POST OF THE DAY
ARM Holdings plc
ARMs Longa, Vita Brevis

Format for Printing

Format for printing

Request Reprints

Reuse/Reprint

By Scollag
September 7, 2004

Posts selected for this feature rarely stand alone. They are usually a part of an ongoing thread, and are out of context when presented here. The material should be read in that light. How are these posts selected? Click here to find out and nominate a post yourself!

I've been doing some tyre-kicking on the Artisan deal, trying to work out how good Artisan really is, and what ARM see in the company, both before and after the proposed takeover.

So far, I haven't listened to ARM's presentation: I do intend to listen to it presently, but I wanted to make some notes on my own impressions first. I may well change my mind one way or another after hearing what ARM have to say. Still, if we are to approve this takeover (I am long ARM), I am inclined to think the deal should stand on its own obvious merits. The following case, then, is based on what I think ARM see, rather than what ARM say they think they see.

My first impression of Artisan, after poking around some of the nooks and crannies of their SEC filings, is favourable. The company operates a silicon intellectual property (SIP) business model that, at least since 1996, shows remarkable similarities to ARM's (for example, there is a similar lag between the license phase and the royalty phase of revenue), and I believe presents similar difficulties in valuation.

If we were to compare the two companies simply by comparing certain parameters, we might easily draw incorrect, or at least invalid, conclusions. For example, I don't think it is necessarily meaningful to compare TTM profits or profit margins, considering ARM took that exceptional non-recurring charge in Q4. Return on Equity is an absurd parameter to use for SIP companies, since the main capital investment lies in R&D, or Product Development, which is excluded from the conventional RoE calculation. Simple Free Cashflow and Price/Sales should be considered in view of the growth history and prospects of each company, especially where there is particularly rapid growth in revenue and earnings. TMF's old favourite metric, PEG, is among the most useful, it seems, in this case. At some point last week, I noticed that Artisan's PEG after the bump in price was still 1.05, based on 5-year expectations. ARMHY's, for comparison, was 1.21.

Still, I think there is no substitute for a good rummage around in a company's business model, to see what makes it tick, and I like what I see of Artisan: a 12-year trend of increasing revenue, and very healthy growth in operating profit over the last two years.

I do have some reservations. The decline in OpEx over the last year seems to be due in part to a reduction in R&D, but there may be some reasonable explanation for this. The smooth line of license revenue growth has also been disrupted over the last few quarters, but this has been more than compensated by rapid growth in net [sic] royalties. I am still trying to decide whether I approve of the gross/net royalty system used by Artisan, but more of than anon.

So, the bottom line is that if I had stumbled across Artisan rather more than two weeks ago, I think I would have felt very comfortable buying at the pre-bid levels, on the basis that the shareprice chart did not yet reflect the strength of the recent operating performance of the company. Having watched ARM for a number of years, I know how rapidly SIP companies can grow, and how well the model can scale up organically under proper, strategically-oriented management. As I think I already posted here, of the top ten SIP companies in 2001, Artisan is the only one that has increased revenue over each of the two subsequent years.

So, purely as an investment, I think I can see some merit in the bid, and at least the potential for a case that Artisan is still undervalued at the present arbitrage-determined levels. Now let's see what ARM can do to increase the value.

Operationally, I don't see huge scope for tweaking the knobs. I think there may be some case for adjusting the gross/net royalties process in favour of more conventional discounting, and I think ARM may be able to use their experience to drive down cost of sales through standardization on replicable solutions.

I see the argument in favour of "one-stop shopping", but I am not yet convinced that this is likely to prove significant in this case.

The original press release from 23 August included some remark, I believe, about ARM being able to benefit from Artisan's proven sales channel, gaining access to 2000 accounts. This is pure horsefeathers, and leads me to treat ARM's case with considerably greater caution than I might otherwise have exercised. In fact, Artisan have only around 1000 licensees, and of those only around 100 have actually paid a license fee. Artisan's direct sales force totaled only 34 people at the end of last year, compared to ARM's 204. I am not saying that Artisan's is necessarily disproportionately low, but let's be realistic about who will be helping out whom in the proposed company. ARM have said in the past that they see their own addressable market as being stable around 450 companies, and I seriously doubt that ARM now suddenly regard the 1550 extra to which Artisan may have some access as the source of their own future growth. The challenge, rather, is for ARM to find some way for Artisan to get cash from the 1900 who have paid no license fee so far, and to ensure that as many ARM deals as possible also include Artisan IP.

