Big data software company Cloudera yesterday filed its proposed IPO terms, and the big takeaway is that the top end of its proposed $12-$14 price range ($2.6 billion fully diluted valuation) is more than a 50% haircut from where Cloudera raised VC funding from Intel Corp. back in May 2014 ($30.92 per preferred share). Blame a combination of stock market forces, software market competition and Intel's strategic objectives.
Market multiples changed: Cloudera looks like it wants to price on 5x projected 2018 revenue and 4x 2019 revenue, per Renaissance Capital's model. That's in line with current enterprise software multiples (and better than where things stood last spring), but well below the ≈10x revenue multiples paid for one or two years out back when Intel made its investment.
Outlier: Intel's investment seemed to be as much about hitting an ownership threshold as it was about underlying VC fundamentals. And it still seems to care about that, given its plan to buy up to 10% of the IPO.
Comps (Part I): Hortonworks (Nasdaq: HDP) is smaller than Cloudera ($184m in 2016 revs vs. $261m) and growing slower (51% vs. 57%), but it's still in the same general business. Trouble is that Hortonworks shares have gotten hammered since its late 2014 IPO. That offering priced at $16 per share, quickly zoomed up north of $26 but broke the IPO price in January 2016 and hasn't recovered. It opened trading today at just $10.40 per share.
Comps (Part II): The bull case for Cloudera and HortonWorks (not to mention MapR, which also is in IPO prep) is that cloud penetration into the enterprise still has a very long way to go. The fields are green for ultra-marathon miles. The bear case is that many of those potential enterprise clients already have existing relationships with Amazon, IBM, Google and/or Microsoft ― and each of them can theoretically provide some sort of price-cutting bundle, or at least the peace of mind that comes from working with an existing vendor. Which leads to one more bull case: Cloudera could get snapped up by one of these companies, but it has a history of saying no to such things (even from Intel, and at loftier valuations than the IPO now projects).