Ben Horowitz, serial technology CEO and cofounder of venture capital firm Andreessen Horowitz, wrote a book with a unique goal: to help investors understand public company managers.

The Hard Thing about Hard Things is built around a claim that the man on the street may find laughable: public company CEOs have a hard job.

Of course, lots of other work hard: piece-work seamstresses in sweat shops, research scientists, and taxi drivers are a few random examples.

But CEOs are extraordinarily well-compensated for their efforts. The median big-company CEO earned over $10 million in 2012. Smaller companies pay less, but the CEO is still likely making more than anyone else. She is also likely to have the best office, the most efficient assistant, the most generous severance package, and the least scrutinized expense account.

The job is made easier because you have no boss. You technically answer to the board of directors and perhaps other constituencies, but that's not the same as a day-to-day boss on-site. There's no benchmark for success, no audit of performance, and no professional standards that need to be maintained.

You could argue that the company stock price is a performance measure, but hard jobs have much more direct, frequent and competitive measures: athletes, politicians, and trial lawyers have to win versus other top competitors, surgeons have to save lives, computer programmers have to write code that works, movie producers need to complete bankable films on time and on budget.

The CEO job requires no specific talents or training. Some CEOs perform other jobs, often helping out in marketing and strategy, or perhaps engineering or finance. That's helpful, particularly in a smaller company, but the core CEO function is executive, not technical, and executing is not rocket science or brain surgery or even tic-tac-toe.

I admit the CEO has to make hard decisions, like firing a friend or shutting down a line of business. These decisions can devalue years of other people's work and blight career prospects. They can take the fun out of jobs, disappoint customers, and strand loyal people in bleak situations.

But as hard as it is to order these things, it's harder to be victimized by them.

So once again, the CEO seems to be in the relatively easy position. And while CEO actions can do damage, compare that to surgeons, police officers, or soldiers, whose mistakes can end lives.

Even more than any of this, what should make the CEO job easy is that by the time decisions reach the top, they have to be pretty evenly balanced to the point where they're basically all close calls. Flipping a coin or using a Magic 8-ball wouldn't be a terrible idea. Some CEOs might add value through strategic acumen, but given the enormous unpredictability and complexity of events, I'm skeptical that the average CEO does appreciably better than random chance.

What Ben Horowitz makes brilliantly clear is that to be a good CEO, you have to make the decisions hard. Although a net decision may be a close call, it can be extremely consequential. For example, cutting research funding might be great for short-term cash flow but catastrophic for long-term innovation. Reducing compensation may be good for the bottom line but destructive of employee morale.

There's no logical way to deal with aggregations of incommensurate considerations. There is no price on human life, no way to balance fairness versus efficiency, no ruler that works for social utility of product and employee satisfaction. But no one wants to work for a person who flips a coin on these matters rather than investigating each aspect and feeling the pull of competing imperatives. It may not change the decision, but it changes the company. People will make great sacrifices for and accept great disappointments from leaders who respect their loyalty. No one wants to be a sucker for an uncaring boss. A CEO who cares, who feels that decisions are hard, will not forget the people who got hurt by a decision; she will try to find ways to make it up to them. A CEO who flips a coin will flip the same coin the next time and possibly double the damage.

It takes active cooperation, not just tolerance, of many constituencies for a company to be successful. People have to want to work there to bring their best efforts and function as efficient teams. Customers have to buy, suppliers have to sell; and both should be providing essential information and thinking of ways to enhance mutual advantage. Investors have to invest, regulators have to okay things, communities and governments have to regard the company as an asset. It's possible to struggle along for a while without satisfying some of these groups, but the probability of success decreases.

Within each group there are multiple considerations. For examples, employees like money, fun work, the right degree of challenge, a pleasant environment, satisfying career paths, resume-enhancing responsibilities, and meaningful work. Different people value these things in different degrees, and in some cases, define them in different ways. CEOs cannot deliver all of these things to everyone all the time, but to keep anyone happy, they have to try to deliver as much as possible most of the time.

When I was an activist investor, I always experienced a great relief when I got to a one-on-one meeting with a CEO. Even as the largest shareholder, which I often was, this could take years. You mostly talked to lawyers, who were making hourly fees of your (shareholders') money to block you from getting anywhere. The longer you sat and the more frustrated you got, the better for the lawyer. You could talk and the lawyer would reply, but the words did not connect with reality -- an infuriating arrangement.

