Hacker News Reader

Comment by woopwoop | Hacker News Reader

I remember an interesting theory about golden parachutes in Steven Landsburg's "The Armchair Economist". In a typical publicly traded company, the CEO is actually much more heavily invested in the company than the shareholders. His salary, and probably the major part of his assets, are dependent on the performance of the company. On the other hand, the average share holder is well diversified. Therefore, a CEO will be incentivized to pursue a highly conservative strategy, while the shareholders may wish that he takes more risk. To offset this, the shareholders may make it known that a generous severance package, as well as generous benefits if the company does exceptionally well, are on the table, to simultaneously soften the blow if he takes a bad risk, and sweeten the deal if he takes a good risk.

In other words, the point of executive compensation may not be to produce better results on average, but rather to increase the variance of the results.

View on HN

> Therefore, a CEO will be incentivized to pursue a highly conservative strategy

I strongly disagree. A CEO is probably highly personally invested in the company, but that investment almost always has an unusual structure: it resembles a call option, not equity. If a CEO generates a large gain over the course of a few years, the CEO makes a lot of money. In contrast, if the CEO generates a large loss over the course of a few years, the CEO loses very little. This can give a CEO an incentive to make extremely risky decisions because the CEO doesn't personally suffer much more from a huge loss to the company than from a small loss to the company.

This problem exists for investment managers. In a hedge fund that charges a performance fee on investment gains, managers have a perverse incentive to take large risks. It gets more pronounced if the fund is already down for the year: if, say, the fund has taken a 40% loss, it can look very attractive to the manager to bet all of the remaining assets on a coin flip. Heads, they get their bonus. Tails, they now have a 90% loss, but they weren't getting their bonus either way and they lose nothing.

Edit: At least the coin flip is neutral on expectation. But the same issue exists for a bet with negative expected value: the manager gets some benefit if they get very lucky, so the manager has positive expected value even if the fund has very negative expected value from their decision.

>This can give a CEO an incentive to make extremely risky decisions

That's the intended effect. You missed the point. Read woopwoop's comment again.

I think you need to read both woopwoop and amluto's comments again.

woopwoop is saying that the equity compensation given to CEOs is a disincentive for taking risks and that it needs to be offset by a promised severance package in order to increase the variance of outcomes.

amluto is saying that this is a mischaracterization of the equity compensation for CEOs, which only exposes them to gains and not losses and thus they are already incentivized to pursue a risky strategy (e.g. golden parachutes are not required to create this incentive).

woopwoop is saying that CEO compensation is structured to benefit the risk-taking CEO no matter how well his risks pay off ("generous severage package" if they don't pay off, "generous benefits" if they do).

amluto is saying CEOs are exposed only to gains and not losses, which is agreeing with woopwoop without realizing it.

The reason a BoD/shareholders would want to set up a CEO with such an arrangement is because the much wealthier shareholders are well-diversified and risk-seeking (they don't depend on their income from this one company to feed their family)- while the CEO is not as diversified, he DOES depend on his compensation from this one company to feed his family- so if the BoD/shareholders expect him to take risks, then he needs to be protected from the consequences of those risks.

Actually hedge fund managers tend have a large portion of their net worth invested in the fund also.

Only the new ones.

Even when this is true, it doesn't necessarily solve the problem. Say you're 28 years old and you're running your first fund. You have a net worth of $500k (because you worked 100-hour weeks as a banker for a few years, paid your student loans, and saved up). You put $400k into your fund and raise $50M. You charge a 25% performance fee.

If you get 10% returns, that's $1.25M for you (at a possibly low tax rate, too) plus $40k on your own investment. If you lose 20%, you're out $80k. What's your incentive here?

Right. That's pay for volatility. Huge problem.

Worst case: Pacific Gas and Electric before and after deregulation. Before deregulation as a public utility, a century of modest profits and modest CEO pay. After deregulation, wildly volatile profits ending in bankruptcy three years after deregulation, and much higher CEO pay.

> Tails, they now have a 90% loss, but they weren't getting their bonus either way and they lose nothing.

Except that their investors will flee (cash out) and they'll not be getting any management fees next quarter.

They'll cash out anyway -- a 40% loss is pretty bad, too.

Also, apparently having a famously bad investment track record doesn't make it much harder to raise money for a new fund in the future. If I remember right, the LTCM managers kept right on going after LTCM blew up.

In hindsite - LTCM should have been the canary in the coal mine. They got bailed out to the tune of about 4 billion. Sounds like chump change after 2009.... And the banks are STILL too big to fail. If they truly are too big to fail then banks should be regulated like utilities used to be. What will it take before a majority of this country realizes that? Another 2009?

