← Weekly Rhyme

MICROSOFT · MSFT

Microsoft’s 2026 Path Has a 93.3 Rhyme With the Summer of Windows 98

Microsoft’s last 90 trading days most closely resemble a stretch ending August 3, 1998. What happened next was anything but a straight line.

Frozen market analog chart

NOW 2026THEN 1998

FROZEN AT 2026-09-11

Today’s Microsoft (MSFT) path and its August 1998 historical analogThe 2026 current path and 1998 historical path converge at one TODAY point equal to 100. The current line stops there, while only the historical 1998 line continues through 90 trading days of historical outcome.60801001201401998 CONTINUED →HISTORICAL OUTCOME · NOT A FORECAST−90DTODAY+90D

Historical outcomes

What followed the matched period.

1M
-6.6%
3M
-2.4%
6M
+54.6%
MAX DRAWDOWN
-19.7%

Historical outcome — not a forecast.

The Rhyme

From May 5 through September 11, 2026, Microsoft’s price path most closely matches the 90-trading-day stretch ending August 3, 1998.

Rhyme Score: 93.3

The resemblance is interesting on price alone. But the historical setting makes it more so.

In both periods, Microsoft was a dominant platform company trying to shape the next layer of computing—while strong execution coexisted with unusually large strategic and external risks.

What Happened Next

After August 3, 1998, Microsoft initially moved lower.

1M: −6.6%

3M: −2.4%

6M: +54.6%

6M Max Drawdown: −19.7%

The six-month endpoint looks spectacular. The path to get there did not.

Russia announced extraordinary financial measures on August 17. The Federal Reserve cut its target rate on September 29 and eased twice more that autumn. Microsoft’s federal antitrust trial began on October 19.

Those events overlapped with the stock’s path; they do not establish a single cause. Company execution, the PC cycle, policy, and risk appetite were all moving at once.

The endpoint alone conceals the experience of getting there.

Why This Rhyme Is Interesting

In 1998, Windows was the gateway to personal computing just as the strategic contest was moving toward the browser and the internet.

Windows 98 launched on June 25. By July 16, North American customers had obtained more than one million retail upgrade licenses, and Microsoft’s fiscal-year revenue had risen 28% to $14.48 billion.

But the same platform power driving Microsoft’s success was being challenged. The Justice Department and state attorneys general had sued Microsoft on May 18 over Windows and Internet Explorer.

The company was executing. Its platform power was being contested.

The technology is different in 2026, but the pattern is recognizable.

Azure provides infrastructure. Microsoft 365 provides enterprise distribution. Copilot connects AI models to existing workflows. Azure and other cloud-services revenue grew 43% in Microsoft’s latest quarter, while Microsoft 365 Copilot exceeded 30 million paid seats.

Microsoft is again trying to turn an installed platform into an advantage in a new computing layer.

Where the Rhyme Breaks

A historical analog becomes more useful when we know where it stops rhyming.

The 1998 Microsoft was primarily a PC-software company built around Windows and Office licensing. Its economics were comparatively asset-light.

The 2026 Microsoft is broader—and AI is far more capital intensive.

Microsoft reported $41 billion of quarterly capital expenditures, with about two-thirds going toward shorter-lived assets such as CPUs and GPUs. The company also said AI infrastructure investment and usage weighed on cloud gross margin.

That creates a challenge with no real equivalent in the Windows era: Microsoft must not only win distribution, but earn attractive returns on enormous physical infrastructure.

Competition is different too, spanning hyperscalers, frontier-model providers, open-source models, and partners that can also be competitors.

The platform pattern rhymes. The economics do not.

Bottom Line

The 1998 analog does not tell us where Microsoft goes next.

It does show a previous moment when exceptional execution, platform power, technological change, regulatory pressure, and market risk existed at the same time.

That makes four things worth watching now: AI demand, distribution power, infrastructure economics, and competition and regulation.

And 1998 offers one useful reminder: a +54.6% six-month outcome still included a nearly 20% drawdown along the way.

Markets don’t repeat. They rhyme.

Historical outcome, not a forecast.

Sources