Software Stocks Drop for Seven Straight Sessions, Erasing All Gains as AI Disruption Threats Weigh on Valuation Recovery
Taylor Wilson
The iShares software ETF (IGV) has dropped for seven straight sessions, wiping out a prior 16% rebound and pushing year-to-date losses past 12% — AI replacement fears are no longer sentiment noise but a structural breakdown of the sector's valuation framework.
What exactly did seven days of selling erase?
IGV ended its brief rally on June 1, then fell for seven consecutive sessions — the entire 16% three-day gain was given back.
Year-to-date losses now exceed 12%, versus a 7.9% gain for the S&P 500. Software is the index's worst-performing sub-sector.
This means → the earnings-driven bounce was a dead-cat rally; the market took one week to reject three days of optimism.
Why did one Anthropic model release tank the sector?
On Tuesday Anthropic launched its new Mythos model; IGV fell 2.8% on the day.
This reflects a deeper fear: not "can software companies use AI?" but "does AI make software products themselves unnecessary?"
RBC BlueBay market strategy head Mike Bell: even after this sell-off, software stocks still have further downside — this is an existential threat.
In plain terms = when AI can do the job itself, who still needs to buy the tool? This is not a growth-slowdown story — it is a "do you still need to exist?" story.
Valuations look cheap — why can't investors buy the dip?
Bell's point: when the uncertainty range is this wide, traditional valuation metrics lose most of their reference value.
This means → even with P/E multiples halved, the market cannot tell "cheap" from "value trap" — because the denominator (future earnings) may itself go to zero.
Adobe now trades at under 10× forward earnings, far below its ten-year average of roughly 30×, yet no one is calling a bottom.
Adobe reports Thursday — why has it become the bellwether?
Adobe shares are down 32% year-to-date; its creative tools are seen as the most direct target for AI image-generation technology.
The stock has fallen after 9 of its last 11 earnings reports; long-time CEO Shantanu Narayen has announced he will step down, raising transition doubts.
Ruffer fund manager Fiona Ker drew a clear line: it is hard to see where Adobe's competitive moat lies going forward — it looks like a company that could be replaced.
Is the entire sector collapsing?
Not quite: 90% of S&P 500 software companies beat earnings expectations, above the broader market's 82% rate.
Snowflake, Datadog, DigitalOcean, and JFrog all posted strong results; cybersecurity names held up relatively well.
In plain terms = plenty of companies are still making money, but the market does not care how much you earn today — it cares whether you still exist in three years.
OpenAI and Anthropic may IPO — what does that mean for software stocks?
Anthropic could list as early as October this year; OpenAI is also considering a fall debut.
Bell's logic: investors want to avoid AI losers and overweight AI winners — these IPOs offer directly investable AI-winner exposure.
This means → legacy software stocks face a double squeeze: business disrupted by AI + capital diverted by AI-company IPOs.
Ker's playbook: pick companies with large enterprise clients locked in by compliance or legal risk — clients who cannot easily switch. Free-cash-flow yield and below-market multiples provide a cushion.
Content is for reference only, not financial advice.