Archive for

Feb 26

Combimatrix Fourth Quarter Update: I don’t think the market has this one right

When you are speculating in small cap stocks that have only 3^rd or 4^th tier analyst coverage (if that), it’s always an adventure when an earnings report comes out.

The market reaction is not always immediate, nor is the first one always correct. Short term considerations; a heavy short interest, a big holder who wants to use the volume to reduce an overweight position, or a long or short pump by a pennystock newsletter, can overwhelm the otherwise expected response.

Unlike large market capitalization companies that generally are well understood and where the news is quickly reflected in the stock price, the sort of tiny micro-caps I play in have responses to news that can be erratic for days or weeks after a news event.

In the face of noisy feedback, you just have to trust in your work.

I first wrote about Combimatrix a little under a month ago. The company performs a number of diagnostic tests centered around reproductive health. Combimatrix has been around a long time and only recently has started to gain traction on insurance coverage with their tests, which in turn has begun to translate into earnings. The reported results last Tuesday.

My thesis centers around the company’s claim that they will be cash flow positive by the fourth quarter of this year. I just don’t think the stock is anywhere near pricing in that possibility.

Combimatrix has a history of burning through cash. The stock trades at a level that says nothing has changed. Even after the recent run-up (when I first looked at Combimatrix it was a $3 stock versus the current price over $4) the stock trades at 0.5x revenue. If they achieve their goal of becoming cash flow positive they should trade at multiples of that.

The fourth quarter results moved them closer to the goal. Revenues were up 32% year over year and 10% sequentially. EBITDA continued its upward march toward the black. They are on track to exceed guidance and be cash flow positive even earlier.

The stock’s immediate after hours response was what I expected. It quickly ran above $5 and my expectation was we would soon see $6.

But that wasn’t to be, at least not yet. After opening a little under $5 the next morning it was pressured the rest of the day and only closed up about 5%. There was more pressure on Friday, it actually got below its pre-earnings close for a short time, but fought its way back to flat by the end of the day.

I don’t know what has held the stock back for the last two days but I don’t think I’m wrong about where its going. Maybe there are shorts making a last ditch attempt to push it down, maybe there is a big holder who has already made the decision to reduce their position into strength and doesn’t care about the results. I don’t know.

The bottom line is the results are good, the company continues to increase revenues, show increases in test volumes for their core tests, and show excellent cost control (G&A, Sales and Marketing and R&D costs are all at levels at or below what they were at the beginning of 2015).

I added to my position. I had already been adding prior to earnings but I added some more. It’s getting to be a big position, especially considering the entire company is only worth about $7 million net of cash. I’m okay with that. I think this is a disconnect in an under-followed, mostly hated, perennially disappointing micro-cap and few are willing to give them the benefit of the doubt. So I will.

Feb 23

Radisys Fourth Quarter Update: Expected and Unexpected

I was bracing myself for a beat down when Radisys reported their fourth quarter earnings. And I was right!

There were two problems with the quarter:

Guidance was poor
The signs that guidance would be poor were ignored by analysts.

Let’s take a look at the signs.

On the third quarter conference call Brian Bronson, the company’s CEO, talked about having low $40’s (in millions) of revenue in the first two quarters of 2017. In January Radisys disclosed an updated credit agreement where they made reductions to the EBITDA covenants for the first and second quarters that were consistent with a level of $40 million of revenue. Finally, if you took in the color around trials from the third quarter call and the Needham conference, you would be led to conclude that material revenue was unlikely before at least the second quarter, more likely the second half of the year.

So it was actually pretty easy to see what the guide would be for the first quarter and full year. I made the following estimates after reading through the EBITDA covenants of the credit agreement (but before the guidance was announced). I shared these thoughts in the comment section here.

My estimates turned out to be pretty close. They would have likely been even closer if it hadn’t been for the fall off in the legacy business (which I will get to in a minute).

