Axcelis: Rock Solid Financials, Excellent Outlook, And An Exceptional Entry Price

| About: Axcelis Technologies, (ACLS)

Summary

Sales, earnings and free cash generation will resume their step function higher in the June quarter after a two-quarter pause attributable to Samsung’s spending pattern.

Q1 will be the low quarter with the upturn magnified by a just-beginning, multi-quarter sales ramp with new customers Micron and Hynix.

Operating income is highly leveraged to revenue and virtually all is free cash. Free cash yield from Q2 forward will annualize in double digits with the stock below $3.

In Q2, the company will reverse split its stock which I believe will prove highly beneficial to shareholders.

The company has $80 million in cash, $300 million in NOLs, and is likely to be acquired by Lam Research once Lam’s pending acquisition of KLA-Tencor is fully integrated.

Note: My original article on Axcelis is here. It provides background information for those new to the name.

Price action in Axcelis (NASDAQ:ACLS) since the February 2 earnings call has been mystifying. The stock sold off more than 10% intraday on heavy volume on February 3 even though Q4 results had been pre-announced and Q1 guidance was in line. Financials are rock solid and the operating outlook is excellent but the volatility is unnerving if one isn't familiar with the company and its prospects. So let's review Axcelis' prospects because I believe the stock will triple in the next 18 months and you should own the shares.

It's likely the recent price weakness is attributable to (1) a former large, activist holder dumping the remnants of its position, and (2) a fanciful - and utterly inaccurate - article on a Korean website with the provocative headline, "Samsung (OTC:SSNLF) Shelves Expansion of Semiconductor Plant in China." I've provided more details in the final section.

Business Model

Axcelis has a hockey stick business model. That means operating expenses are largely fixed and earnings are primarily a function of gross profit, which is the product of revenue and gross margin. Importantly, both revenue and gross margin percent will increase quarter by quarter in 2016. This is the step function referenced above.

The reason operating expenses - R&D and SG&A - will barely budge as revenue increases is because costs have been wrung out - down 13% over the last three years while revenue is up nearly 50%. That's an amazing achievement. Further, the company believes it can support another 50% increase in revenue with no more than a 10% increase in operating expenses.

Gross margin was under 30% three years ago, is 34%-35% currently, and is guided to 36%-38% by Q4 of this year. Here is the CFO talking about gross margin on the February 2 earnings call: " ... we expect this year to exit somewhere in the high mid-30s... right now based on where we are with our improvement roadmaps we are on plan to achieve those gross margins (regardless of) the revenues this year ...." (emphasis added).

Further, the company expects a gross margin of 40% in 2017 when several large greenfield projects are scheduled. I'm referring to average gross margins, of course. For the average to rise to 37% this year and 40% next, the gross margin on incremental revenue has to be well above 40%.

Here is what Axcelis' hockey stick operating model looked like in 2015. (I've rounded to whole numbers for ease in reading but the margin % in the bottom row reflect reported results.)

(in millions)

Q4 2014

Q1 2015

Q2 2015

Q3 2015

Q4 2015

Revenue

$63

$73

$78

$79

$70

           

Increase vs Q4 2014

Revenue

$11

$16

$17

$8

 

Oper. Income

$3

$7

$8

$2

           

Profit Margin on Incremental Revenue vs. Q4 2014

 

29%

43%

44%

25%

Click to enlarge

In Q2 and Q3 when revenue approached $80M, more than 40% of revenue dropped straight to EBIT. Keep this hockey stick model in mind because the current flattish quarter is probably the last one below $80M for as long as Axcelis remains a public company. Q2-Q4 are likely to average near $90M and in 2017 quarterly revenue should easily top $100M, when multiple large projects are set to launch.

Note, too, that the company has NOLs (net operating loss carryforwards) from years ago of $300M and will not pay taxes until it has earned in excess of the current market cap.

Revenue prospects - better than the market expects

Axcelis guided revenue down mildly to $65M for Q1 and will not comment on full-year prospects until mid-year but they did share their expectations for full-year implant market share. They are confident of attaining mid-20s% share in 2016 and mid-30s% in 2017 due to strong momentum with new customers. We can work with that. I am assuming 25% share in 2016, which I believe is low considering that Hynix and Micron (NASDAQ:MU) are coming on as new customers.

