discussed a paper  showing that 162(m) could be expected to increase the extent to which executives would be rewarded based on random components of company performance rather than the results of their own efforts. And the author noted evidence that stock options options increased from 24 % of average CEO pay in 1993 to 47 % in 2002. Of course many have criticized stock options for rewarding based on luck rather than merit. And this is apart from the backdating and other manipulation stock options invite.

To complete the analogy,  Vic notes that bailout executives' unhedged restricted stock compensation that vests only once the government is repaid is really "more like an at-the-money or slightly out of the money stock option."

When firms are not subject to government manipulation, they can contract to minimize these effects.  Consider Axelson, Strömberg, and Weisbach's analysis of private equity compensation in Why are Buyouts Levered? The Financial Structure of Private Equity Funds. They note that the limited terms of buyout funds give managers an incentive to go for broke just prior to termination.  The funds plan for that by forcing the managers to pool investments from several buyouts so they have to subtract the failures from the successes, and to seek financing from third parties rather than simply drawing from the fund. Heavy handed government regulation of incentives inhibits this sort of complex fine-tuning.

Vic notes that banks could avoid the bailout restrictions by outsourcing to management companies and paying uncapped management fees. Zweig in the article linked above makes the same observation and observes that Harvard did just that when its endowment managers' compensation was criticized -- just moved them outside.  The result was large losses by the "outside" managers.  Now that could just be luck or coincidence.  But you could also argue that moving the managers outside reduced supervision.  Indeed, that's the classic tradeoff in the "make or buy" decision.

Finally (at least for now), I noted in my initial paycap post the danger that this is a potential "model for the future."  Consider that former Senator Obama was the Senate sponsor of say on pay, which he's now stuck in the bailout pay rules.  Consider also that once this "quack corporate governance" (as I've described it) gets into a regulation, it's no longer just an idea, and it acquires new supporters -- the companies that are already subject to it and want the same for their competitors.

The idea that these proposals are an innocuous joke is, in my view, flat wrong.

Backdating and the story that got away

Remember backdating? The story of the century? Winner of a Pulitzer Prize? Surely you must, because a guy got sentenced to 21 months in jail and fined $15 million for it.

That comes to mind because John Carney tells us:

In his Capitol Hill testimony today, Markopolis revealed that he contact[ed] a WSJ reporter in December of 2005. The reporter seemed interested in the story but his editors never greenlighted an investigation, Markopolis says. * * *

Around the time that Markopolis was talking with the WSJ, the paper was gearing up to launch its ground-breaking investigation into a scandal involving the misdating of options grants. That story got played up big time in the Journal, picked up by both the SEC and journalists across America, and eventually won the Journal a Pulitzer Prize. Of course, it quickly fizzled out as regulators came to realize that it wasn't quite the big deal almost every one thought it was. It became obvious that, far from uncovering the financial crime of the century, the WSJ had uncovered some minor accounting errors that had become common practice in the tech community. Meanwhile, the WSJ was missing the actual financial crime of the century. They were so convinced that corporate CEO pay was somehow criminal that they missed the real criminality even after it was delivered to them on a silver platter. No one would expect that the Madoff story, which is basically about a fraud at a little known securities firm, would garner anyone a Pulitzer.

As I said at the time, "backdating is not a business scandal. It is a scandal of credulous (or worse) business journalism."  Now we're learning that this scandal also had opportunity costs.

What would you give to see somebody with a real company making actual money and employing real people doing an IPO today? How much would it bother you to find out that the CEO got stock options on a different day than he said he did?

Meanwhile, when this stuff was front page news, Bernie Madoff was stealing billions from investors. Markopolos couldn’t interest the SEC in the story, either. If Greg Reyes loses his appeal, he'll go to jail. I wonder if Madoff will ever see the inside of a jail.

Backdating as optimal contracting

I've been writing for sometime about my skepticism concerning the backdating so-called scandal. The posts are collected in my executive compensation archive. My basic line is that backdating was not in fact the blatant theft from the shareholders it has been portrayed as. There are reasonable explanations, and it's not clear the shareholders suffered any harm (though the plaintiffs' lawyers who jumped on the backdating bandwagon are regretting their investments). There may have been disclosure violations, but those don't make for the scandal of the century.

The latest theory and evidence along these lines Mahmudi and Gao, Backdating Executive Stock Option Grants: An Agency Problem or Just Optimal Contracting? The paper shows

that backdating could be the consequence of optimal contracting when shareholders are maximizing their value. We first establish a theoretical model which predicts that managerial backdating benefits shareholders by (1) reducing the management compensation cost and (2) increasing managerial incentive. Using a large dataset, we provide strong empirical evidence supporting the model's predictions. Providing further support for our theory, we document evidence that strong corporate governance reinforces backdating. Overall, our evidence supports the optimal-contracting view of option backdating, and contradicts the prevalent agency explanation.

The compensation cost thesis is based on the idea that risk-averse, under-diversified CEOs want a strike price lower than the grant-date stock price. Thus, the authors say, backdating yields "lower compensation cost for a given level of managerial incentive."

Indeed, the data show that

backdating is negatively related with the CEO’s total annual compensation and his cash compensation. This evidence suggests that shareholders simultaneously grant less cash payment to CEOs when allowing them to backdate their option grants. What is more important, the total compensation cost is actually reduced in the presence of option backdating, which is contradictory to the view that backdating makes shareholders overpay. This evidence strongly supports our optimal-contracting explanation that backdating saves compensation cost for shareholders.

With respect to incentives, the authors show "a strong positive association between backdating and the pay-performance relation in executive pay."

And the authors "document robust evidence that strong governance is associated with more backdating. This result directly supports our optimal-contracting explanation and goes against the prevailing agency argument."

Finally, the authors give the standard tax/accounting explanation for why backdating rather than in the money options. So, yes, backdating was an end run around accounting rules that effectively penalized in the money vs other options, but the rules made little sense. This doesn't make manipulation of the rules right, but it's a far more benign explanation than the one that launched myriad costly investigations, lawsuits and prosecutions.

CEOs aren't overpaid

Marc Hodak has some thoughts worth sharing in Forbes:

Consider that the average S&P 500 company has a market value on the order of $10 billion. If one had to choose among CEO candidates, and the board believed that one candidate's leadership was likely to yield a return on capital just one percentage point better than the next best candidate's, that difference would be worth $100 million per year to investors. A conscientious board with the shareholder's interests at heart could hardly risk letting the best candidate go elsewhere over even a few tens of millions of dollars. * * *

Nearly every reform attempting to rein in CEO pay has been based on some version of the managerial power thesis. These attempts have proved ineffective or backfired in exactly the way one would expect if executive compensation were driven by a reasonably well-functioning market for talent. * * *

[T]o continue to recommend and implement reforms based on a theory that directors are lazy, incompetent or corrupt--or that competition for talent is an afterthought in setting pay--only invites policies that impose costs on companies for which there may be no offsetting benefits, and which may create distortions that actually undermine shareholder value.