Oracle's plans for PeopleSoft have never been harder to read than they are now.
For most of the past 16 months the chances of Oracle taking over PeopleSoft have been firmly in the "unlikely" category. The expected opposition of anti-trust regulators was one factor, but the biggest obstacle for Oracle was the resistance of PeopleSoft's management.
While PeopleSoft was led by a chief executive officer (CEO) claiming the company was not for sale at any price, Oracle had little hope of successfully acquiring PeopleSoft through its tender offer to shareholders, now valued at $7.7bn (£4.3bn). Thanks to PeopleSoft's poison pill, the company's board of directors can essentially deflect Oracle's advances for as long as they remain in office.
But the dynamics of the situation between Oracle and PeopleSoft have changed from stand-off to detente with the firing of Craig Conway as PeopleSoft's CEO and the testimony by PeopleSoft director Steven Goldby this week that he would be open to discussions with Oracle if the price were right.
So how likely is the deal to be completed? On the one hand, David Duffield, the new CEO selected by PeopleSoft's board, strikes long-time customers and observers as an odd choice if the board hopes to quickly strike a deal with Oracle.
As one of PeopleSoft's founders, Duffield has a reputation for being deeply committed to PeopleSoft's customers, who overwhelmingly oppose the deal.
On the other hand, with shareholder interest building in the cash Oracle is offering and with anti-trust blocks to the deal nearly gone, PeopleSoft's board may not be able to continue its resistance.
Goldby said he would consider negotiating with Oracle "if there is ever an indication that Oracle is willing to pay what we consider to be the right price for the shareholders to get for this company, and there is a high certainty of being able to close a transaction quickly."
The Yankee Group said in a research note that it believes PeopleSoft's executive team "is clinging to the hope that Duffield can muster support to prevent shareholders from voting in favour of the buyout, but is largely resigned to the takeover".
Oracle has asked a Delaware court to void PeopleSoft's anti-takeover "poison pill" provision and its "customer assurance programme" promising customers compensatory payments if PeopleSoft is acquired by a company that disrupts its product development and support plans.
The poison pill provision in PeopleSoft's bylaws allow its board to inflate the company's outstanding shares total and make a hostile takeover prohibitively expensive.
On the first count, Oracle faces long odds, experts say. Delaware courts have not overturned a poison pill since the late 1980s, according to William Lawlor, a Dechert partner who leads the firm's mergers and acquisitions group.
But if Oracle opens discussions with PeopleSoft's board and wins its support for the acquisition, the poison pill issue becomes moot. How likely is that? It's impossible to tell.
While Goldby, in his testimony to the Delaware court, seemed to open a door to Oracle, he also reiterated that PeopleSoft's board considered and rejected as inadequate the $26 per share cash offer Oracle made earlier this year, before Oracle lowed its bid to the current $21.
He also emphasised that decisions about the Oracle bid are made by a five-person committee of PeopleSoft's independent directors. While Conway, and now Duffield, could comment on Oracle's offer, the CEO has little direct influence over whether it will be accepted - meaning that the resistance so far to the bid has come unanimously from all of PeopleSoft's directors.
Oracle could advance its bid by raising its offering price, but there's only so high the company can go before it risks alienating its own shareholders. Financial analyst Chris Kwak, of the Susquehanna Financial Group, said that if the paying price creeps much higher, Oracle would generate better value for its shareholders by spending the billions it would pay for PeopleSoft on a stock buyback instead.
"A lot of people believe Oracle will have to raise its bid," Kwak said. "Our view has been that $21 [per share] is a pretty generous offer.
"It may be that the intrinsic value of PeopleSoft is less than $21 [per share], but the value of PeopleSoft to Oracle is more than that, if they are successful in finding synergies and reducing headcount. Still, if the price gets too high - beyond $23 or $24 [per share] - Oracle would do better to back away and buy back its own stock."
As PeopleSoft customers continue to face uncertainty about the future of their supplier, the best thing they can do is update their software to the latest versions to protect their investment if PeopleSoft's development is disrupted, advises Meta Group analyst Barry Wilderman.
He gives the acquisition an 80% chance of happening at this point, but whether or not it would be a good move on Oracle's part depends on how well it supports PeopleSoft's customers and how many they retain, he said.
"Some people will migrate to Oracle's applications, but some will go to rivals like SAP, and some will stay on PeopleSoft and go to third-party providers to maintain the applications," Wilderman said. "If Oracle can carry over, say, 80% of PeopleSoft's customer base, this is a good deal for them. If it's 60%, it's not a good deal."
Customers relying on PeopleSoft's customer assurance programme to protect their investment may be out of luck: Lawlor said that while the Delaware judge is unlikely to void PeopleSoft's poison pill, he might "split the difference" and invalidate the programme.
"Oracle has put together a pretty good case," he said. "They may get a sympathetic hearing from the judge. If I were betting, I would say he may very well strike down the [customer assurance programme] as being an overly draconian measure. It's a little muddy how the court would fix the problem there, but I think there's a good chance he will look at it."
Stacy Cowley writes for IDG News Service