Analyst: "... you believe you're going to pick up further market share in 2016. Is that accurate?"

CFO: "Yes. That's exactly what we are saying."

Business remains solid. The company confirmed it has no indication of order pushouts from any customer as of mid-Q1 and "no one is slowing down." The relentless declines in energy prices and interest rates bring about silver linings elsewhere.

In broad terms, the market for implant systems is $900M plus/minus 10% per year. There are a number of variables that make it difficult to pin down a specific number but the industry pays attention to Gartner's estimates so I will, too. Gartner estimates that 2015 implant sales were $950M-$975M, but will drop 10% to $850M-$875M in 2016. That seems overly steep when industry participants such as Lam Research as well as other consultants are expecting semiconductor capital equipment spending to be flat to modestly higher this year, but let's go with 25% of $862M as the base case for implant system sales this year.

In addition to sales of new systems, Axcelis has a sizeable, steady aftermarket business (upgrades, spares, used tools, maintenance). Aftermarket revenue is steady at about $30M per quarter, with small variations related to sales of $1M-$2M used tools. Let's call it $120M annually.

Putting the two together, base case 2016 revenue is $335M, up 11% from 2015 with two new world class customers ramping up over the next few months and scheduled to take deliveries throughout the year.

Recall Axcelis guided Q1 revenue to $65M. First, be aware they routinely issue conservative guidance and then exceed it as a matter of investor policy. Second, they likely have allowed themselves extra wiggle room this quarter because the bulk of purchase orders for shipment in late March runs the risk of a slip into Q2. At $4M-$5M per tool, a single slip makes a difference. I'll be surprised if the reported number is less than $70M.

If we take Q1 revenue guidance of $65M at face value, it leaves $270M of our $335 base case revenue to be realized in Q2-Q4. That's an average of $90M per quarter. This is what the market is missing, and the sell side won't write about, because the company won't comment on Q2 or the full year until the May earnings call.

Here is what $90M revenue quarters in 2016 will look like.

(in millions)

Q4 2015

Q1 2016

Q2-Q4 2016

Revenue

$70.5

$65-$70

$90

Operating Income

$2.4

$2-$3

$12.4

       

Change vs Q4 2015

Revenue

flattish

$20

 

Oper. Income

flat

$10

       

Profit Margin on Incremental Revenue vs. Q4 2015

 

n/a

50%

Click to enlarge

Operating income of $12.4M per quarter represents an EPS run rate of $0.36. More importantly, it represents a free cash flow run rate of $46M-$48M, which is a 12%-13% free cash yield on market cap with the stock at $3.00 and an absurd 24% free cash yield on enterprise value at the current price of $2.25. This is true free cash - EBITDA less interest, taxes ($0 due to NOLs) and capex. The company could be taken private in a heartbeat at the current price.

The two crucial points to realize as we move into Q2 are that half of incremental revenue above $70M per quarter will drop directly to operating income and that sell-side consensus estimates of down revenue for the year are way off base.

In defense of the analysts, they haven't yet received any guidance on 2016 from the company and have little incentive to spend time modeling the financials because their institutional clients aren't going to touch a $2 stock. It's one thing for you and me to do our own work and decide to buy shares of a penny stock, but when one manages other peoples' money the priority is to retain the account and its fee income. One sure way to raise a question of professional competence in a client's mind is to put him in a few $2 stocks, which brings me to my next point.

What sort of management proposes a reverse split of its own volition?

Short answer - a smart one.

Longer answer, one that is serious about bringing its stock to the attention of a larger pool of investors in expectation its valuation will be re-rated.

Most institutional investors don't research/cannot invest in stocks that trade below $5. They may be prevented by a fund mandate, by internal rules, or by a general bias toward penny stocks of low quality, illiquidity, and inability to put enough money to work to justify the research effort. Those factors lead to pricing inefficiencies and present opportunities for us.

The point of the 1:4 split Axcelis will implement in Q2 is to lift the price to a level that facilitates discovery by institutional buyers, which should lead to greater liquidity and a higher valuation. In addition, the share count will be brought in line with peers, which itself has been a potential flag.

There may be a psychological aspect to valuation in play here, too, because we are all more accustomed to dealing in dollars than pennies. As an example, see if these two identical versions of my base case for Q3 2016 have equal intuitive appeal to you.

