The best investors are able to check their emotions at the door and understand when their thesis is wrong. Unfortunately, most investors suffer from an inability to update their opinions even as conditions change.
AMD (AMD -3.81%) bears are a prime example. Three years ago, many were predicting the company's demise, noting AMD was tethered to the rapidly declining PC market with a disproportionately large debt pile to service. Since then, shares have increased tenfold, comparing favorably to NVIDIA's (NVDA -2.87%) 12-fold increase.
Still, many are still unwilling to give CEO Lisa Su the benefit of the doubt. It's time for AMD bears to admit they're wrong.
The danger of belief perseverance
AMD bears once looked smart. After peaking in the $40-per-share range during the tech bubble of the early 2000s, the company found itself trading at less than $2 per share 15 years later.
Bearishness peaked in 2015 as net revenue in AMD's computing and graphics division fell 42% from the year before, leading to a consolidated revenue decline of 27% and negative operating income.
Quietly, the company had already begun its transition plans. The year prior, CEO Rory Read had retired and appointed Lisa Su as his replacement. Read's background was in the PC industry, while Su's background with Texas Instruments and Freescale Semiconductor (now part of NXP Semiconductors) was in semiconductors.
Su's approach is working: AMD averaged revenue growth of 16% in the next two years, and finally reported positive operating income in fiscal 2017.
Time to revisit the new bear thesis
While many have updated their opinion on AMD, there are still holdouts with new bearish opinions. The pessimists focus on:
Strong competition from NVIDIA in many of the same markets: Even the biggest AMD bull will likely admit NVIDIA is a better-run organization and outpaces AMD in innovation. but this naively assumes data center applications and graphics processing units (GPUs) are winner-take-all markets. With the data center market projected to grow to $70 billion by 2021, the industry is big enough for market leaders like Intel and NVIDIA as well as smaller players like AMD.
As such, NVIDIA's continued success portends healthy markets for AMD as well. Increasingly, AMD is winning market share from Intel in the data center. And with revenue of $5.3 billion versus $63 billion for Intel and $9.7 billion for NVIDIA, even small market-share wins should keep AMD's market cap growing.
GPU demand is driven by crypto-miners and will substantially cool: While this is a risk, as with most subjects that deal with cryptocurrencies, benefits are often overexaggerated. In the first quarter, AMD reported 10% of its revenue was from crypto and blockchain.
Second-quarter crypto-revenue showed a sequential decline. However, AMD still reported a 64% increase in total GPU revenue. The company expects growth from other end markets to continue to offset crypto-related sales, with Su predicting "very little from blockchain" in Q3 revenue while still guiding for 7% year-on-year revenue growth.
The company is richly valued, particularly on a P/E basis: Perhaps the best argument against the company is a fundamental one. AMD is expensive on traditional metrics; it currently trades at 39 times projected earnings, double the greater market, and about 3.5 times sales. If the second-quarter blowout is any indication, growth should account for the pricey valuation.
Surprisingly, AMD's recently reported earnings were boosted by the maligned PC business. Earlier, Gartner noted that PC shipments increased in the second quarter, the first year-on-year increase in six years. Fellow analysis firm IDC noted "growing consumer demand from gamers" as a reason for the increase, which augurs well for AMD's GPU business.
It's time for bears to finally admit AMD is no longer the AMD of yesteryear. The company is well situated in high-growth data center and gaming GPU markets, and should continue to outperform the greater market.