For more than a year now the tech trade was simple — own the chips, sell the software. The narrative was that AI would eat the software stocks alive as “SaaS-pocalypse” spread like The Walking Dead show. I think this article is quite timely as not only have the software stocks been slowly building relative strength in recent weeks, but the semiconductor trade has been acting very heavy in recent sessions. Here’s what the charts say, and two ways we’re looking to play it. Software is turning up against semis After more than a year of software stocks’ underperformance relative to the semiconductor stocks, the ratio of IGV / SMH ( Software ETF / Semiconductor ETF ) bottom near 0.13 last month and has turned back higher, breaking through the black dotted trendline resistance indicating a possible reversal of the software stock misfortune. This has become one of the most crowded trades in recent memory in technology with the masses long semis and short software. As shown in the top-10 holdings of both ETFs (remember you must know what you own when holding ETF’s and mutual funds!), money rushed into Nvidia , TSM , Micron and out of Palantir , ServiceNow , and Salesforce . It’s not just semis – software is turning up against everything! Here’s why I think a structural shift may be happening that is more than just a multiday profit taking in semis ahead of second-quarter earnings. In the bottom panel of the chart below, you’ll see the software ETF is turning higher against the technology ETF XLK after over a year of relative weakness. In the top panel, you’ll find software is actually turning higher against the entire S & P 500 after a triple bottom starting early in 2026! This possible reversal has been in the making for about five months! Software showing relative strength versus the semis — or even technology — is one thing, but when you’re gaining ground on the boarder market, that’s a significantly more compelling story. I spent most of my morning reading through a 65-page research report from the brilliant research team at Jefferies called “What’s Ailing Software.” It practically broke my brain, but here’s the synopsis. It distills the market’s AI fears into 10 “walls of worry” and concludes that most are already priced in — its bottom line being that “software is not dead, but weak software is.” The firm argued systems of record stay durable because AI agents have to plug into them, which strengthens incumbents rather than replacing them, and that near-term margin pressure is largely embedded in today’s beaten-down valuations. Jefferies’ bigger-picture is simple: unless one giant AI model ends up running everything, the corporate world stays fragmented because every company’s data is locked up and walled off from outside vendors — nothing like the clean, uniform data a company like Google gets to train on. That’s why AI has taken off quickly in tidy areas like coding, but will be a lot slower to spread across the rest of a business. And that’s why the edge goes to incumbent companies already sitting on the messy enterprise data. Its playbook is to stay overweight the “token path” — the infrastructure and hyperscaler names positioned to win as AI usage scales — while being more selective on application software. Within that framework, both of the stocks below are named as Jefferies’ top picks: ServiceNow, in application software, and Snowflake , in infrastructure software. I tapped Claude to build this software graphic to show a barbell approach to trading the software recovery. For the deep-value end of the barbell, we’ll start with ServiceNow. It’s one of the more punished names in the group, down roughly 50% from its high. But the daily chart is now carving an inverse head and shoulders, a classic bottom pattern around the $100 zone. The stock is starting to move higher, targeting the 200-day moving average at $132, as well as the blue downtrend line. In the panel below, you’ll see that NOW relative to the S & P 500 has been carving out an inverse H & S as well, suggesting we might see strength against the broader market. I don’t yet own ServiceNow, but I’m getting close to adding it to the growth portfolio for our investors. Three reasons the fundamentals back up the chart: Wall St is flipping bullish. Guggenheim upgraded ServiceNow to buy on July 1 in a note it framed as “Armageddon called off.” Of the 48 analysts that cover it, 10 rate it as a strong buy and 34 as buy. The average price target is $141.12, with shares closing Tuesday at $110.73. Jefferies says the fear is overblow. Their verdict: “software is not dead, weak software is.” ServiceNow is a system of record that AI agents have to plug into, which strengthens it rather than replacing it. Jefferies names it as a top pick, and the margin dip that scared investors is a temporary cost of its Armis acquisition that management expects to reverse in 2027. Valuation. EBITDA has been growing around 30% over the past 3 years and is expected to slow to a respectable 21% in the next 4 years. Earnings per share for 2027 are expected to be $5.03 giving us a cheap 22 forward market multiple. Now to the other end of the barbell, with the software name that continued as a growth stock as most software stocks fell apart. SNOW is trying for a third time to breakout above the last 2025 highs around the $300 level after an initial pullback that traced out a flag pattern in the prior month following the massive gap higher. From a fundamental point of view, Snowflake gets paid when AI runs, not when people log in. Snowflake charges for the data and AI workloads its customers process, which is purse consumption. So the AI boom that threatens the seat-based software is a direct tailwind for SnowFlake. Jefferies puts it in the “token path,” the infrastructure layer built to win from AI, and also names it a top pick. Further, its growth is speeding up, not slowing down. Last quarter, EPS beat expectations by 22%, product revenue grew about 34% beating the company’s own guidance by a wide margin, with a record operating margin and a 126% net-retention rate, plus a new partnership with OpenAI. This was the massive gap in the chart that set up the current 1-month long flag pattern mentioned earlier. 2027 expected EPS is $2.69. Assuming a $268 stock price, it gives us a very expensive 100-times next year’s earnings. But with EPS growing year-over-year since 2022 at 1,823%, 292%, (15%), 51% and next year expected to be up 54%, we’re going to pay a high price for that growth. Bottom line The most one-sided trade in tech — long chips, short software — is likely starting to unwind. Software is now turning up against semis, against the tech sector, and against the whole market at once. Two clean ways to play this at opposite ends of the barbell are ServiceNow, for a beaten-down turnaround with Wall Street coming back to it, and SnowFlake, for proven strength and growth riding the AI-usage wave. — Todd Gordon, founder of Inside Edge Capital We offer active portfolio management and financial planning for retail investors, as well as regular market updates like the idea presented above. Visit us at https://www.insideedgecapital.com/cnbc DISCLOSURES: Todd does not yet owns SNOW or NOW personally and for clients of his wealth management company Inside Edge Capital, LLC. Charts shown are Koyfin All opinions expressed by the CNBC Pro contributors are solely their opinions and do not reflect the opinions of CNBC, or its parent company or affiliates, and may have been previously disseminated by them on television, radio, internet or another medium. THIS CONTENT IS PROVIDED FOR INFORMATIONAL PURPOSES ONLY AND DOES NOT CONSTITUTE FINANCIAL, INVESTMENT, TAX OR LEGAL ADVICE OR A RECOMMENDATION TO BUY ANY SECURITY OR OTHER FINANCIAL ASSET. THE CONTENT IS GENERAL IN NATURE AND DOES NOT REFLECT ANY INDIVIDUAL’S UNIQUE PERSONAL CIRCUMSTANCES. THE ABOVE CONTENT MIGHT NOT BE SUITABLE FOR YOUR PARTICULAR CIRCUMSTANCES. BEFORE MAKING ANY FINANCIAL DECISIONS, YOU SHOULD STRONGLY CONSIDER SEEKING ADVICE FROM YOUR OWN FINANCIAL OR INVESTMENT ADVISOR. Click here for the full disclaimer.
A crowded tech trade is starting to unwind. Charts suggest two stocks to play the move