Nevertheless analyst estimates heading into the quarter were much higher. Most had $50 million of revenue for the first and second quarter. So when the first quarter number came in lower, the estimates were slashed.

So that was a big part of why the stock has done poorly. But it’s not the only reason. The company also announced news that the legacy business revenue, referred to as embedded systems, would decline more in 2017 than previously anticipated. Embedded systems revenue is now expected to be $55 million in 2017. As recently as January Radisys had said it would be around $75 million.

Its unfortunate. But not crippling. The embedded systems business is not a reason to own Radisys. Its just a distraction.

The reason to own Radisys are the new products and services offerings. And here the company did okay, though not enough to offset guidance revisions and the negative perception around embedded.

My hope for the quarter was a firm new order for DCEngine. Given where Radisys had stated they were in trials I knew this was a long shot. Indeed, no such big announcement came to pass.

Instead they gave a positive but mostly qualitative update. Here’s a summary:

Making progress with their North American Tier 1 communication service provider (CSP) that is in trials with their DCEngine rack – they have three DCEngine units in the lab and Radisys expects commercial revenue could begin as early as the third quarter.
A second DCEngine trial with a customer that will be, in turn, selling to a Tier 1 customer who I am guessing is AT&T, giving Radisys a foot in the door there.
A new FlowEngine use case for an existing customer (Verizon?).
In Europe, an expansion of the existing hardware and services contract with the Tier 1 (this was a contract related to the open-source Central-Office-Redesigned-as-a-Datacenter or CORD initiative that I wrote about in my earlier post on Radisys)
Also in Europe, a new Tier 1 engagement again revolving around CORD.

Overall the company said they were ahead of their plan to have 10 tier one engagements by the summer.

It’s all directionally positive, but as of yet there has been no big win to add to Verizon and Reliance Jio. And its clear that color alone is not going to move the needle unless revenue can be tied to it.

So what do you do from here?

I’ve waffled a couple of times but in the end have added a little to my position. I think the bad news is most likely out. We know from the Needham conference that the European CSP engagements are initially targeting professional services around CORD but that they will likely include a DCEngine component as well (if you go back and listen to the Needham conference again the language is pretty clear: expect that little can be announced by the February earning call but that more engagement with this CSP, around DCEngine, is to come).

We know that the new FlowEngine product will be launched mid-year and in fact a press release just came out today that it is now available for field trials (this is ahead of the end of March date they had previously suggested). We know that Verizon (and likely others?) are waiting on the new version before purchasing more FlowEngine products. Interestingly, Radisys mentioned Gigamon, F5 and A10 as competitors whose space they looked to infringe on with this new product.

There was a press release earlier this month describing a partnership with China Unicom to build open-source PODs for mobile 5G. If this is successful one would expect follow-on orders for DCEngine. China Unicom is a very large service provider.

And finally there was a second press release today describing a new 5G RAN version of Cell Engine. It is probably not coincidence that also today CORD and xRAN announced project alignment. Three large telecoms, AT&T, Deutsche Telekom or SK Telecom, are all mentioned in the release. Radisys mentioned that the new Cell Engine was “developed in close collaboration with a leading mobile operator”, likely one of the three mentioned in the CORD/xRAN collaboration.

Away from the news flow and looking at numbers, when I parse the guide for 2017 it’s not hard to spin it bullishly. We all knew the first and second quarters were going to be poor, low $40s revenue at best. But given $40/$40 (million) in Q1/Q2 the implied Q3/Q4 guide is a range of $55/$55 to $70/$70 (million). These are solid revenue numbers particularly given that they are going to be more heavily weighted to the new products than in the past.

So I’ll wait. Some more. But as so many of the other positions in my portfolio scream higher (Identiv, Combimatrix, Hortonworks, Ichor and Attunity of late) and I still sit with a decent percentage of my portfolio tied up in Radisys, I certainly hope it’s worth it.

