ARM Holdings (ARMH) is one of the world's most successful semiconductor IP licensing firms. After nearly 10 years, the firm's revenues, profits, and most importantly stock price have broken out to new all-time highs, on the back of the mass adoption of ARM's CPU IP across the vast majority of smartphones and tablets. While I also believe ARM's micro-controller share gains also continue to drive upside to both licensing and royalty revenue, these chips have a less tangible impact to the top and bottom lines as the ASPs for such devices are much lower. In this article, I'd like to make a simple observation about Imagination Technologies' recent fall from grace and why this illustrates the fragility of ARM's rather lofty share price.
Who Is Imagination?
Imagination Technologies (LON:IMG; OTCPK:IGNMF) is another semiconductor IP vendor whose primary purpose is to develop low power graphics IP and license it to SoC vendors, much in the same way that ARM licenses both CPU and GPU IP to semiconductor design companies. On the back of its recent similarly explosive success in the mobile market (Imagination's graphics IP is in every iPhone, iPad, and many Android phones), the share price has gone "parabolic" in much the same way that ARM's has over the last couple of years:
The company's business model is identical to ARM's, albeit its portfolio of IP is (currently) less rich and certainly less pervasive than ARM's, but the core business model is essentially the same -- license IP to chip designers, collect royalties, rinse and repeat. Imagination's graphics IP is generally considered the world's lowest power/highest performing for the ultra low power space.
Imagination's Growth Has Been Superb...
ARM has had an absolutely wonderful boost in revenues/profits since the launch of the iPhone in 2007 (and the subsequent birth of an industry in which ARM could produce more substantial, higher royalty bearing designs):
The "Cortex A" designs found in many smartphones and tablets probably get much more of a licensing "oomph" upfront, and then royalties thereafter are also likely much higher than the micro-controllers/"Cortex M" series of processors. Imagination, too, has had a very similar boost in top and bottom lines with 2009 revenues coming in at ₤64.1M, and 2012 revenues coming in at a whopping ₤127.5M, a doubling very similar to ARM's action (of course at different revenue levels). Imagination's stock recently hit 712 pence (~$10.82) per share before over half of its peak value. Just what triggered this drop?
A Temporary Catalyst And Expectations Of Competition
The first catalyst that drove the share price downwards was the following from its recent earnings report:
As highlighted in the IMS on 12 March 2013, despite an active pipeline of licensing engagements, the closure timing of a number of deals has been subject to higher levels of uncertainty due to structural and organizational changes among a number of semiconductor partners in various regions. As a result several agreements have been subject to short-term delays, impacting FY13 licensing revenues. As a consequence of this we now expect licensing revenue (excluding MIPS) to be around £27m for the year. We are confident this timing issue is a short-term phenomenon related to the "lumpy" nature of licensing revenues.
Despite having record revenues, strong licensing, and increasing market penetration, Imagination's share price now trades at 336 pence (~$5.11), about 50% off from its 712 pence (~$10.82) high achieved in March.
Further, I expect that investors are wary that ARM might actually eat into Imagination's graphics cash cow with its "Mali" IP. I am not so confident that this is a short-term problem, but clearly more competition in the space is enough to shave points off of your earnings multiple.
I now caution investors that the same things that hit Imagination could also hit ARM.
Imagination's MIPS and Intel's Atom
This went largely unnoticed by the financial community in the States, but Imagination Technologies now has full rights to MIPS Technologies' CPU IP. MIPS was an ARM competitor that ran an identical business model to ARM (but had been around much longer, and built its business in segments besides mobile) that eventually got carved up and sold. Imagination got the CPU bits.
Outside of the threat that I believe is coming from Intel's (INTC) "Atom" products in phones and tablets (and a strong defense in the datacenter), MIPS could also be a huge problem. In particular, while the "cheap Chinese SoC vendors" today use ARM IP in their SoCs, the relationship that many of the Chinese CPU design teams have with MIPS, coupled with the fact that Imagination may be very aggressive on licensing/royalty rates, means one of two things:
- ARM's ability to materially increase royalty rates may be limited, and there may actually be downward pressure over the longer term
- Given how levered to much higher royalty bearing IP ARM is in the smartphone/tablet space, any outright socket losses at major customers could have a major impact.
A barrier to entry that may be cited is that the "ecosystem" is all ARM, but Google's (GOOG) Android will work perfectly on both X86 and MIPS, thanks to the high level way that code is written for this platform. While on an OS such as Microsoft's (MSFT) Windows, the legacy software is all natively compiled code, the vast majority of Android applications only need an implementation of the Dalvik VM for the particular instruction set architecture.
Another unforeseen problem for ARM that many seem to ignore is that cost-per-transistor for the ARM ecosystem is actually going up substantially, and it gets worse at "14nm/16nm" since the back-end does not shrink with the front end from the 20nm node; the upshot is that while "14nm/16nm" will offer better leakage/performance characteristics over the 20nm node, it will likely offer very little in the way of cost benefits. Intel, on the other hand, begins mass producing a "true" 14nm process at the end of 2013 with volume shipments in both high end and SoC products occurring during 2014 -- as TSMC/GloFo ramp their 20nm processes:
The Catalysts Are Not Obvious Yet
The share prices don't respond to these potentially large shifts until they are very dramatically obvious to everybody (not just technology geeks). These arguments are not "obvious," nor do they impact ARM's top and bottom lines today, so there is no reason for ARM to fall on these arguments.
However, the day you see a major design win in the high end go to Intel, or if you see some of the cheaper Chinese SoC vendors move towards Imagination's "MIPS" IP, then sentiment will begin to change, as it will then be clearer that it's only a matter of time before the full impact of these catalysts is visible to all.
For those of you unwilling to fight momentum, but believe these arguments, I still recommend that you wait until the evidence is here. Just as with Imagination's fall, or as with Apple's (AAPL) recent decline, it is usually much safer to get in on the short trade when it is already clear to everybody. You won't "catch the top" if this all comes true, but the share price will go into free-fall, and you can easily short on any/all pops even following an initial "gap down," since the trapped longs/profit takers will use any opportunity to get out of their positions. Buying puts during such a downtrend is usually like pushing a button that says "give me free money."
Of course, there is real risk that Imagination does nothing meaningful with MIPS to go up against ARM's monopoly in this space, and there's a real risk that Intel's "Baytrail," "Merrifield," and "6331" fail to gain traction in the tablet/phone spaces, but if you see that either of these two catalysts (or both) are unfolding, then ARM becomes a dream short as multiples quickly compress to account for uncertainty.
Shorting ARM right now is not the safest play in the world, but I would be watching very, very carefully for signs that the "blue skies" scenario that is built into the share price starts to look less plausible. I wouldn't want to be long when the initial drop hits, and certainly not after it, which is why at these lofty levels, I have to caution you to look at what happened to Imagination when even the slightest hint that things weren't perfect came along. It got cut in half (although it looks to be bouncing somewhat), and created legions of bag-holders.
Disclosure: I am short ARMH, and I am long INTC, MSFT. 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.