During this decade, a lot of companies have been forced to go through a transition period. Legacy businesses are dying and new growth drivers are taking their place. IBM (NYSE: IBM), Microsoft (NSDQ: MSFT) and Oracle (NYSE: ORCL) are just three that come to mind immediately.

IBM and Oracle probably have it the worst because of their unique predicament of being caught between their declining-too-fast hardware and software legacy businesses and their not-growing-fast-enough cloud divisions. Both are yet to prove to investors that there is a clear path to higher revenues and, thereby, to higher valuations.

Today, I’d like to talk about Oracle. Their core business was almost entirely dependent on the on-premise IT infrastructure model that nearly every company of scale employed before being introduced to the cloud. The whole world operated on that model, making Oracle the king of databases. They still are, but a growing Cloud IaaS (Infrastructure-as-a-Service) industry led by Amazon (NSDQ: AMZN), Microsoft and IBM are forcing Oracle to take a different stand.

The Oracle of Old

There was a time when Oracle CEO Larry Ellison ridiculed the very concept of cloud, calling it “gibberish.” In hindsight, why wouldn’t he? After all, he was talking about a new technology that could potentially threaten the apparently immovable force that was Oracle’s enterprise business. It was a reflexive defense mechanism; a forgivable one; but not anymore.

The world doesn’t stop moving because of one man’s opinion. And this was the fallout of that opinion: a top line that nearly doubled in the years between 2007 and 2012, and then struggled to maintain that $37 to $38 billion level for the next several years.

In fact, the revenue slowdown was so bad that the company was jolted into action and quickly jumped on the cloud bandwagon. Their focus is now so intense that they’re probably the only company of their size to report their cloud earnings in such a granular manner, specifying incomes from SaaS and PaaS, as well as IaaS, as standalone numbers.

The Oracle of Today

The good news for Oracle investors is that their cloud business is indeed growing. Total SaaS and PaaS revenues have grown nearly 48%, from $1.86 billion in 2015 to $2.75 billion in 2016. This is even more significant when you see comparable growth among the cloud leaders Amazon, Microsoft and IBM.

Unfortunately, there’s a downside to that, and it is this: cloud revenues as a whole only accounted for 8% of the company’s top line in fiscal 2016. Moreover, it is the only component that is growing at such a rate, while every other reporting segment shows negative growth.

Of course, Oracle isn’t going to suddenly lose the bulk of its enterprise customers to other cloud players because legacy businesses are expensive to get rid of. What will likely happen is that this slow bleed will continue indefinitely as more and more clients move to the cloud. The only way to survive in this scenario is if they’re able to successfully transition existing clients from on-prem to off-prem, or to on-prem hybrid the way IBM is doing.

Source: IBM Big Data Hub

The Investment Case: Mind the Gap

There’s obviously a gaping chasm between what they’re earning from older business units and new growth drivers. The question now is whether they can keep growing cloud revenues at the same 50% level into the next several quarters. That’s the only way they can keep their overall losses to a minimum.

We’ve already seen that the gap currently equates to 3% negative growth. Oracle can no longer permit such losses to occur. The guidance for FY 2017 seems to sum up that need:

“Over the full-year for FY 2017, I expect SaaS and PaaS revenue growth will be higher than the 65%, up from 62% in FY 2016. SaaS and PaaS gross margins should exit Q4 FY 2017 much higher than the 57% reported today, as we show steady and continued improvement through the year.”

If they can achieve this over the next four quarters, it will be a very important milestone for the company. Of course, even if they manage to double their cloud revenues every two years, they still have about four years to go before those can match their current software and hardware revenues combined. I honestly don’t see that happening for another 5-7 fiscals at least - and that’s if they keep growing their cloud business at a CAGR of at least 20%-30% during that entire period. The market potential is definitely there, but whether Oracle will seize that opportunity is the question now.

This is my opinion: HOLD for now until you see at least two quarters with zero overall loss. That’s the only way to be relatively sure that legacy decline is being matched by cloud growth dollar for dollar.