(FORTUNE Magazine) – I first met Dan Vasella last February at a bar in Manhattan's St. Regis Hotel. The CEO of Swiss pharmaceutical giant Novartis was in town for the World Economic Forum, and between heady conference sessions on global security, global imperialism, and global warming, Vasella had offered to sit down with a FORTUNE editor for an informal interview. My boss made me go.

It was 5 P.M. My wife was making dinner, and I was sure I was in for a 30-minute discourse on drug pipelines and the revolutionary age-spot cream that was then in Phase II clinical trials. (I ordered a very expensive Scotch.) Strangely, though--and it seemed to begin the moment I arrived--we descended into subjects both personal and raw. Sitting in the dim light of the hotel bar around a cramped table, we talked about the physical pain of cancer; about poverty; about the unforgivable cost of drugs for the uninsured; about the loss of family to disease. His face seemed to defy characterization--swapping empathetic pangs for boyish grins, then suddenly shifting to sober, almost absent, meditation. During the hour or so we sat together he barely mentioned his company, and I began to wonder if Vasella, who as recently as the late 1980s was working as a medical doctor in Bern, was really a CEO. He seemed too, well, candid. Too trusting of a journalist. Lord, I thought, doesn't he know what I do for a living?

Indeed, it was his fearless candor that made FORTUNE think of Vasella when the special CEO issue now in your hands was merely a wish list in a hot summer editorial meeting. We wanted a bigtime executive to speak boldly and honestly about the pressures that Wall Street analysts put on the leaders of public companies--specifically with regard to quarterly earnings targets. How hard is it, we wondered, for a chief executive to plan the long-term success of a company when performance is judged--often in the narrowest terms of reported profits--every three months on Wall Street? And if not on a quarter-by-quarter basis, how should securities analysts--and more important, investors--evaluate public-company bosses? What the heck is an appropriate yardstick for success, anyway?

What makes Vasella a particularly good candidate to answer these questions is that he has indeed been successful at his job, making money for shareholders with relative consistency. Novartis, with 70,000 employees worldwide, notched over $4 billion in profit on roughly $19 billion in revenue last year. Since it was formed from the merger of Sandoz Pharmaceuticals and Ciba-Geigy in December 1996 (and Vasella was named chief executive), the company has returned more than 62% to shareholders. Year to date (including dividends) it's up over 3%, compared with the Amex drug index, which is down 21%. (Pfizer is down 22%; Merck, off 8%.)

But Vasella's company has been equally successful in other ways--most notably with the introduction of its breakthrough cancer drug, Gleevec. The company has also been laying the foundation for future growth, spending heavily on R&D during the economic downturn. (It is now building a new billion-dollar research facility in Cambridge, Mass.) And quietly--make that very quietly--Novartis has given away, or offered at dramatically reduced cost, drugs that fight cancer, leprosy, and malaria throughout the developing world.

So we put Vasella to the test again, sitting down with him for three conversations in October--the first at a mall in Short Hills, N.J., where he scrawled the psychological "flow chart" pictured above; the second on a long-distance cellphone call; and the third at a Japanese restaurant in lower Manhattan. The following is a compilation of his thoughts. --Clifton Leaf

Failure. It is the prospect of failure, of course, that keeps CEOs up at night. But I would argue that there is another, less talked-about risk that may be more treacherous in the end. That is success. Or rather short-term success--what chief executives and Wall Street analysts call "making the quarter."

The practice by which CEOs offer guidance about their expected quarterly earnings performance, analysts set "targets" based on that guidance, and then companies try to meet those targets within the penny is an old one. But in recent years the practice has become so enshrined in the culture of Wall Street that the men and women running public companies often think of little else. They become preoccupied with short-term "success," a mindset that can hamper or even destroy long-term performance for shareholders. I call this the tyranny of quarterly earnings.

Let me say straight off, there is nothing inherently wrong with delivering consistent results, quarter after quarter--that is the mark of a great company, after all, when it is done over time. (And to be sure, one cannot achieve substantial long-term rewards for shareholders without delivering many, many short-term periods of outperformance along the way.) But tyranny is a slippery thing. Rarely does it make itself known for what it is right from the start. Once you get under the domination of making the quarter--even unwittingly--you start to compromise in the gray areas of your business, that wide swath of terrain between the top and bottom lines. Perhaps you'll begin to sacrifice things (such as funding a promising research-and-development project, incremental improvements to your products, customer service, employee training, expansion into new markets, and yes, community outreach) that are important and that may be vital for your company over the long term.

For me, for example, the challenge might be whether or not to strike a deal with a biotech company that is currently running a trail of red ink. Should I decide not to make the deal because I'd then have to consolidate the losses for three years, even though the value of the investment would in all likelihood be positive for the company over time? Do I personally have the strength of my convictions to face the impact of short-term negative results--to say to shareholders, "We invested in this company, and our earnings will grow 3 cents less than expected"? Trust me, it is often a difficult question to answer. And then there is the opposite scenario. Should Novartis leverage its balance sheet to do risky acquisitions or ones that may simply be a poor fit for the company instead of investing in organic growth? True, our earnings and revenue streams may improve in the short term, but at what cost? Or--and we have seen this too many times in recent months--you start to play games with the numbers, digging a hole that over time becomes deeper and deeper, always hoping that you can fill it back up later. In reality, though, you can't cover up that hole forever.

