Financial Model Overview Of AMD And Price Target Initiation

| About: Advanced Micro (AMD)


Initiating 12-month AMD price target at $20.42 and 24-month price target at $29.89.

Shares have yet to price in market share/pricing gains in the MPU space, though GPU trends are well understood.

Furthermore, AMD’s pricing curve seems steeper than what initial price leaks suggest, which provides a strong argument for market share gains.

Expect significant gross/operating margin expansion in the next couple years, given conservative cash management paired with R&D/SG&A efficiency.

I reiterate high conviction buy and forecast dil. EPS of $0.41 and $1.30 for FY’17 and FY’18.

*Subscribers to Cho's Tech Research got early access to this report last week.*

I'm initiating my price target on AMD (NASDAQ:AMD) as I believe there's enough embedded certainty in the current trajectory of financials to make a predictive assessment of forward looking multiples, revenue drivers and margin levers. Though, I believe shares could be subject to some near-term volatility on confirmation of known information, the stock still appeals to long-term investors given the implied upside to unit compares in consumer desktop/laptops versus prior year.

Given the launch cadence of various product releases it's in my view that we can get incrementally aggressive as each quarter progresses, as my financial model would suggest investors are pricing $29.89 into the stock for 2018 and $20.42 into 2017. There's an easily identifiable argument for why investors should anticipate substantial value accretion from recent product developments and modest market share attachment in the OEM channel.

Furthermore, AMD's competitiveness on price and positioning of Ryzen SKUs below-$100 could materially alter the investment thesis without harming expectations on profitability. Furthermore, given the low base ASP (average selling price) figure we are currently operating from (roughly $50), the introduction of a disruptive product that can address price points in the higher volume segment would make me even more aggressive on near-term fundamentals.

This would run contrary to the widely and wrongly held belief that AMD must move aggressively up-market. Though efforts there are appreciative, the situation on the volume channel is the trickiest, and any confirmation that AMD can counter Intel's aggressive rebates would be a net positive.

High-end pricing by AMD seems fine, but I would want the low end of the pricing curve to outperform, not the high end. This is what professionals are fixated on, given the implied pricing/performance leverage AMD is now currently sitting on top of.

Defining the product mix, timing and revenue

On February 22, 2017, we have just received confirmation that AMD is launching three high-end SKUs, which suggests gross margins could swing above estimates, though revenue could be a little tricky to model given the initial introduction of three SKUs.

I stick with the base line scenario because I expect AMD to have made their revenue outlook on the assumption of OEM channel sell-in as granular insights on initial four weeks contribution is measurable from the Q4'16 earnings release announcement date.

Quoted from Credit Suisse

Inventory Dollars/Days Grew Below Seasonal. MPU inventory dollars declined 1.9% q/q below the normal seasonal pattern of up 0.8% q/q. On a days' basis, inventory was down 1.7 days q/q to 81.7 days vs. the normal seasonal pattern of up 1.4 days q/q - days are still above the five-year median of 72.9 days.

Days in inventory in the semiconductor space has dropped in-line with Lisa Su's commentary of burning inventory in the channel. More specifically MPU (microprocessor units) inventory trended below seasonal comps, suggesting OEMs are looking to burn pre-existing units before replenishing with new CPUs, more specifically MPUs. Hence, we can view current channel commentary as a positive forward indicator of replenishment, which is why AMD's ramp up of Ryzen volumes is more heavily Q2'17 weighted as inventory replacement is operative over two months. Given data on sell-in is made readily available when AMD reported in January, the forward-looking outlook is likely to conform with Q1'17 actual results.

Though I expect in-line expectations on actual volumes/revenue. The initial launch seems to carry enough supply of high-end SKUs to drive expectations on gross margins. AMD announced the launch of Ryzen 1800x, 1700x and 1700 which are priced at $499, $399 and $299, respectively. The initial pricing leaks suggested $499, $389 and $319, respectively. The Good news is that AMD's pricing curve when adjusted using conventional calculus (derivative analysis) suggests that the curve is slightly sleeper than what was originally leaked.

Source: Alex Cho

The orange line includes the actual pricing from R7 1800x to R7 1700 and gets projected downwards based on derived figures. This is based on a revised calculation on reported pricing and relative pricing differential. The derivative from the announced price tiers suggests a steeper curve, which basically means pricing gets progressively cheaper the further we move down the stack.