So where's the beef? I don't see any of the above as compelling drivers for the deal.

I am sure that for Artisan, having the imprimatur of ARM must count in their favour. Artisan's market seems to be still a little more competitive than ARM's, and I am sure they can build on the advantages of interoperability and reliability that the ARM endorsement creates. I have seen some suggestions that this could work even if the companies remain independent, but I strongly disagree: ARM's endorsement would provide Artisan with, effectively, a franchise, and ARM would be entitled to expect a financial return, which could get gruesomely complicated. Ownership achieves that return simply. I suppose that counts as synergy, but I still have the feeling we are missing the real big picture.

What does ARM see that we don't, yet?

I think it may be a "Vision Thang" thang.

ARM want to be the architecture of choice for the digital world, but it is their strategy for fulfilling that objective that interests me. I have long argued that ARM's success is due not only to the performance/price/power characteristics of their cores, but to the value proposition they offer chip designers: a faster, cheaper, safer route to market.

I suggest that the Artisan acquisition is not so much an extension of that strategy as simply another aspect of it.

Let's think through the chip lifecycle. Starting from a twinkle in a salesman's eye, a development project is conceived that will create a new product. The potential market for the product is assessed, and the sales potential estimated. Against these benefits, the management team will weigh the various costs, timescales, and risks associated with the proposed development (bear in mind that both time and risk inherently carry cost too).

The costs are divided into those that are proportional to production, and non-recurring engineering (NRE) costs. These NREs are effectively an investment, on which a return must be created through the production run, which must also cover the production costs. Obviously, the longer or greater a production run, the lower the cost base of the finished product. For most purposes, chip design costs are included in NREs, which includes R&D and product development.

To date, ARM have been remarkably successful in creating revenue by driving down the costs of development in the design phase, and Artisan have been successful in creating revenue by driving down the costs of the transition to production. I see both trends continuing for the foreseeable future, but I don't think the market has yet realized the scope for Artisan to grow in precisely the same way ARM have.

As Artisan tell us in their SEC filings:

We serve as an interface layer between integrated circuit designers and manufacturers. This manufacturing process interface layer is important since every integrated circuit design must be mapped into a given manufacturing process to achieve specified performance and desired yield.

That interface layer represents an opportunity to drive down costs for the designers. That is Artisan's value proposition as I see it at present, and the source of revenue growth potential that may not yet be apparent to the stock markets.

Here is the CEO of MIPS (!), in an article that is undated but seems to have been written around 2001-2.

Question #2: Can it meet cost needs?

For years, the big cost-related question has always been, "How large is the die?" Soon, it will be, "How much will it cost to get the die into production?" The trend is crystal clear: The cost of developing advanced SOCs is skyrocketing.

Just a few years ago the cost of a mask set seemed very high at a few hundred thousand dollars. Now, it's roughly a million dollars for a 0.10-micron set. If you've got more than two or three spins of silicon, your costs are out of control. And look at what's happening in the area of non-recurring engineering (NRE) charges. Around 1995, mask and development costs for a typical SOC, including three mask spins and a total run of 250,000 units, ran to about 13% of the total cost of the product. Not so by 2003. That 13% is rising to 62%.


(My emphasis. Don't worry about the rest of the article: it is mainly an effective argument against licensing a MIPS core).

We are familiar with how ARM have capitalized on the huge range of semiconductor design activity over the last ten years, and to see the scope of Artisan's market we should examine the scale of anticipated activity in the foundry market. The transition to production, especially for fabless companies, offers huge opportunities for economy based on standardized, replicable solutions.

Of course, not every management team has the ability to seize the opportunity, but ARM have become overwhelmingly dominant through constant focus on customers' needs, and on the Artisan side Mark Templeton (CEO) and Scott Becker (CTO) have been with the company since it was founded in 1991. They have grown the company they co-founded, through changes in the model, through IPO, through strategic acquisitions, and, most importantly, through consistent and impressive revenue growth. The proposed combined ARMisan executive team represents the greatest pool of SIP expertise in the world, by a huge margin.

The proposed combined ARMisan value proposition to semiconductor designers is extremely compelling.

At present, I am rather inclined in favour of the takeover.

Now, lemme see about that presentation...

Cheers,
Scollag.


Become a Complete Fool
Join the best community on the web! Becoming a full member of the Fool Community is easy, takes just a minute, and is very inexpensive.