And here I would contend that the culprit that drives this cycle isn't the fear of failure so much as it is for many the craving for success. For the tyranny of quarterly earnings is a tyranny that is imposed from within. That's right. For all of the faults ascribed to Wall Street these days, CEOs can't really blame research analysts or institutional shareholders or other investors for the pressure of short-term performance. Rather, for many of us the idea of being a successful manager--leading the company from peak to peak, delivering the goods quarter by quarter--is an intoxicating one.

It is a pattern of celebration leading to belief, leading to distortion. When you achieve good results--first one quarter, then the next and the next--you are typically celebrated, and you begin to believe that the figure at the center of all that champagne-toasting is yourself: "Hey, look at me. I'm this smart, talented CEO with all the right answers!" You are idealized by the outside world, and there is a natural tendency to believe that what is written is true. It isn't, though--no CEO is as good (or as bad) as the media makes him or her out to be. Nevertheless, many come to believe their own press. Then it becomes difficult if not impossible to change the course you and your company are on. Your mind whispers, You must make the targets--must keep delivering record results at whatever cost to continue the celebration. But in reality things don't go in only one direction; they go up and down.

It's hard to break out of such cycles. It is better not to get in them in the first place. Unfortunately that is no easy task either. From the beginning you must ask yourself one question again and again and again: What really matters? Which actions are really fundamentally important for the success of this company and its constituencies (namely shareholders, customers, employees, and the community at large)? Yes, there is a second question that will creep in, but the answer--as controversial as it is to say it--is almost always no. And that's, Does stock price on any given day really matter? Here, I believe, there are only two times when it does: when the company is being bought or when it wants to buy another. Otherwise, who cares? If you do your job as CEO and strive for what really matters, your shareholders will benefit over the long term.

It is not enough to be truthful to yourself only. To me transparency means that I will communicate truthfully what I do and don't know about my company's performance and prospects, the doubts that I have, and the things that I don't doubt. The goal of transparency is to give the shareholder an opportunity to form an opinion about you, to make a judgment. That's not to say one has to be naive and publicly share information that will harm your company from a competitive standpoint. But in general one has to be transparent to a degree that allows fair judgment of both the company and the strength of the underlying business.

As important as what you do say is what you don't. I don't believe that I, as CEO of Novartis, should give specific quarterly guidance to analysts. That's their business. The analysts can, on their own, make their own estimates. If we say exactly how much we will earn in the next quarter or the next three quarters--and then hit it each and every time--it means we are playing with the numbers. Nobody knows exactly, to the comma, how much we will produce in sales. Give the tools to the analysts to make the interpretations and the calculations--be transparent where it counts--but don't do the calculations for them. (This is really a policy decision, however; therefore, it is important for the board to make it collectively. When you don't hit those external targets, a CEO had better have his board standing alongside him.)

Again, the issue is that expectations can often lead to distortions. If you predict everything and say you are focused on achieving ambitious short-term goals, then people think they have to own the stock now to make money quickly. And if they don't make that money, they are upset. If you say, "I am interested in creating long-term shareholder value, and the company is likely to go through ups and downs en route to that goal," then you get a different shareholder base.

It may sound trite, but I truly believe my ability to keep shareholders' faith in our company depends in the end not on whether I make the quarter but on who I am, what my guiding principles in life are, my behavior. What counts is who you are personally. And this relates not just to the CEO but to the entire leadership team and the company culture it establishes. We work in a system with certain safeguards, but they are imperfect. We can circumvent safeguards for our own interest and for our reputations short term or for short-term success. Temptation is all around us.

Let me offer a strange example. I made relatively little money--much less than my competitors did in similar jobs--as CEO of Sandoz Pharmaceuticals, for example, before the Novartis merger. Even after the merger my compensation was still significantly lower than that of my peer group. Today, however, I'm being paid competitively. The strange part is, the more I made, the more I got preoccupied with money. When suddenly I didn't have to think about money as much, I found myself starting to think increasingly about it. Money corrupts the mind. By the same token, you can find yourself in a situation where you worry more and more about your reputation and become its prisoner. You shouldn't be worrying about it. It is wrong to worry about whether you will be the hero next month. One day the glitter will be gone anyway. Only once you are unencumbered by that will you become free to do the right things as leader of your company. Otherwise you'll always have to do what others expect of you. Or what you believe is expected of you. If you have a great reputation as a consistent performer and you're worried about losing that lofty status, then you become constrained. You become a slave to the master of public opinion.

For me the question is, Am I paying attention to my own unspoken motives, to what's going on inside my head and heart? The environment, the law, the way the board and the CEO interact, can help a person stay on the right track. But in the end it always comes down to one's own values. As Buddha says in verse 121 of the Dhammapada, "Do not think lightly of evil, saying, 'It will not come to me.' By the constant fall of water drops, a pitcher is filled; likewise the unwise person, accumulating evil little by little, becomes full of evil."

An individual can be corrupted by what's going on around him, and if you're not careful and suspicious of your own strength, then you will fail. Therefore, I do believe that the social framework in which we live gives us some boundaries and is absolutely needed. But all of these boundaries and all of the laws will be too weak and too imperfect to keep somebody on the right path who doesn't want to walk on the right path. So if you want to cheat, you will always find a way. You might get caught, but many don't. And every time they don't, the tyranny of self-deception gets stronger.