No force in financial markets is as powerful and predictable as "mean reversion," the tendency of stock and bond prices that are extraordinarily high or low versus history to return to their long-term norms. The big drop we're now seeing in tech shares looks like a classic case in point, as economic gravity takes charge, pulling high-flying prices closer to where they've traditionally hovered relative to earnings.

We're hearing sundry explanations for the fall—"a second lockdown is on the way," "chances for new stimulus are fading," or "the trade war with China is endangering Big Tech"—but the one that makes the most sense is most basic: mean reversion that comes and goes, but always comes again.

In a previous story, this reporter studied the amazing trajectory of the 10 highest market-cap technology companies in the Nasdaq 100, a group I dubbed "StarTech." I treated StarTech as one big company. Since StarTech's record close on Sept. 2, its valuation has cratered from $10.47 trillion to $8.76 trillion, shedding 16.4% or $1.7 trillion, equivalent to one-eighth more than Microsoft's current market cap.

The fall is so substantial that you'd think we might be nearing bargain territory. More likely, mean reversion has just begun, and it has a long way to go. That's because StarTech hasn't come close to retracing the giant spike that took its shares from moderately priced to outrageously expensive. Amazingly, that moonshot began less than a year ago, in the fall of 2019.

On Sept. 30 of last year, StarTech comprised the following members in descending order of valuation: Microsoft, Apple, Amazon, Alphabet, Facebook, Intel, Cisco, Adobe, PayPal, and Tesla (which bills itself, and gets valued, as a tech pioneer). Since the Nasdaq counts Alphabet's A and B shares separately, StarTech actually encompasses not 10 but 11 members. At the time, StarTech's market cap stood at $5.822 trillion, and its combined GAAP, four-quarter trailing net earnings were $224 billion. So its price/earnings multiple was 26, a big increase from its level of 19 in 2015, but in line with its average over the previous half-decade.

From Sept. 30, 2019, to Sept. 2 of this year, Intel and Cisco left the group, replaced by Netflix and Nvidia—a significant shift since the prior pair's profits were a lot higher relative to their valuations than for the newcomers. Over that period of just 11 months, StarTech's valuation soared to that nearly $10.5 trillion peak, a jump of 80%. But earnings didn't keep pace. In fact, StarTech's profits dropped by 6% to $210 billion. The combination of the explosion in prices and slide in earnings doubled the multiple from 26 to 50. In less than a year, Apple's P/E went from 19 to 40, and Microsoft's from 29 to just over 40.

In the past three weeks, the 16% retreat still leaves StarTech's valuation of $8.76 trillion 50% above its level a year ago. Its multiple, what investors are paying for each dollar in earnings, remains highly elevated at 41.7. That's half again its mid-to-high 20s norm since 2015.

What will investors reap if StarTech's P/E falls to its average of 28 of the mid-to-late 2010s? Today, its dividend yield is a puny .32%. We'll assume that StarTech spends half of its total earnings on share repurchases, Big Tech's long-favored vehicle for rewarding shareholders, increasing EPS each year by 1.5%. So dividends and buybacks would deliver total returns of 1.8% combined. We'll also make the incredibly optimistic forecast that overall profits rise by 10% a year, well above the average for the past half-decade.

In that scenario, StarTech's profits would swell from $210 to $338 billion by the fall of 2025. At our future, normalized P/E of 28, its market cap would be $9.5 trillion, just 8% above today's $8.8 trillion. So investors would be getting annual gains of 1.8% from buybacks and dividends and 1.6% from growth in overall profits, for a total of 3.4%. That example illustrates the strong pull of mean reversion. The fall in the multiple to normal levels works against the 10% gains in profits, leaving meager, low-single-digit returns.

Keep in mind that overall profits are highly unlikely to expand at anything approaching 10% in the years to come as Big Tech keeps pouring cash into buybacks over big investments in growing their businesses. So zero or negative returns are more probable. Big Tech could rebound to new highs, of course.

But in the end, mean reversion will win. It always does.