Can Oracle survive Larry Ellison?

As the storm raged about him, claiming the lives of six sailors on other boats, the Oracle chief executive relinquished the wheel to a professional sailor.

"I couldn't handle it," he later explained. "There were no stars; I didn't know what to steer by. I had to turn the wheel over (to someone else). You discover your own limits."

You could say that Ellison finds himself in a similar position today, metaphorically speaking. Still at the helm of the company he founded almost a quarter-century ago, the flamboyant CEO faces heavy turbulence in Oracle's two primary lines of business.

In its database stronghold, Oracle is seriously challenged by IBM and Microsoft. In business applications, Oracle's hostile bid for PeopleSoft appears to be failing. PeopleSoft has completed its acquisition of J.D. Edwards and regulators on both the state and federal levels are closely scrutinising Oracle's attempted takeover. And within Oracle, Ellison has few seasoned professionals to whom he can turn.

His senior management team is woefully depleted, with the loss over the past several years of Ray Lane, Gary Bloom, Robert Shaw, Randy Baker, Polly Sumner and many more. Not only is there no one at the top to challenge the often-mercurial Ellison, there is no clear successor to take over in an emergency.

So Oracle's stakeholders--customers, partners, shareholders, and employees--are at the mercy of a nearing-60 CEO who indulges in high-risk behaviour and whose interest in his company is fitful. Two years ago, he told the Oracle AppsWorld conference that if he had to do it over again, he'd probably go into genetic engineering rather than computing. After three generations of computing (mainframes, client-server and Internet), "there will be no new architecture for computing for the next 1,000 years," he proclaimed. "The computing industry is about to become boring."

Scarcely encouraging words from a CEO whose company is now extraordinarily vulnerable to his whims.

Even if Ellison's energy level and commitment to being CEO remain high, no lone executive, no matter how capable, can single-handedly run an operation as complicated as Oracle and also focus on how to reposition the company for the future. While competitors like IBM and Microsoft smoothly transitioned their leadership from, respectively, Lou Gerstner and Bill Gates to Sam Palmisano and Steve Ballmer, and built up teams around them, Ellison coyly toyed with executives like Lane and Bloom.

It's becoming obvious that most of Oracle's problems are internal, related to its loss of management at the top, its alienation of everyone from customers to partners, its conflict-ridden culture that sucks energy into the black hole of corporate politics and, last but not least, the flawed personality of the Oracle himself, Larry Ellison.

It's unlikely that Ellison would ever be forced out of Oracle, since he owns one-fourth of its stock and has a relatively weak board of directors. But, given his propensity for fast jets, fast cars and sleek yachts, some kind of accident is possible. Or he could decide that Oracle has become boring and simply leave to start a biotech company. It is just that kind of scenario that causes the CEOs of most major companies to do succession planning.

But not Ellison. One wonders if he can even conceive of Oracle's existence without him.

"If Larry was incapacitated, the cult would dissolve," former executive Marc Benioff says. "It's unclear if Oracle is a sustainable enterprise without Larry, because his personality is so firmly entrenched." Even if Ellison leaves voluntarily, whoever replaces him would face monumental challenges.

He or she would simultaneously have to fix a troubled company, overhaul a counter-productive culture and fill the shoes of one of the world's most high-profile corporate leaders.

"Oracle has had its best days," Meta Group analyst Will Zachmann maintains. "There's no way they can grow like they did in the 1990s. More likely, they go into a decline. If they lose their core (database) business, the rest of it will crumble."

Analysts note that an aggressive, expansionist management style works, as long as you're winning. It's less clear how well that works when a company like Oracle is in a defensive mode, in which it's not leading a market. One conclusion is clear: There are no more quick victories.

But Oracle will need to come perilously close to running aground, as happened in 1990, before Ellison even considers relinquishing the helm, or the board of directors gets enough of a backbone to pry him from it. He gave up the wheel voluntarily in the Sydney-to-Hobart yacht race, but only because lives, including his own, were in jeopardy. It may be that he's no longer capable of letting go of the company that he formed--and that now defines his very being.

With or without Ellison, his company's future seems so murky that even the ancient Delphic oracle could not predict it.