Hortonworks: Scratching my head

First, I’m sorry I haven’t weighed in on 2018 Q1 for Hortonworks (HDP) before now. I’ve had people ask me about it on and off board, and I was hoping to get back to all of you sooner, but here we are.

The quarter was mostly fine. Subscription revenue was up 46%, which was lower than it has been recently, but obviously still very good. Customer count was up about the normal amount, net expansion rate stayed high (120%), but the deferred revenue was a head-scratcher. Total deferred revenue actually decreased from 275 million to 250 million. It’s not as bad as it looks, because it was mostly due to an accounting change, but even without that it would have been down a little - I think they said 4 million - whereas it’s usually up more than 10 million. They were asked about it on the CC and the response was simply that instead of deferred being up, actual revenue was up. Yes, the 79 million recognized this quarter was higher than any quarter before now, but the point is, that was expected already. And it was probably also expected that deferred revenue would increase as well. So I think it’s fair to say that total growth was slower than expected this quarter.

Deferred revenue doesn’t “go away.” They just recognize it. So of the $79 million of recognized revenue in Q1, let’s make up a number and say $20 million had been sitting in deferred revenue, and the other $59 million just came in this quarter (those may be waaaaaaaaaaaay off – I just made them up). That would mean they only added about $16 million of NEW deferred revenue in Q1. Whereas in the March 2017 quarter they recognized $56 million, but also saw an almost $13 million increase in deferred revenue. So I believe that means total revenue was $75m this quarter vs $69m last year. Not a rosy picture of growth. However, I could be wrong, or they could bounce back next quarter with a huge increase in deferred, or the accounting change could be having some other effect that I don’t understand…any number of things could be happening. It may not be the end of the world. But it was weird.

The flip side was that they were cash flow positive yet again – that’s 2 quarters in a row! Operating Cash Flow was 8 million, even more than the 6.4 million they turned in for Q4 2017. However, even here I have to scratch my head, because it appears the main driver for this cash flow was accounts receivable. You can only tap that well for so much, so I don’t expect they’ll be cash flow positive every quarter unless deferred revenue increases again.

Another point in their favor is that OpEx was actually down sequentially, and only up 5.5% YoY. I try not to get too excited about that, because they’d have to cut it pretty drastically for it to make a big impact, but I do think it’s very positive that they can keep it in check. It’s actually within 10% of what it was in the March 2016 quarter, two years ago.


I do think there’s a margin of safety here. The good part about the $250m of deferred revenue they have is that they are almost guaranteed to be able to recognize more and more each quarter. So recognized revenue will grow smoothly even if it’s slower. And also, at a PS under 5, slower growth may still be enough to move the stock in the right direction. But the growth concerns related to the deferred revenue issue bothered me enough that I decreased my position. I am still hopeful they’ll bounce back next quarter with faster growth. But if growth slows, I’ll be out. For now I’m going to watch and see.

Hope this helps.



I know that I posted a while back that I was out of HDP, bu I bought back in at 16.81. They have a forward EV/S less than 4, are growing well, positive cash flow. I stick with my original conclusion back when i originally wrote up HDP. If they can continue to grow and continue to prove they can execute then eventually they will have a higher ev/s. Right now there is so much negativity around hadoop that I think there multiples are unfairly low. This should get figured out one way or another over the next year or so.


oh look at that, I guess it is my 10+ year fool anniversary.


Thanks for the analysis. I’ve been thinking about Hortonworks - I also picked up some Cloudera on the drop. Hadoop is still the best tech for big data, and big data is still exploding, so I think the current negativity on Hadoop is temporary.

Cloudera is what I have seen in the workplace, but Hortonworks is nicely integrated into all the clouds and has a good reputation. I’m taking initial positions in both as we see who gets better traction in the enterprise