Feb 15

Oclaro Second Quarter Update: No sign of the inflection yet

I’ve owned Oclaro since last spring, first wrote about the company here, and subsequently here. Oclaro has their year end in June, so the company recently reported their second quarter results.

Oclaro had another very good quarter, which had been expected after a preliminary announcement two weeks prior. Revenue grew 14% sequentially and 64% year over year. I calculate that adjusted EBITDA was $41 million in quarter, more than double what it was in the first quarter.

Digging a little further, the growth came almost entirely from the telecom/line side of the business. Just to recap (see my earlier posts for more details), Oclaro makes transceivers and laser components that are used for optical data transfer. Oclaro divides this business up between a client side, which consists of shorter length lasers (they have names that start with CFP or QSFP 28), and line side, which consists of longer length lasers of which the most significant is their new CFP2-ACO product. Below is the breakdown between the telecom and datacom sides.

Growth coming from telecom/line means that the aforementioned CFP2-ACO product was responsible for much of it. This is a relatively new transceiver (see here for a bit of background) that was pioneered by Oclaro. Being first to market, they are enjoying a window of 100% market share as the competition races to develop their own versions of the transceiver. The CFP2-ACO was expected to be the growth driver in the quarter.

Competition is going to come into the market beginning in 2017, as the company conceded on the conference call. Acacia, Finisar and Neophotonics is are developing their own CFP2-ACO products.

Still Oclaro has a significant lead and has used their head start to lock down customers for this fiscal year (ending June) and their fiscal 2018. From the conference call:

[We] negotiated or are completing negotiations of several multiyear contracts or extensions contracts for some key product like the ACO that run as long as through calendar year 2018

The CFP2-ACO sales have centered on North America and Europe so far, but Oclaro did say on the call that they are seeing sales from China begin to pick up. China buys a lot of Oclaro products (40% for the quarter), and in the past these have been lower end of the spectrum transceivers (the modulators and lasers that go into the transceivers). China revenues for Oclaro grew 9% sequentially in the second quarter.

While the telecom side drove growth last quarter, the client side products continued their strength of prior quarters but are constrained by production limitations. The lower end CFP product remains sold out (probably in part because of China?). Oclaro hasn’t gotten as much traction on the higher end products (CFP2, CFP4, QSFP28) but they expect this to pick up, and they “do expect to shift [to] the CFP2, CFP4 and QSFP28 transceivers to accelerate through the summer”.

Its worth noting that the CFPX and QSFP28 will have lower average selling prices (ASPs) than the CFP, which could be a drag on top line revenue, but that margins for these products will be higher than the CFP, so the bottom line should improve.

At the Needham conference in January (here is the replay) Oclaro said that the growth of their business relied on 3 end markets: Datacenter, Metro 100G and China. There wasn’t anything on the second quarter call or in the guidance Oclaro gave to suggest any of these 3 markets have slowed yet.

For the third quarter Oclaro guided $156 million to $164 million, which is another sequential increase from the $154 million in the second quarter. To a large degree growth is being driven by the sold-out conditions of many of their products. Management gave some color around how they would like to ramp production faster than they have been able to.

I’ll continue to hold Oclaro. I’ve expressed my concerns in the past about not knowing when this cycle is going to peak. The color I get from listening to the company and its competition is that we are still at least a few quarters off. There is much talk of consolidation in the space, and I would hate to sell just before an offer comes in for the company. So I will continue to management my uncertainty via a reasonable position size, and hold out for a selling price in the mid teens.

Feb 12

3 Comments

Hortonworks Fourth Quarter Update – The hadoop adoption cycle continues

Hortonworks is the second of the three companies I own that reported on Thursday night. The first, Ichor Holdings, I wrote about yesterday. The third, Radisys, I’ll get to shortly.

I originally wrote about Hortonworks along with another company Attunity in November after doing some research on Hadoop and concluding that its adoption presented a good growth opportunity for the companies involved. At the time the stock was trading at $6, had recently been issued a sell recommendation from Goldman Sachs, and was pretty hated all around.