It will take a collapse of the moneyed politics, probably through bloody revolution, for this to change at this point.

A suggested solution is 'career shares'. The CEO is required to hold shares granted from incentive scheme until retirement. Incentivising them to develop long-term value - http://www.business-school.ed.ac.uk/about/people/65/Brian/Ma...

I'll try and remember the name of UK CEO that successfully extracted £200m in incentive based remuneration over a ten year service. A service that saw the share price boom in early years, then by the end of the 10 years crash to below the starting price.

(1) The CEO is actually much more heavily invested in the company than the shareholders. His salary, and probably the major part of his assets, are dependent on the performance of the company. On the other hand, the average share holder is well diversified.

(2) Therefore, a CEO will be incentivized to pursue a highly conservative strategy, while the shareholders may wish that he takes more risk.

1 is true (assuming heavy equity-based compensation), but 2 does not follow from 1.

What does follow is that the CEO will do whatever he can to make short term gains for the company before he exits, even if it means kicking the can down the road for future managers and long-term shareholders to take a hit on at a later date. When the whole management team is on in this short-term focus, it is colloquially referred to IBGYBG (I'll be gone, you'll be gone).

Shareholders are not a homogeneous group of investors. Some pension/mutual funds invest with the goal of exiting in 10 years. Some activist shareholders invest with the goal of pumping up the price, then cashing out within 3 months (see Carl Icahn and Apple). A sizeable (and growing) portion will be owned by index funds while some will be owned by hedge funds, etc. Then you also need to factor in employees, executives, etc.

One (newer) way to make this group happy as a whole and solve the IBGYBG problem is through clawbacks - punishing ex-executives for the things that happened on their watch. Check out Wells Fargo to see how this is playing out in Congress and in court.

The question the paper is trying to answer - what motivates CEOs to "exert effort" in able to return industry-beating results - seems like it has a simple answer to me: no CEO wants to fail at his/her job and be known as the one who presided over company X's slow decline into bankruptcy/irrelevance. Doesn't look so good on the resume when shopping around for your next C-suite role.

There are rules (from the SEC as well as often the company board) which prevent executives from selling significant portions of their stakes quickly.

And the original golden parachutes were created to incentivize a good acquisition. A CEO whose company gets acquired is often firing himself. So the whole idea was to give them a huge score if it happened. Otherwise CEOs might fight takeovers just to keep their jobs.

And, if a lot of the CEO performance results is just luck, then bad luck can ruin them. If a company is on the brink of failure, nobody is going to take the job without a parachute. It could be their last job if they take the blame.

So why not make the golden parachute "if we get acquired, you get this" instead of "if you get fired, you get this".

Because CEOs didn't get there by not knowing how to negotiate and insulate yourself from risk

What's to stop the board from firing him and then doing the acquisition to save the money that they would have had to pay him.

In theory? Not much. In reality? For one, that's a lot of work, especially to screw someone who did a decent job (presumably), who's compensation probably is a drop in the bucket compared to the stock price, and, let's be honest, has a chance to be on your board. The amount of incest in company boards is pretty high. So if you do it to them, there's a chance they'll do it to you.

They did originally. I think even now most big CEO's don't have a really huge golden parachute. And at a certain point, it's just differed compensation. Instead of 2 million a year and zero parachute, they get 1.75 million a year and a million dollar chute.

There is also another form of golden parachute, which the person is getting paid for getting on a sinking ship. Otherwise nobody is going to become the CEO of a failing company. They are basically getting paid for the hit to their reputation.

This exists, it's called a "Change in Control" arrangement.

> a generous severance package, as well as generous benefits if the company does exceptionally well, are on the table, to simultaneously soften the blow if he takes a bad risk

So a pay gap of up to (and sometimes over) 1000x between CEOs and rank-and-filers, given all the saving and investing that the CEO could do just like any Average Joe, would not be enough to "soften" any "blows"? Huh.

The gigantic pay gap may be an injustice, but I think you are missing the subtlety of the theory. The worry is that the CEO will become too personally dependent on that 1000x salary to take "appropriate" risk. For example, assume that launching a new product line has a 50% chance of doubling the companies profit, a 25% chance of doing nothing, and a 25% chance of cutting the profit in half. For the shareholders, this is probably a good bet to take.

