Templeton Warning Casts Shadow Over Memory Chip Valuations
Record profits for Micron and SK Hynix are driving a narrative that memory chip cycles are obsolete, but historical parallels suggest investors are ignoring familiar capital expenditure risks.
Micron Technology and SK Hynix are generating record profits in 2026 as the artificial intelligence boom turns memory chips into a severe infrastructure bottleneck. As large language models scale, memory has become a critical component required to package with AI accelerators and graphics processing units. Because expanding semiconductor manufacturing capacity requires significant lead times, the sudden spike in demand has driven a commensurate surge in chip pricing.
This pricing power has transformed Micron and SK Hynix into two of the biggest winners of the 2026 market rally. Market participants have aggressively piled into these names, operating on the belief that the current AI build-out is far from its peak. This optimism has led investors to assign premium multiples to the memory chipmakers' current earnings.
The bull case centers on a fundamental shift in the end market. Micron and SK Hynix are no longer relying on consumer device manufacturers, which traditionally exposed the sector to the whims of consumer sentiment and broader macroeconomic fluctuations. Instead, the vast majority of their chips are now being absorbed by AI hyperscalers.
Proponents argue this transition effectively removes the variability that has caused sharp earnings cycles in the memory industry for decades. The structural demands of AI, the logic goes, justify paying a higher price for earnings today than historical norms would typically allow. Yet this prevailing market thesis warrants skepticism when viewed through the lens of historical technology investment cycles.
The late investing legend John Templeton famously warned that "the four most dangerous words in investing are 'this time it's different.'" Templeton utilized this phrase to highlight the risks inherent in market bubbles, where underlying reasoning is used to justify valuations deviating far from established historical norms. That exact phrase is now forming the backbone of the bull thesis for memory stocks.
The more frequently investors assert that cyclical risks have been permanently engineered out of the market, the more caution is typically warranted. Historical precedents offer a sobering parallel to the current AI infrastructure build-out. Massive capital spending projects in sectors like railroads, telecommunications, and internet infrastructure all eventually collapsed under their own weight.
AI is almost certainly a transformational technology on par with those previous innovations. However, assuming the current trajectory of capital spending will grow indefinitely ignores the inevitable saturation of demand. For investors holding Micron and SK Hynix, the current record profits are undisputed, but betting against the historical cycle remains a hazardous proposition.