The blue line represents the rumored pricing levels (which didn't sound correct to begin with, as I expound upon in ridiculous detail here). But basically the curve was too flat, and was aggressive to begin with. It didn't even conform with industry standard pricing across the stack. So, I'm incrementally optimistic on AMD's announcement of mid/low-end pricing as I'm expecting unit pricing to get progressively cheaper.

In terms of product release cadence, I'm confident that the low-end volume segment gets a major boost upon the release of Raven Ridge APUs in 2H'17, which has an on-die feature package where both a GPU and CPU are combined resulting in better power efficiency and improvements to form factor. However, the launch of Ryzen series CPUs is more of a desktop product, while many have mentioned that the low end is addressed with APUs. I believe this year is different, and we'll see greater dependence of off-package embedded graphics where the embedded chipsets are sold via the motherboard as opposed to desktop APU configurations. Given the timing of an APU launch, and the suggested market it's relevant in (Ultrabook and laptops), I believe APUs drive the notebook narrative more so than towers. Hence, AMD went ahead with the launch of Ryzen CPUs and integrates with the GPUs on the same die package at a slightly later point.

This makes sense from a hardware engineering standpoint because it's a more complex design. Hence, I'm confident that AMD's fixation on APUs as a volume desktop product gets shifted to 2018 as Raven Ridge is operating at a TDP of 4W to 35W, which is the standard power envelope for tablets to high-end laptops.

Therefore, I model a ramp that's more incremental as I expect consumer market penetration to be more of a cumulative trend rather than a lumped together move. Furthermore, the emphasis on both desktops and servers seems sufficient for 1H'17 y/y revenue acceleration even in a modest market share/pricing scenario.

Breaking down the financial model

Source: Alex Cho, FBR, AMD results, Credit Suisse, Mercury Research

Source: Alex Cho

Granularity is still difficult given the lack of a specific unit count/pricing break out, but we do know with relative certainty the MPU (micro processor unit) is going to contract marginally on a y/y basis with perhaps modest ASP gains. I believe I'm very conservative here, but I eventually model significant contribution in Q2'17 onwards as I anticipate sell-in to move considerably higher on cleared OEM and retail channels. Hence, I'm confident that MPUs (both server and desktop) become a Q2'17 narrative that gets reflected in the outlook figures.

As such, I model a scenario where Q2'17 q/q revenue comps increase 36%, which is driven by 2.5 percentage point market share improvement paired with a $5.00 improvement in ASPs for MPUs. Though, there's potential for upside in pricing, I keep things conservative here, so we're operating off something investors are willing to discount.

That being the case, I forecast ASPs (average selling prices) to accelerate from $49.25 to $79.25 between FY'16 and FY'18. This ASP assumption implies 60% upside to average pricing over a cumulative 24-month period. Though it's worth noting that Intel's ASPs hovers at $150 in consumer MPUs. So, even with AMD's aggressive pricing strategy, it's still possible to boost average pricing while competing with Intel on price (as indicated in the Ryzen price curve). Therefore, we should anticipate significant ASP improvements despite shipments likely concentrated in the $100 to $200 price band.

Furthermore, I'm expecting the GPU-business to materially accelerate given the launch of Vega, which would weaken the investment thesis for Nvidia (NASDAQ:NVDA) holders, but would be net accretive to AMD investors given the quarterly outlook implies 50% y/y growth in the graphics segment. Perhaps a decent chunk of this is driven by share gains in the low/mid-end segments where GPUs are sold for less than $200. That being the case, efforts to drive datacenter accelerators could significantly alter my stance on GPU revenues.

Furthermore, it's difficult to attach a reliable estimate on datacenter MPU/Accelerator sell-in. So, I've modeled something extremely conservative (perhaps 3% y/y growth) for the business in 1H'17. I'm more aggressive on semi-custom revenue on the back of refreshed consoles, which isn't driven by volume, but rather an improvement in pricing due to PS4 Pro and Xbox Scorpio. As such, I'm expecting 20% y/y growth in semi-custom, datacenter, embedded and licensing in 2H'17. I'm certain there's substantial upside to this figure and willing to re-examine upon better than or above-seasonal figurers in Q2'17.