Then assume the CEO will be replaced if profits drop. In the absence of a "golden parachute", a CEO might reason that a 25% chance of being fired and loosing their cushy job makes the plan not worth pursuing. The theory (OP wasn't necessarily endorsing it) is that a rich severance package can better align the interests of the parties. While I personally think "cronyism" is a better explanation for the current situation, the theory is better than others I have seen.

> I think you are missing the subtlety of the theory

There's nothing "subtle" about a plain racket. And there's nothing subtle anymore, either, about bringing in the Ayn Rand style of "debating" by putting people down.

At 1000x pay gap, the CEO needs only work a few months to have nothing to worry about for the rest of his life. Anything on top of that, anything, is a luxury item.

"Risk" of what? Having to manage with only 2 yachts instead of the customary 3? Right, that's a serious existential threat, gotta mitigate it.

Shave a couple orders of magnitude off that pay gap, and then the argument might, just might, begin to make sense.

Why do you think that the shareholders don't have a right to offer generous compensation packages to those they consider good leadership talent?

Why do I think something I don't actually think?

Shareholders, in theory, should have the right to make any decision that is legal and is advantageous to them. That's what I think should happen, in an ideal world.

The practice, however, is different. The entity that decides CEO compensation is not "we the people". It's more like the smoke-filled room at the Republican Party's HQ in The Simpsons.

Shareholders do decide. You seem to be implying that shareholders are being cheated, however, if they are actually unhappy they can a. sell (or not invest in the first place if they don't like the representation they'll receive), or b. vote in a new board, or c. sue.

You are struggling hard to make a victim where everyone involved is happy.

I don't think it is that simple.

There is a tremendous amount of back scratching that goes on when setting executive pay.

The board hires pay consultants, the consultants are not going to rock the boat and suggest pay cuts, so they approve a nice increase over the current going rate. Rinse and repeat and you have out of control executive compensation.

Majority shareholders are often in the same position - very well paid, and well connected with other CxOs.

That leaves the rest of the shareholders along for the ride.

> You are struggling hard

Whenever I argue with a staunch defender of social inequality, the put-down tactics are par for the course. "You struggle", "you fail to comprehend the subtlety of the argument".

If doesn't take a PhD in psychology to understand why that is. For someone to be pro social inequality, they must believe they are already topkek - or are standing a very good chance to make it. If that is the case, then it's a simple matter of statistics that any random stranger is likely placed lower in the pecking order. Therefore they must be "struggling" to comprehend, etc.

There is no "struggle" when you're telling the truth. The higher cognitive load happens when you make stuff up from whole cloth - there are studies showing it, look it up.

> everyone involved is happy

In this reality, the trends in political news in the last several years show otherwise. Not sure about other realities.

It's not even that. A high-paid CEO is a prestige thing, like a status symbol, for a corporation. "We are great because our leader is great and as proof look how much he/she gets paid". It's like an individual flashing their gold Rolex (or in Marissa Mayer's case, a Cartier).

Who are you to tell a company how to manage itself and how much they can pay their employees? If the board feel a CEO generates that much value for their company they should definitely pay him/her what they are worth.

Who are you to tell a company how to manage itself and how much they can pay their employees?

A citizen, which is to say, part of a group that collectively decides the "rules of engagement" for corporations. Society has chosen to grant certain rights to corporations because society feels that properly regulated corporations are a net benefit. At the point where enough individuals decide that the benefits are no longer sufficient, new rules can and will be made.

There is (to me) nothing inherently sacred about allowing amoral corporations to operate in the manner that they decide is most effective for their interests. We've decided that minimum wages are acceptable, that maximum work weeks are a good thing, that certain styles of contracts are unenforceable, and if we wanted to, we could decide that excessive differences in pay are legally unconscionable.

There's always a danger that societally we'll make the wrong choices, and that new rules will do more harm than good, but I don't think there is any question that society has a "right" to make such rules? At least no more than there is question that society has the right to make and enforce any other law? Turning it around, who are you to tell us that we can't regulate these companies how we choose?

And there's nothing subtle anymore, either, about bringing in the Ayn Rand style of "debating" by putting people down.

I may be the one missing the subtlety now, but if you are referring to my phrasing, I didn't intend any insult. From your comment (voted below zero before I upvoted and replied) I felt you were being unreasonably dismissive of a potential explanation that I found interesting and had not heard before.

There's nothing "subtle" about a plain racket.

The theory offers an explanation of why stockholders might rationally offer a generous severance package to a CEO. I agree that the "racket" theory has greater explanatory power, but I found it worth contemplating alternative explanations. Maybe "subtle" is the wrong word? "Counterintuitive"?

the CEO needs only work a few months to have nothing to worry about for the rest of his life

In practice, though, CEO's worry tremendously about their ability to accumulate further wealth. I wouldn't be surprised if this worry increases the greater their current salary. And from the shareholders' point of view, it's the CEO's perceived risk that affects their decision making, not the actual risk.

