AMD

Here is the problem, which is nearly at a primary school level. A simple logical puzzle. A shopping street has two grocery stores. One of the stores is much more popular than the other. But both shops are full of customers every day. So both shops are raking in money. Sales output of a more popular store roughly doubled over the past year, from $14.5 billion to $30.8 billion - oh, yes, it's a very big shop - which led to tripling of its market value. Meanwhile, sales in the second store have already grown by 69%, albeit by its lower standards, namely from $2.3 billion to $3.9 billion. Please draw a conclusion, by what percentage the market value of the second store could increase, assuming that professional appraisers are rather objective. It seems ridiculous, but the correct answer is that the second store's market value lost 35% within the same year, and it even dropped by 50% from its peak price of the last spring. Holy Cow! That was a story of some failed expectations of mine. Since the big store is, of course, Nvidia, and the small one (and also, in fact, quite a prosperous marketplace) is Advanced Micro Devices (AMD). And their goods are not essential food, but chips for artificial intelligence (AI) related data centers, which are also in high demand.

Moreover, AMD shares reportedly tumbled 10% additionally on February 5, only because the firm's AI chip revenue failed to be exactly in line with elevated projections of Wall Street analyst pool, which somehow bet on a 80% pace of data centre growth to as much as $4.15 billion YoY. Okay, one might say that Nvidia's "store" sells 8 times more chips that everyone needs. And even remember that Nvidia chips are of better quality, that Nvidia occupies about 80% of global chip market share. Again, Nvidia's last quarter will be finally counted only by February 26, when Nvidia's financial report is scheduled, a month later than in AMD's case. Like most large investment houses, here I have provided growth metrics regarding the major data center segment, which is a proxy for the AI playground, where AMD struggles to compete with Nvidia. Well, AMD CEO Lisa Su admitted that her company's data center sales in the current quarter may go down about 7% from the just-ended quarter, but this announcement was exactly in line with an overall expected decline. Is it really such a big deal that AMD shareholders have to experience pain from seeing their chosen stock falling to a 14-month low, with further need for a 100% rally just to match last year's record prices?

The same Lisa Su declined to give the particular forecast for the company's AI chips, but she said that AMD expects "tens of billions" of dollars in sales "in the next couple of years". And I see no reason to doubt her words. AMD CEO added that the firm is now working to compete against Broadcom (AVGO) in collaborating with its customers like Meta and Microsoft to create custom AI chips for their purposes, as Broadcom helps its partners to design their own chips, contrary to mostly "off-the-shelf" processors by AMD and Nvidia. They know their weaknesses as opportunities for strengthening to work in that direction, so what's wrong with the market's adequacy of perception? Perfect Nvidia is the enemy of good AMD, according to the crowd's opinion. Besides AI chips, AMD is also one of the largest providers of personal computer chips. Until recently, this point was generally the source of their main income. Consumers continue to buy new PCs, which also can handle generative AI tasks, by the way.

Actually, AMD has been the only loss-making company in my large portfolio for a long time, so it even makes me smile now. At least, because it is only a matter of time before AMD's pogo stick ultimately uncoils to come loose. Record annual revenue and earnings have to entail recovering to record market value eventually. I am not sure this will happen in the first half of 2025, even though AMD forecasts its revenue rise between $6.8 billion and $7.4 billion for the current quarter, with the market consensus midpoint being slightly lower at $7.04 billion. If you don't believe me then analysts at Stifel are of the opinion that AMD is well positioned for AI compute and "It is likely" that some of its customers "are waiting for 325/350 systems, which should drive a much stronger second half". Again, the median estimate by the Wall Street's analyst pool was now declined to about $150 per share vs $166.5 before the last downside move, yet even $150 sounds much better compared to $112 on closing price this Wednesday or an intraday low at $106.56 during the last trading session. Anyway, there is a strong technical and psychological support zone near the round figure of $100, from where AMD stock had begun its cool ascension in late 2023.