Memory chipmakers have been among the biggest winners of the artificial intelligence boom, and investors have noticed. Companies like Micron Technology and SK Hynix have watched their profits soar as large language models demand ever more memory chips to package alongside AI accelerators and graphics processing units. With manufacturing capacity taking years to expand, prices have surged alongside demand, delivering record profits and attracting a wave of investors convinced that the AI build-out is still in its early innings.
A growing chorus of market participants argues that the brutal earnings cycles that have haunted the memory chip industry for decades are finally a thing of the past. The reasoning goes that selling chips to AI hyperscalers rather than fickle consumer device manufacturers brings a structural stability the industry has never enjoyed. Investors have been willing to pay up for that narrative, bidding shares higher on the belief that this shift represents a fundamental break from history.
That optimism is exactly what would have drawn a cautionary word from Sir John Templeton, one of the most celebrated investors of the twentieth century. Templeton famously warned that the four most dangerous words in investing are “this time it’s different,” a phrase he used to flag moments when valuations detach from historical norms and rationalizations replace discipline. The more often those words surface, he believed, the more skeptical an investor should become. The memory chip sector is now echoing with precisely that sentiment, and there are reasons to wonder whether the confidence is warranted.
History is littered with transformational technologies that triggered massive capital spending sprees before supply eventually overwhelmed demand. Railroads, telecom networks, and internet infrastructure all changed the world, but they also produced spectacular booms and busts as investment outpaced real need. Even Micron and SK Hynix seem to sense the risk, having secured long-term customer agreements to protect pricing on the downside while simultaneously committing hundreds of billions of dollars to new manufacturing capacity over the coming decade. Those contracts may soften the landing, but they will not stop new capacity from eventually pushing prices lower or prevent depreciation costs from eating into margins.
Templeton himself acknowledged that roughly twenty percent of the time, things genuinely are different, so it is impossible to say with certainty that AI will not prove the exception. Still, after the tremendous run these stocks have enjoyed, investors may want to consider whether the current prices already reflect perfection. Trimming positions or at least exercising considerable caution before adding new exposure might be the wiser course, especially when some of the smartest minds in investing history would be sounding the alarm.