- A $2.50 stock with 75 cents in cash and earning at a run rate of 36 cents.

- A $10 stock with $3 in cash and earning at a run rate of $1.44.

I would look at the $10 version immediately every time and put the $2.50 version on my to-do list, where it might remain for a while. The low price itself is an impediment as it implies a need for extra due diligence.

If you have experienced a typical reverse split, as I have, the lesson learned was "sell immediately and don't look back." That's because the typical reverse split is an act of desperation by a financially troubled company to avoid delisting. Unsurprisingly, its stock continues to do poorly post-split. However, the smaller subset of companies that are growing, have strong cash flow and balance sheet, and a sensible explanation for their action - which describes Axcelis exactly - usually have excellent stock performance post split. One that I have experience with is NewMarket (NYSE:NEU), which split 1:5 in 2002 and has been a spectacular money maker ever since.

All of which is to say, I'm very much in favor of Axcelis' reverse split.

Looking ahead

2016 is just a springboard for Axcelis. In 2017, several big projects are scheduled to launch that are implant intensive. It would not be surprising for implant system sales to reach $1 billion and for Axcelis to capture 40% share and achieve 40% gross margin. That works out to EPS of $1.00 (untaxed) but I doubt that number will ever print because I expect the company to be acquired by Lam Research before the end of 2017.

The next several years are expected to be a time of heavy investment in non-volatile memory, which is implant intensive. (Non-volatile memory enables the "instant on-off/your work is always saved" feature of smartphones and tablets that soon will be standard nearly everywhere).

Lam Research/KLA-Tencor (NASDAQ:KLAC) and Applied Materials (NASDAQ:AMAT), the two $20B market cap, $10B sales gorillas that dwarf all other competitors, will be battling to supply end-to-end solutions for multi-billion fabs. However, Lam has a hole in its end-to-end offering. It lacks implant capability.

There is only one solution for Lam because implant is a two-company oligopoly. Applied bought its implant capability in 2011 when it paid 5x sales plus the cash on the balance sheet to acquire Varian. In comparison and though smaller than Varian, Axcelis sits at 1x trailing sales, with new technology superior to Applied/Varian, rising sales, rising market share, and $0.65 in cash that increases every quarter. It's a sitting duck.

Perspective on the recent price weakness post-earnings

At the close on February 2, the day of the earnings call, ACLS was unchanged since year end despite the sharp decline in the tech sector and the broad market. However since the call, price action has been atrocious even though there were no surprises in Q4 results or Q1 guidance.

I believe two company-specific, non-fundamental events, both of which are now past, put short-term pressure on the stock and sowed confusion - but have created an exceptional entry point for investors.

1. Vertex Capital is a probable seller since the call, for reasons unrelated to earnings prospects or valuation. It's likely they see better opportunities for their brand of activism after the Axcelis CEO reiterated on the call there are no plans for a share buyback even though the company is holding excess cash. Vertex has been drumming on Axcelis to buy back shares ever since it filed a 13-D disclosing 7% ownership in September 2014. Axcelis refused to go along and Vertex reduced its position to 4.9% last May, after which it was no longer required to report on its holdings. Further, the standstill agreement between Axcelis and Vertex expired two days before the earnings call, serving as a prompt to mount a proxy fight or to move on. I suspect they opted for the latter and blew out their remaining shares over the last week.

2. This article appeared on a Korean website in the wee hours U.S. time on the day after the earnings call. It is sobering at first glance because this is how chip industry downturns start, with order pushouts and expansion delays. A story like this will trigger a "sell now, investigate later" reaction, especially during a time of extreme selling pressure in the tech sector. There is just one problem with the article - it is completely incorrect. The plant that is the subject of the article was designed from the start to produce 100k wafers per month, as it is now doing, and Samsung has never had plans to expand it. All of Samsung's expansion plans are focused on a huge site it owns in Pyeongtaek, South Korea. It has instructed suppliers to begin putting infrastructure in place because it plans to build multiple fabs on the site in coming years. Ground breaking occurred last May and production is scheduled to begin in early 2017, although Axcelis thinks it may be brought forward into Q4. Should that happen, my current 2016 earnings projection will look downright timid.

Disclosure: I am/we are long ACLS.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.