I recently posted an article on SeekingAlpha (Oracle’s Sun Acquisition is Paying Off in Spades) that drew from the data about operational leverage that I posted previously on this blog. The SeekingAlpha posting received several good questions, and I wanted to post those questions and my answers here.

In a nutshell, the questions are:

  1. Oracle seems like its obfuscating its financial picture. How do you know that it is making good acquisitions in light of the continuing shrinking in the hardware business?
  2. Your operational leverage argument may not be causal, but coincidental. How do you know the software update segment wouldn’t have done better without the Sun acquisition.
I split edited versions of the questions and my answers in two blog postings since each Q&A is fairly long.
How Do You Know the Sun Acquisition Helped Oracle’s Business?

Your main contention is “The Sun acquisition was good, and we can show that by pointing to the increasing operational leverage of the software update segment since the acquisition.”

I don’t see your argument for a causative connection here. There is a temporal overlap which, in and of itself, isn’t explanatory and might be coincidental. It’s perfectly possible that the software update segment might have been doing even better without Sun.

Thanks for your comments and questions–I’m glad the article was thought-provoking for you.

Regarding causality, you are right that it is hard to draw causal connections with any data, and I agree that one should be careful about doing this. The points you raise are valid and you’ll notice that in the text of the article, I am careful to use verbiage like “suggests” and “likely” rather than “proves”.

To address your question of causality, from a statistical standpoint, the first step is to see whether or not the new data is really anomalous to prior results. In other words, before we show that X has changed due to Y, we should show that X has really changed.

Looking at the leverage data, I would say that there is the strong indication that the results of the last three years have been anomalous in comparison to its prior history:

Estimated Operational Leverage for Oracle’s Software Update Business
Source: Company Statements, IOI Analysis

Looking at the above, it is clear that the firm had fairly stable and maybe even declining operational leverage in the years preceding the merger; after the merger, the measure increased markedly.

There is not enough data in this series to do a scientific comparisons of means with a high statistical confidence, but this kind of study is rare in the analysis of single companies, so we do have to make a bit of a leap of faith. Certainly, looking at the above graph, it would be harder to make the case that something substantive had not happened after 2010 than that something substantive had happened.

So, I am comfortable saying that something in the business likely changed subsequent to the Sun acquisition. The next step is to consider if there is a plausible causal link between the the acquisition and the increased profitability. Indeed, I can think of two:

Bundling Advantages
Oracle may be bundling its software in such a way as to generate greater profits from every sale. This might come from telling Solaris owners:

“You don’t have to buy a new box, but we can sell you an inexpensive upgrade that will help your old box run faster. If you buy your normal db upgrade, we’ll throw in this Solaris software upgrade for just 10% more and you’ll get 50% faster access speeds.”

Obviously, I’m making this conversation up, but this kind of “vertical integration of the stack” mechanism is one of the stated strategic goals of ORCL, so it’s probably taking place in some form or another. This is the first mechanism that hit me as I was looking at these data.

Java Monetization
Along with the purchase of Sun’s hardware lines, ORCL also got access to Java. There is evidence that ORCL is doing a better job at monetizing its ownership of Java than Sun was. Certainly, considering that Sun was mainly a hardware company and wasn’t nearly as concerned with the software side, this improvement under ORCL shouldn’t be a surprise.

A lot of programmers were worried that ORCL’s ownership of Java would mean that Java would be too commercialized. It seems like ORCL has so far been very skillful in walking the fine line between playing nice in the game of the Java Standards Committee and making sure that it was getting paid for the programming resources it was expending on Java.

There is good evidence that the programming community now believes that Java has improved under ORCL’s stewardship, and it is hard to believe that Ellison would not make sure he was getting some economic return on these improvements.

Regarding opinions whether the Sun acquisition was positive or not, there is anecdotal evidence on both sides of the argument.

Some ORCL insiders have criticized the Sun acquisition saying that the hardware business is terrible. Some salespeople are not happy that they’re not getting credit for hardware sales that come about because of their relationship with software clients, etc. These are the kinds of things that take time to work out in an acquisition, so I tend to discount the stories as anecdotal and figure that the issues will be worked out over time.

On the other hand, Ellison has said that the Sun acquisition is the best one he has ever made. He is a software guy and the profits of his company largely stem from the software business. He’s not an idiot–he can see that the paring down of the hardware business is affecting the top line–so he must be thinking that the Sun acquisition has helped his software business somehow. Looking at the data on operational leverage, we see that after the Sun acquisition, the software business is rapidly an anomalously increasing its profitability.

Taking all of this into consideration, I am pretty comfortable in making the contentions I have in this piece. Of course, I cannot be completely sure, and I am constantly looking for proof that my contention is wrong.

Thanks for your excellent question–your point is well taken,