Software Stocks' Decade-Long Outperformance Ends as IGV Returns Now Match S&P 500
Claire Weston
The software ETF (IGV) has seen its ten-year annualized return fall to 15.7%, matching the S&P 500 exactly; AI displacement fears are repricing the entire sector from the ground up.
How did a decade of outperformance vanish?
In the ten years before ChatGPT launched, IGV returned 335% cumulatively (~16% annualized) versus the S&P 500's 238% (~13%) — software led by a wide margin.
Since the start of 2026, IGV has dropped nearly 16% while the S&P 500 has risen roughly 10% — a 26-percentage-point gap in just months.
This means → three and a half years of S&P catch-up have completely erased the excess return software once held.
How exactly is AI hitting software companies?
The core fear: legacy software products — subscription SaaS, CRM suites — could be directly replaced by AI-native tools, thinning the old moats.
Salesforce is down 43% year-to-date, facing its longest losing streak on record; Microsoft, despite holding an OpenAI stake and running massive AI cloud operations, is still down 23%.
In plain terms = even the software giants with the strongest AI narratives cannot hold up — the market is re-examining the old "software equals high growth" thesis.
How far has the valuation multiple compressed?
IGV now trades at 23.6× forward P/E — price divided by expected earnings over the next twelve months — down from 26.5× at the start of the year and well below the 28.6× when ChatGPT launched.
At the same time, IGV's forward EPS has risen from $2.78 last year to $3.52 for 2026 — earnings themselves are not deteriorating.
This means → companies are not earning less; the market simply refuses to pay the same price for the same earnings — confidence is shrinking.
Where is institutional money moving?
Invesco stated in its mid-year outlook: "We favor semiconductor and hardware names and remain cautious on software."
This reflects a broad institutional read: the spoils of the AI arms race are flowing toward chipmakers and energy suppliers on the hardware side, not traditional software vendors.
In plain terms = the arms dealers look more certain to profit than the troops using the weapons — capital is picking the higher-certainty side.
Can the software sector stabilize here?
Volatility remains, but the "high growth plus wide margins" narrative that once justified premium valuations is being steadily eroded by AI displacement expectations.
Whether the multiple can hold at current levels depends on how the market ultimately prices AI substitution risk — and that question has no answer yet.
This means → until the market finishes pricing "how much software AI actually replaces," the sector is likely to stay under pressure.
Content is for reference only, not financial advice.