24/7 Wall St.

Sometimes you see analyst calls that stay negative for too long, and sometimes those same analysts decide to get bullish after a massive run. It happens. After all, analysts often have years and years of evaluating a company, and sometimes it is just too hard to see that good news can last after years of disappointment. So what should investors take from a major upgrade of Advanced Micro Devices Inc. (NASDAQ: AMD) when the analyst report admits that it is very late to the party?

AMD was raised to Buy from Underperform at Merrill Lynch. That is a rare two-notch upgrade, and it comes after AMD shares have risen exponentially off of their lows. Merrill Lynch’s team covering AMD is led by Vivek Arya and Adam Gonzalez, and their price objective was raised to $12 from $5 in this call.

This call also has a lot of overlap with Intel Corp. (NASDAQ: INTC) and NVIDIA Corp. (NASDAQ: NVDA) due to processors, graphics processing units, deep learning, artificial intelligence and even personal computers (PCs). Merrill Lynch’s team is the same for Intel, which they rate as Buy with a $42 price objective. They also rate NVIDIA with a Buy rating and just raised their price objective to $112 from $95 on this same day.

What investors need to consider here is that AMD was at $9.56 prior to the call, and its prior 52-week range was $1.75 to $9.79. After about an hour of trading on Thursday, AMD shares were up 7.7% at $10.30 and more than 33 million shares had already traded hands. The average daily volume is close to 47 million shares.

So, the first thing is Merrill Lynch’s admission that it is late to the party. The team said that it is “better late than sidelined” and that its “growth options too large to ignore.” They even said that AMD is uniquely addressing a $50 billion total addressable market and is valued at just two times sales over enterprise value.

Merrill Lynch’s call even assigned a $15 bull case and a $7 bear case for AMD. This report said:

In our view, AMD is the only vendor that can challenge both Intel in PCs and servers and NVIDIA in gaming, pro-graphics, and in deep learning markets. Collectively this is a $50 billion duopolistic addressable opportunity, with AMD holding less than 5% value share currently, with any success highly accretive to AMD’s current operating model. … Second, AMD has shown strong capability to build semi-custom chips, and monetize its patent portfolio, which creates more growth leverage.

One issue brought up is that just 25% of the street analysts really have Buy ratings. That leaves substantial potential for higher earnings revisions ahead. They see the next catalyst being the next-generation Zen gaming PC in the first quarter and the Zen server in the second quarter. They also see high-end graphics in mid-2017 and a third semi-custom win in the second half of 2017.

As far as why the firm was late, the team said that they were too concerned about AMD’s historically weak execution and high debt levels. They also were concerned about competitive risks versus chip industry giants like Intel and NVIDIA. Now Merrill Lynch’s proprietary PC gaming analysis and analysis of artificial intelligence and deep learning suggest that AMD’s growth markets are still in the early stages. They also feel AMD can regain market share. Other positives were shown as follows:

• CEO Lisa Su, who was appointed in late 2014, has delivered a strong turnaround and is focusing on relevant markets and improving past deficiencies in software.
• AMD’s recent refinancing actions have pushed out debt maturities and increased its cash balance to $1.2 billion, which is well above the required $600 million minimum levels.

As far as the $7 bear case and $15 bull case, AMD’s value of just two times enterprise value over sales is far cheaper than peers. Merrill Lynch’s new earnings per share estimates of $0.10 in 2017 and $0.27 for 2018 are roughly double the consensus estimates.

Merrill Lynch’s new investment rationale says:

We rate AMD Buy since we view it as the only company that can challenge two large incumbents Intel and NVIDIA in a $50bn addressable market opportunity in PC, server, high-end gaming, deep learning and related markets where AMD has less than 5% value share currently. Second, the company’s model allows for semi-custom product and licensing options with potential customers, which would create further leverage in the model.

The team said back on October 21 that AMD had a strong execution and pipeline but tough competition and tough business model. That summary, which was again with an Underperform rating at the time, suggested that Intel was unlikely to give up any market share in data centers and that the Alibaba server announcement was not a meaningful topline driver. They also felt that AMD was still far behind NVIDIA in deep learning. Their rationale at the time:

We rate AMD as Underperform due to its exposure to the low-end consumer PC segment which renders the company susceptible to continued market share losses to Intel and amplifies weakness of the PC market. The company is doing well with custom processor wins in next generation game consoles at Sony (Playstation 4) and Microsoft (Xbox ONE) however, but we expect that will be offset by declines in low-end consumer PC’s and limited traction in commercial PC’s.

As a reminder: AMD is a heavily shorted stock due to years of negativity. The most recent short interest report was from mid-November, and that showed more than 90 million shares short.

As far as what else might stand out about this AMD $12 target, that is now a dollar higher than the prior highest analyst price target. Sometimes two-notch upgrades happen, and sometimes analysts are way late to the party. Stay tuned.