Sep 18, 2014 · 4 minutes

When the Roman Emperor Tiberius had to choose a successor, he actually chose two: His grandson Gemellus and the notorious Caligula. History books tell us that Tiberius was being shrewd: He knew he was widely despised as a ruler, and had an idea that Caligula would end up in charge and leave Rome with an Emperor that, by comparison, left his legacy looking pretty good.

The news of Larry Ellison handing over the CEO role at Oracle to two successors (the release refused to use the infelicitous phrase “co-CEO”) made me think of that piece of ancient history, not because there is a Caligula awaiting to take over the enterprise-software giant, but because I couldn't help but wonder: Is all this an elaborate plot to make Larry Ellison's legacy look good?

On the one hand, it's a crazy thought. No sane CEO would ever consider such a thing. On the other hand, there's a crazy logic to it. Recall the impression, widespread as recent as a few months ago, that Tim Cook was leading Apple astray. It made Apple and Cook look bad, but the legacy of Steve Jobs shined that much brighter in retrospect. This month has shown Cook is leading Apple capably, but for a while the worse Cook appeared, the better Jobs looked.

A lot of the initial reaction to the Oracle transition today focused on the co-CEO idea, which rarely works well. When it does work, as with Whole Foods Market and Deutsche Bank, one of the CEOs is a founder of the company. In Oracle's case, it's the co-founder who is stepping down. In his place a former investment banker (Safra Catz) and a former executive of an Oracle rival (Mark Hurd) are taking his place.

In most other cases, the co-CEO thing is problematic. At Research In Motion, joint leadership worked until it didn't. Once the company faced a true crisis, the co-CEOs were slow and hamstrung in responding to things quickly and the Blackberry franchise fell into a tailspin. Oracle rival SAP had dual CEOs, until the company decided to just choose a single leader.

Beyond the dual CEOs, there's also a crazy-quilt structure to the new arrangement. Ellison is becoming Executive Chairman of the Board, bumping longtime Chairman Jeff Henley to Vice Chairman. In Oracle's release, Ellison said, “"Safra and Mark will now report to the Oracle Board rather than to me." Which sounds good, until you recall that the board is run by Ellison.

Ellison, long one of the most headstrong personalities in Silicon Valley, is also staying on as CTO. And so the reporting structure isn't split two ways, but three. As the press release explained, “All manufacturing, finance, and legal functions will continue to report to Oracle CEO, Safra Catz. All sales, service and vertical industry global business units will continue to report to Oracle CEO, Mark Hurd. All software and hardware engineering functions will continue to report to Oracle Chairman and CTO, Larry Ellison.” It's as if Tiberius was staying on to co-rule his Empire with his two successors.

Ellison maintains that the new structure allows the current management team to stay in place, and if everything were going well at Oracle, the plan might work. But the company is in the midst of a long-term, secular change in the enterprise software industry that is working against incumbents like Oracle.

Two years ago, Sarah Lacy wrote a piece detailing how Oracle was vulnerable to a long-term shift from the old enterprise-software market – where the best product mattered less than stability of existing wares – to one where customers could expect, even demand, quality and innovation in their business software. Since then, that shift has become more evident, and Oracle more vulnerable.

During the past three years, Oracle's stock has risen 42 percent, lagging signnificantly the S&P 500's 65-percent gain. Over its past three fiscal years, Oracle's revenue has risen a total of 7 percent. Yet its earnings per share has more than doubled to 48 cents a share from 21 cents a share. Some of that has come from cost cutting (operating profit has risen 23 percent). But buybacks have also reduced the number of outstanding shares by 10 percent, pushing up the EPS figure.

Some Oracle critics have noted the company is spending vastly more on share repurchases than on capital spending. Over the past three years, the company has spent more than $20 billion buying its shares back from the stock market. Lost in the news of the executive changes was Oracle's announcement that its board has authorized the company to buy back another $13 billion in stock.

The buybacks have blunted the impact of Oracle's business slowdown on its stock price. But it's becoming harder for investors to ignore. The 62-cents-a-share profit Oracle reported today missed analyst expectations by two cents a share. That marked the third straight quarter Oracle has fallen below Wall Street's consensus estimate. And that's after billions in stock buybacks.

The 3-percent drop in Oracle's stock following today's news likely had less to do with the management changes and more to do with that miss. Oracle also said its earnings and revenue guidance for the rest of the year would be below what analysts had been expecting.

It's not so much that Oracle is refusing to adapt to the cloud economy – it adapting slowly but surely. It's more that the cost-saving potential of new enterprise wares mean the overall pie is shrinking, even as more companies with new business models ensure that the industry keeps evolving quickly. And that raises the possibility that, despite Oracle's longtime influence in shaping the enterprise software market, the company's best days are behind it.

Whether Oracle has peaked or not, Ellison's own legacy will endure. If the company continues to thrive, he can take credit as Chairman and CTO. If it keeps slipping further into decline, he knows that it's the CEO (or CEOs) that will take the blame. And Ellison the founder will look all the more brilliant by contrast.