Shave a couple orders of magnitude off that pay gap, and then the argument might, just might, begin to make sense.

Well, no. As a theory to explain current behavior, I think this theory is actually more applicable the greater the discrepancy in pay between the CEO's current job and the other jobs available if they are fired. You are right that the situation would be different the if the gap was less, but that doesn't mean the theory is not relevant when the gap is large.

Maybe you'd prefer looking at it from a "Prisoner's Dilemma" perspective. Assume you have a extremely-well-paid self-interested psychopath making the executive decisions for your company. Ideally, this would not be the situation you find yourself in, but let's assume it's the case.

Assume that as a rational (read, "amoral") shareholder you wish to maximize your gains, which is positively but nonlinearly correlated to the company's profitability. How do you best achieve this? Perversely (and to me counterintuitively) reducing the downside to the CEO for risky decisions may be a strategy with a positive return.

The kind of CEOs who get golden parachutes earn enough money in a single year to live a whole life on it without having a job again. There is absolutely NO risk whatsoever taken. There is not even any investment that they could lose either, it's only profit + more profit.

> The kind of CEOs who get golden parachutes earn enough money in a single year to live a whole life on it without having a job again.

You probably earn enough in a single year to live the rest of your life in a different part of the world, yet you continue to work, in part because you want your current standard of living.

Likewise, a CEO will want to continue to live at his current standard of living, or even better.

> a CEO will want to continue to live at his current standard of living, or even better

Al Capone and the robber barons had the same goals.

The only thing that matters is your annual income and your expenditures. If you earn 1x as much as you spend, you're staying afloat. If you earn 30x as much, then a year of work covers 30 years of retirement.

The CEO of Kroger earned $13 million in his last year there, and called his own compensation "ludicrously high". I doubt he's spending a million a year, but even if he is, the ratio is indeed ludicrous. And he's nowhere near the high end for executive compensation.

Are you suggesting that the average top-tier executive lives a lifestyle in the same proportion of their income as the average software developer, or blue collar worker, or...? Really?

I never really understood why people think the pay gap between the CEO's and the rank-and-filers is such a negative thing.

If you look at walmart for example, and entirely eliminated the compensation of the CEO (19.4 million), and all executive vice presidents and CFOs (2.8, 10.8, 8.4, 10.1, 11.5, and 8.5 million respectively) and total them all up you get ~$71.5 million.

71,500,000 / 2,000,000 million employees = $34 dollars per employee per year of a raise. That's almost negligible, and if you assume minimum wage (7.25) 50 weeks a year, for 40 hours a week gives $14,500 / year, you get $35 / $14,500 = a .24% raise. That's less than a 2 cent per hour raise per employee if you completely slashed the salaries and compensations of the top executives.

When people talk about pay gaps between executives and rank-and-filers, they generally mean pay gaps between individuals, not pay gaps between collectives (in this case, the 7 executives vs the 2 million employees).

There's a popular narrative that CEO compensation is at the expense of workers, that workers could be paid a livable wage if only executives could do without their extravagent bonuses. It's commonly referenced with respect to Walmart. I believe this is the myth the OP was attempting to dispel.

Walmart is an extreme example. If you read "Capital in 21st Century" you'll see a lot of the worries of inequality is caused by the general rise of "super managers." If you don't have time to read that massive book, I think this is a good summary: https://jesiwagle.wordpress.com/2015/01/16/super-managers-do....

You picked the an extreme example there. Walmart employs a zillion unskilled people. Similar numbers for Google would yield something like an extra $6,000 per employee per year.

But I think it's a negative thing because it's a sign of a shift in behavior. Look at the graph here: https://en.wikipedia.org/wiki/Executive_compensation_in_the_...

CEOs have gone from making 20x the average worker to 200x. Are CEOs ten times more amazing? If anything, I think they're worse. I think that, broadly, they've shifted their focus from economic value creation to generating the numerical appearance of success in ways that let them extract lots of money.

There's also the problem of worker motivation. Companies are human enterprises. The greater the wealth differential, the less the people doing most of the work will feel part of something bigger than themselves. As with Walmart, they may show up and do what they're told, but they are less likely to actually give a shit.