Given the above factors on product introduction and sustained momentum in the MPU/GPU business between FY'17 and FY'18, I'm anticipating revenues of $6.03 billion and $8.6 billion, respectively. I attach a modest market share scenario for the MPU business as I'm expecting market share of 11.10% and 17.30% in FY'17 and FY'18, which is comparable to the ramp-up in the Athlon X2 generation from 10 years ago. I'm expecting notebook shipments to be a material driver in 2H'17, which is where AMD would surprise the current consensus as many are modeling estimates off historical revenue trends on a seasonal basis as opposed to unit shipment attach rates into the OEM channel.

Most of my assumptions on expense/gross margin ramp is comparable to sell-side expectations. I'm certain that AMD will re-invest a decent chunk of sales back into its R&D efforts. I'm expecting gross margin to trend higher, but albeit modestly given the transition to 14nm will drive costs upward given higher wafer defect rates in early production runs and wafer supplies costing an implicit amount per unit. Therefore, I'm expecting gross margins of 34.49% and 37.59% over FY'17 and FY'18, which is significant and drives my profit narrative.

Though OpEx is likely to increase, I'm expecting AMD to layer it on gradually. As such, I'm expecting OpEx to increase by around 14% per year, which is a much slower cost trajectory than revenue. However, I'm confident that AMD's strategy is inclusive of managing costs and would add headcount where it makes sense, given the conservative nature in which Su and her executive team is currently operating. The emphasis on efficiency can be justified given the strength of the product portfolio, relative to what they invest into R&D. Furthermore, the company doesn't carry as much overhead as Intel given the lack of in-house fab, which is a cost reflected in its current wafer supply agreement with Global Foundries. Given my conservative stance on gross margins and more incremental expansion on R&D we can anticipate a scenario where net profit margins increase to 7.23% and 17.68% in FY'17 and FY'18, respectively.

When combining these inputs, I'm expecting AMD to report dilutive EPS of $0.41 and $1.30 for FY'17 and FY'18. This is substantially higher than conservative estimates of $0.06 and $0.27, though I don't believe conservative expectations have anything embedded into them that are this granular. The only estimate that I reviewed that was based on any improvement in assumed share was published by Morgan Stanley, but the analysts there are also conservative on their FY'18 estimate given competitiveness of Intel, and costs being driven higher on a more modest revenue ramp.

Despite concerns over competitiveness in the next year and beyond, I'm certain AMD's current product roadmap positions it for reasonable levels of success. The improved profitability and stable R&D/SG&A ramp would re-initiate AMD for long-term growth and increase size of development teams to reach lofty goals on 7nm design execution. Furthermore, GlobalFoundries has found a stable base to operate from to reach their internal deadline of 2018 or early 2019 volume production of 7nm wafers. Hence, I'm not only optimistic on revenue/expense in the near term but anticipate sustained product deployments to keep pace with both Nvidia and Intel in the immediate five-years.


I'm raising my price target to $20.42 for 2017 and also assigning a 2018 price target of $29.89. I value AMD at 23x FY'18 earnings, which is based on peer based multiples. Obviously, expectations on forward sales/earnings could push the stock substantially higher than this over the next two years, but for sake of providing something tangible, let's settle for $30 over two years.

To arrive at my 2017 price target, I discount my 2018 valuation by 21%, which is based on AMD's current WACC (weighted average cost of capital). My 2017 price target values AMD at 15.7x FY'18 dilutive EPS, or $20.42.

There are numerous factors that may drive my estimates on revenue/earnings higher. Furthermore, my estimated value premium is extremely modest, but until more of these factors materialize, I won't have to revise my estimate too much higher.

As such, I continue to reiterate my high conviction buy recommendation.

About Cho's Technology Research

Cho's Technology Research is Seeking Alpha's premier technology research package that gives investors and traders an additional edge when investing into tech companies. It's a trade publication, research service and an idea generator. It includes short-term trade ideas that yield 5%+, long-term small cap investment ideas that yield 25% and best-in-class data/news on tech sector themes.

Subscribers get an early preview on Cho's price target/recommendation changes. This services comes at a low annual subscription cost of $20/month. The promotional price period will end soon! Subscribers get a free two-week trial upon checkout and can cancel at any point in time. For more information click here.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

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.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500.
Tagged: , , , Semiconductor - Broad Line
Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here