Micron is the 13th company to hit the $1 trillion mark: what does it mean?

Silicon chip maker Micron is the 13th company to currently carry a stock market valuation of $1 trillion or more, as earnings and price target upgrades from investment banks fuel momentum in a share price already carried higher by strong demand for its Dynamic Random Access Memory (DRAM) products.

A strong trading outlook

The strong trading outlook today is hard to dispute, as artificial intelligence’s need for computing power and inference capability drives voracious demand for memory. But the share price chart looks awfully similar to that of 1999-2000, just as the technology, media and telecoms bubble enjoyed a final blow-off surge before it burst.

Micron’s sales and profits have gone into orbit in the past two years.

Meanwhile its share price has hardly looked back since the launch of ChatGPT by OpenAI on 30 November 2022. Micron’s shares are up by 1,454% since then, compared to an 85% advance by the S&P 500.

The good news is that analysts think Micron’s profit momentum will be maintained. The Idaho-headquartered company is expected to rack up record sales and after-tax profits in each of its financial years to August 2026 and 2027.

If analysts’ forecasts prove accurate for the current financial year, Micron will make $66 billion in after-tax income, compared to $49.4 billion in total across the preceding 30 years.

That helps to explain investors’ enthusiasm for the shares, and, as a rule, semiconductor companies are momentum stocks extraordinaire. They feed off positive earnings momentum, in the form of estimate upgrades, and recoil from negative momentum and downgrades. They are generally very operationally geared, as small changes in sales turn into a much bigger change in profits, up or down.

Chip market is volatile and cyclical

The silicon chip-making industry has tended to be volatile and cyclical over time, despite its long-term growth trajectory which is the result of increases in GDP, new gadget and technological advances and releases, and also higher silicon content per device or product.

DRAM has a reputation for being particularly boom and bust, as Micron’s own history shows. Profits soared in 2000, 2005-06, 2010, 2013-15 and 2017-19, only for the company to record losses in 2001-03, 2007-08, 2012 and 2016, while Covid-19 interfered with 2020’s operational and financial results.

But now the argument is that the age of boom and bust is behind us, thanks to demand from AI, the lengthy lead times involved in building new multi-billion dollar fabrication facilities (‘fabs’) and how Moore’s law no longer works as well as it did, or indeed at all, with the result that chipmakers can no longer increase capacity by boosting output of chips per silicon wafer from their existing facilities.

Consolidation across the industry also means there are just three major DRAM producers right now. These are Micron of America and SK Hynix and Samsung Electronics of South Korea, with the Japanese, Taiwanese and Europeans knocked out of the race long ago.

Add to that talk of the DRAM makers looking to sign supply contracts and set prices for the next two years and you can see why analysts think it could be different this time, and that the boom will not inevitably lead to a sudden, unexpected bust.

But we have been here before. That all sounds great, but similarly bullish noises did not work out as expected when the TMT bubble burst, for three reasons.

What went wrong in the dotcom boom?

First, components shortages began to gum up the works for end-customers, who could not ship finished equipment and systems as fast as their customers wanted, they hoped, and markets expected.

Second, in response to this, customers had begun to double-order in their scramble to find supply of vital components. That bloated order books to deceptive levels and once the market softened, fat backlogs disappeared far more quickly than expected, as orders were cancelled, not confirmed.

Finally, share prices began to yawn. The logic was that if a company was sold out for 12 to 24 months, then there could be no more incrementally good news, or earnings upgrades, for 12 to 24 months. That was an eternity for momentum-driven markets, which craved upgrades, not least to justify prevailing lofty valuations, and prompted some investors to start to look elsewhere. After all, if the news could not get better, then the danger was the next development was that it could get worse, and if the news got worse then valuations would be exposed on the downside.

The first cracks in the TMT boom appeared, and share price momentum began to slow, when companies further down the food chain began to announce they were sold out for a year, or even two years ahead. Passive component, silicon chip and semiconductor production equipment (SPE) companies, and their laser and lens and component providers, all flagged big order intake and fat backlogs, and in the process gave reassurance and visibility on the durability of the expected surge in future sales and profits.

No sign of share price weakness yet

There is clearly no sign of share prices yawning yet, as Micron joins rival Samsung Electronics in the $1 trillion-market-cap club, with SK Hynix just a 3.5% gain in its share price away from joining their ranks.

But the Koreans have had their ups and downs, too. Anyone who invested in SK Hynix (or Hyundai Electronics as it was) at the top in autumn 1999 would have had to wait until this year to get their money back. At least Samsung’s more diversified revenue streams and market share gains across multiple areas, from mobile phones to LEDs and consumer appliances, meant its share price had recouped the TMT bust’s losses within three years.

But anyone who paid the top tick of $70 a share for Micron in March 2000 rode the shares all the way down to $3 in 2009 and finally got into the black in December 2020.

The best cure for high prices is high prices, as they stoke fresh supply and also cool demand, as buyers seek cheaper options or redesign products as best they can. It is not hard to imagine how the finest brains are already looking at ways to adapt to DRAM shortages, while China is looking to make its presence felt in the market, particularly through ChangXin Technologies and JHICC – even if neither currently produce leading-edge chips and both are potentially hobbled by international trade sanctions.

If anything unexpectedly bad does happen, and the DRAM cycle reasserts itself, some investors could be regretting short-term memory loss in more ways than one.