Nevertheless the company was growing like a weed (40% annually). It was also bleeding cash flow like a sieve. But at an enterprise value of less than 2x revenue I found it difficult to pass up the growth. Knowing Wall Street loves those growth stories, I figured a couple solid quarters would put the stock quickly back on its feet.

Fast forward a few months and that is exactly what you got. The company is still growing like a weed (revenue was up 39% in the fourth quarter, guidance was for 28% year-over-year growth in 2017), they are not bleeding cash flow quite so materially (cash flow in the fourth quarter was actually close to flat), and the stock isn’t hated quite so much.

At $11 and with a $1.40 of cash the stock is still trading at 2.4x revenue. So the valuation is actually not that different than when I bought it. One key difference is that back then the cash level was higher (roughly $130 million), the shares outstanding were lower (the company issues stock like toilet paper) and the price per share was lower, so cash was a much bigger piece of the overall valuation and that was partly what I found interesting.

The most interesting thing about the fourth quarter was that growth in their Hortonworks Data Flow platform has really taken off. I wrote about HDF in my original write-up. I’ll repeat how the company described HDF at the Pacific Crest conference last year:

Now [customers] want to have the ability to manage their data through the entire life cycle. From the point of origination, while its in motion, until it comes to rest and they want to be able to drive that entire life cycle. It fundamentally changes how they architect their data strategy going forward and the kind of applications and engagements they can have with their customers. As they’ve realized this in the last year its changed everything about their thinking about how they are driving their data architecture going forward starting with bringing the data under management for data in motion, landing it for data at rest and consolidating all the other transactional data. So it’s a very big mind shift that’s happening.

I still think HDF could be a big growth driver for the company and we are starting to see that. They said on the fourth quarter conference call that HDF grew 6x year over year in the fourth quarter.

So there is lots of reason to think growth will continue. Nevertheless, I am reluctant to add. It’s the cash flow that still gives me pause. While the fourth quarter number was good, they’ve approached break even cash flow in the past only to diverge again into big losses the following quarter. They said on the call that they expect mid-teens negative operating cash flow in the first quarter.

More optimistically, they also said that they expect break-even cash flow in the third or fourth quarter. So that would be a turning point. But then in response to a question about whether we should expect free cash flow after that, if felt like they were trying to scale back expectations:

Yes, I don’t want to talk beyond that yet, Q3, Q4 seems a long way out, but from a – if you think about free cash flow we have been running may be $2 million to $3 million a quarter on CapEx. Q4 was a little light, I think it was under $2 million, but I think once we get to the sort of breakeven number sometime between Q3 and Q4 we will reassess to how much above that we want to punch.

So I’m not sure what to think.

Hortonworks also issues a lot of stock, which while it doesn’t factor into the cash flow number, does dilute shareholder value. The shares outstanding have gone up by almost 3 million in the last couple of quarters.

On the other hand is my experience with Apigee. Another high growth company, with cash flow, that was issuing lots of shares, and the company never really sorted any of that out yet the stock tripled from the $6 price I bought it at to the almost $18 where it was bought out by Google. Hortonworks could easily follow that path.

There are certainly reasons to add. Strong growth, ramping of HDF. They announced another new product launch, enterprise data warehouse in February, and are gaining traction on their Azure and AWS offerings. They also stand to gain visually from accounting changes enacting in 2018 that will allow them to defer less revenue and spread out commission expense, in turn improving the income statement.

Nevertheless my gut, informed by the aforementioned concerns about cash generation and stock issuance, is telling me not right now.

I think if the stock pulled back on a market pullback I would be more likely to add. But it’s hard for me to double up at this price, as I have been prone to do with other ideas that start to work.

So I’ll probably sit with my 2-3% position and watch what the stock does. I prefer to take the safe route when my gut is giving me pause.