BUSINESS leaders are obsessed with how to begin new assignments. One of the bestselling business books of all time is “The First 90 Days”, by Michael Watkins, which provides a template for newly appointed managers to start with a bang. When Mitt Romney, a private-equity tycoon, ran for the American presidency in 2012 he had a 200-day action plan for turning round the free world, replete with floor plans and flow diagrams. Far less thought has gone into how bosses should leave successfully.

Lord Browne, who ran BP for 12 years, wrote in his memoirs that, as he pondered retirement from the oil giant: “My emotional self prevailed over reason. I did not know how to leave.” Niccolo Machiavelli bungled his exit from the government of 16th-century Florence. “I am rotting away,” he said after being fired. He went on to write a masterpiece on manipulation. But “The Prince” mainly dwells on how to acquire and retain power, not how to relinquish it. To fill the void, Schumpeter has drafted six rules to govern bosses’ departures.

First, the wise executive is neither tardy nor rushed. Sometimes he has no choice in the matter. Hopeless bosses may be forced out fast. Great leaders may be ambushed by fate. Lord Browne left after a newspaper delved into his private life. Akio Morita, the co-founder of Sony, suffered a stroke and Emilio Botín, the patriarch of Santander, a bank, a heart attack. But for those with the luxury of choosing when to go, timing is everything.

The average boss of a big American firm is in the job for eight years. The longer the tenure, the greater the chances of a sticky end. Staying in the post too long greatly increases the chances of making a calamitous error or becoming stale. Some chief executives linger by heading for the chairman’s office: think Larry Ellison at Oracle. Such cohabitation rarely works.

Premature evacuation is no better. Instead wise executives wait until their firm meets its long-term targets, ensuring them a bumper bonus. They are patient enough to manoeuvre their preferred successor into pole position and defuse any ticking time bombs. Sir Alex Ferguson achieved one but not the other at Manchester United, a hitherto successful British football club. He ensured the job of manager went to a friend and fellow Scot, David Moyes. But he left behind an ageing team of declining talents and slumping form. A poor run of results meant that Mr Moyes lasted just ten months in the job. Bosses should behave like a brilliant guest who leaves a banquet while it is still in full swing.

Whether wicked or glorious, all leaders must steel themselves for the emotional shock of their abdication. This is the second rule and the most easily ignored. The unlucky will find that investors cheer—Microsoft’s stock rose by 7% on the day Steve Ballmer said he would step down in 2013. Others will be showered with flattery. But all will see their power recede like a fast-ebbing tide. As Sandy Weill, who created Citigroup, recalled with horror, “I had become a lame duck…I felt numb.” He eventually consulted a psychiatrist.

Even as they indulge the collective’s need for nostalgia and tributes, resolute executives are as unsentimental in their last 90 days as in their first. They retain the counsel of trusted outsiders and focus on the important things, especially themselves. They obey the third rule; to keep a beady eye on their compensation. Most pay-offs are now governed by long-term contracts. Yet boards retain sovereignty over the interpretation of these agreements, including non-compete clauses. Wise chiefs flatter and dazzle the board members until the very end. They negotiate directly, not through lawyers or compensation consultants. They do not ask for petty perks, such as the use of country clubs and cars, that will look idiotic if they become public.

Glory is just as important as treasure and so the fourth rule is to create a narrative about the future and past. Leaders must be seen to be leaving for another great and noble task, and must mythologise their legacy. Winston Churchill recommended writing your own history. The best bosses go one better by writing the history while it is still happening. Lou Gerstner, who ran IBM between 1993 and 2002, published an account of his triumphs two months before he retired. A leader who is still arguing about his legacy years after he has gone has lost control of the narrative, as Terry Leahy, an ex-boss of Tesco, a fallen British retailer, is discovering. Laying the blame for the firm’s slide since he left on Philip Clarke, his (since-sacked) successor, has won him few admirers.

The wise chief will grit his teeth and commit to the fifth rule: do not make big decisions in the last 90 days. Jack Welch, the boss of General Electric between 1981 and 2001, put together one of the most elaborate succession plans in corporate history, which included helping the two runners up find jobs at other companies. But he botched his departure, postponing retirement to oversee an ambitious takeover bid for Honeywell, a huge rival, which then collapsed.

When Harry met Dwight

Their legend secure and treasure-chest full, cunning leaders should obey a final rule: ensuring that the next occupant of the job does not outshine them. If they despise their successor, departing bosses can block access to a firm’s important clients, suppliers and investors, and fail to explain vital procedures. President Harry Truman budgeted only half an hour to brief his successor, Dwight Eisenhower. The meeting lasted 15 minutes.

Even if they love their replacement, sensible departing bosses will keep back some information about their once-and-not-future subjects: the malcontents and the treacherous, the astute and the virtuous. They should make a vague offer of advice, safe in the knowledge that their successor will rarely, if ever, accept it for reasons of pride. And then they should walk out and never look back, praying that the new boss is successful—just not as successful as they were.