The refreshing thing about the CEO, and this was as much true of bad ones as good ones, is she understands that everyone has to be satisfied for the company to prosper. She is concerned with reality, not words. She may not give you what you want, but she understands that the problem is to balance your desires with all the other stakeholders in the firm. She's looking for a creative solution acceptable to everyone, not erecting a wall to block your influence on firm decisions.

She'll try to change your mind on some issues, give in a little on others, and change her mind on some of the rest. But the most important satisfaction you get out of the meeting, and you usually do get this, is knowing there is a meeting of the minds and that shareholder interests are valued.

Unlike all the other groups mentioned above, shareholders are not in the office every day, and their active cooperation is not needed unless the company wants to raise new equity capital. Even when the company needs shareholder votes, the elections are sufficiently rigged that it is a small constraint. Bondholders demand extensive legal covenants to protect their interests, employees have laws and government agencies behind them, and customers can go elsewhere.

Shareholders have to rely on the CEO. If you think shareholders can also look to the SEC, the Delaware courts, or their elected representatives on the board to guard their interests, you might be too naive to invest in stocks. Shareholders cannot trust any institution that claims to represent them but fought for poison pills or against expensing employee stock options.

Ben Horowitz, despite being a professional investor with CEO experience, is not sympathetic to shareholders in his book. He is bitter that at one point, his company's stock price was less than its per-share cash, as if managers could be expected to burn through the cash and then declare bankruptcy.

That's precisely what investors expected. Perhaps if they had known Ben Horowitz or the company better, they would have had a different opinion, but they cannot be faulted for their cynicism given what usually happens in these cases. He had some brushes with accounting issues. In one case he hired a CFO who was later convicted of fraud in a prior company. In another, an accounting treatment was overruled by the auditors' national office at an extremely inconvenient time for the company. In neither case does he seem to appreciate that accounting information is oxygen to shareholders and the slightest hint of taint induces immediate panic.

Yes, his businesses were at risk of major disruption or ruin due to technical accounting disagreements. I can see why that would infuriate a manager. But companies have also failed over minor personal differences among managers, technical issues in patent filings, or a slight carelessness in quality control.

Why would he take these so much more seriously than accounting? I think it's because these problems hurt the business directly while accounting problems kick off regulatory sanctions, lawsuits, and stock price declines that attack the business from the outside. With a business problem, the CEO can go to executives, customers, or engineers directly and find a solution. With accounting issues, Horowitz was dependent on distant autocrats (although he managed to eke out solutions anyway). That can lead to thinking of investors as enemies, and accounting as a perimeter defense. And those who challenge accounting are treated as traitors if they are on the inside, or seen as on the vanguard of the attack if they are outsiders.

Perhaps some of this attitude comes from running companies that required frequent equity infusions to execute the business strategy, or increasing stock prices to keep equity-owning employees motivated. It could also stem from experience as a hands-on venture investor who knows far more about the company than is communicated by financial statements. Perhaps public shareholders with diversified portfolios who know nothing about individual companies seem ignorant but dangerous, and the legal and regulatory systems designed to protect them from their recklessness seems stifling.

I take a different view, although I'm intimately acquainted with the problems Horowitz describes. Efficient, diverse public equity markets accomplish two great things, greater than anything ever accomplished by a government or other top-down organization.

First, they allow ordinary people to achieve financial security through their own efforts, without investing acumen or detailed attention.

Second, they turn slack economic resources into extremely cheap capital that can be accessed by almost any idea with a positive expected return. Someday, things like Bitcoin and crowdfunding may accomplish the same goals more effectively, but for the last half-century, nothing has come close to the stock market.

The hard thing about the stock market is it's pretty much a random walk. That's hard for investors who would like predictable returns, and hard for CEOs who want a rational partner in building the business. But "random" only means unpredictable, not uncaused. CEOs should treat the silliness of accounting debates with the same respect they would give the silliness of employees' emotional feelings, consumers' tastes, or legislators' whims.

That makes the CEO job less easy, but Horowitz's book makes clear that the harder you make the CEO's job, the better job you're doing.