I definitely agree that they are not 10x more amazing, but I also find that in the case of a company like google any kind of pay differential has far less of an impact. Employees at google get paid enough, and have such great benefits that they are easily in the top few percentages of the economy. The same can be stated about many tech companies and their employees. Engineer salary and benefits are very comfortable at this particular point in time, because it is very in demand, and requires high levels of skill.

It seems at least from what I've noticed that the companies with the highest differential in terms of rank and file -> CEOs, tend to be the ones with some of the lowest barriers to entry and the lowest skill levels. This is a subject that I would love to see a detailed study on because what "it seems like" is not always the reality of the situation.

I used Walmart precisely because besides being an extreme example, it is one of the largest employers in North America by a good margin.

Edit: forgot a word.

Reasoning from extreme examples is unhelpful when you are asking people to consider the general case.

Let's not guess.

According to Google search: "Google CEO Sundar Pichai made $100.5 million in 2015, according to a regulatory filing released Tuesday. The filing revealed that Pichai was paid a salary of $652,500, awarded restricted stock worth $99.8 million"

and according to Business Insider, Google has roughly 57k employees.

If we just take Pichai's $652k salary, it is +$11/employee/year which is a couple cups of coffee.

Alternatively, if take his entire compensation of $100M, it becomes +$1754/employee/year which is more tangible but still nowhere near the $6k/year cited in parent.

If we just consider salary, I think the multiplier between the lowest to highest paid is going to be lower than you expect in tech. It's really only visible in more low skilled industries.

I'm following charrondev's approach; he was using total compensation for the easily findable executives.

Oh, and, discussion on executive compensation always includes total compensation. For historical and legal reasons, much of the compensation growth happened in non-salary categories.

I believe the reasoning is that it is negative if it hurts the company's bottom line. It's hard, if not impossible, to causally correlate CEO compensation with company profits so there is no reason to assume for example that "golden parachutes" pay for themselves. They don't. If company profits scaled with C-level compension, why not double their bonuses already? That's silly.

And what would compensation per employee be indicative of? Nobody is criticizing that. There can be tension in public companies between the board's interests and the shareholders'. Questions of compensation is a common source.

I think all these comments are on to the same well-documented idea: Principal-Agent problems

https://en.wikipedia.org/wiki/Principal%E2%80%93agent_proble...

The way I look at it, executive stock options aren't to reward behavior, they're to ensure loyalty. It's normal for any human being to feel more empathy for hard-working employees that they work with every day than shareholders who are more distant and who contribute in a passive way.

Stock options give CEOs a large financial incentive to place the interests of stockholders above those of employees whenever those come into conflict (for instance, in decided whether to use profits to increase compensation or increase dividends).

Anyways, that's my theory of why large executive stock bonuses might be a good for shareholders despite unintended consequences like decreased risk taking.

> To offset this, the shareholders may make it known that a generous severance package, as well as generous benefits if the company does exceptionally well, are on the table, to simultaneously soften the blow if he takes a bad risk, and sweeten the deal if he takes a good risk.

This concept strikes me as grossly unfair towards the employees. Like the CEO, they are also not diversified. Their risks may be even bigger than that of the CEO: if the strategy fails, they may loose their job/income, and they make less to begin with. How are they rewarded for pursuing the more risky strategy?

They will likely also have some level of stock based compensation.

On the downside most tech employees can easily find a job elsewhere, whereas the CEO may take a reputation hit that makes it hard to get another position at a similar level. So it probably comes down to the regular employee not valuing downside protection enough to give up some base compensation or upside reward.

I'm curious if these incentives really work at all.

For example it's common knowledge a real estate agent won't go for the most money per deal. They'll go for the quickest money per deal. No real incentive to spend extra time pursuing the most money.

I get that's not a direct analogue for the CEO case but how do we know the getting 80 million for hard work and success is more motivation than 15 million for easy work and failure.

It seems like if you want to motivate success you'd also need to disinsentivise failure . Not by making it so the CEO gains less if they fail but so they actually lose.if they fail. In otherwords their net worth has to go down for failure.

> In other words, the point of executive compensation may not be to produce better results on average, but rather to increase the variance of the results.

The hypothesis you relate isn't implausible, but I think you either got this part backwards, or, more likely, the way you state it is misleading. The strategy would be nuts if it were intended to produce worse expected outcome in terms of monetary return at the (additional) cost of increased variance.

That sounds very academic and theorized in a vacuum. If that were true you might expect contract terms to be explicit about this, and quarter-oriented management to be less commonplace.

But as an investor I don't care much if one particular company takes more or less risk. I just want to know how much it takes so I can adjust my portfolio accordingly.

Reply on news.ycombinator.com