This edition of Market Factors begins with the market trend of the day - the mauling of software stocks thanks to Anthropic’s Claude LLM update. The next section outlines stock picks in the more positive end of the AI investment story - hardware - and the diversion covers science fiction movies that should have won the Oscar.
Dado Ruvic/Reuters
Sector selloff
Time to buy on the dip in software?
Devastation in the software sector is the market story of this week and things might be reaching a point where it makes sense for investors to step in. The sell-off has been broad and severe and at least one analyst thinks a buying opportunity is close by.
JP Morgan technology analyst Mark Murphy began his most recent report by noting the breadth of the sell-off in software. Nothing was safe - small caps and large caps, AI-centric and AI-independent software stocks all got bludgeoned. He noted that Cloudflare Inc., Datadog Inc. and Snowflake Inc. were all down roughly ten per cent on Tuesday despite minimal exposure to data centres.
Anthropic PBC’s Claude LLM is the main catalyst for the sell-off. The company announced new functionality in corporate departments like legal, sales and marketing, data analysis, productivity, finance and accounting. Claude can write bespoke code for corporate users, potentially displacing software companies that sell products for use in these departments.
The software sector was weak ahead of Anthropic’s latest news. Mr. Murphy highlights a lack of headcount growth in the industry and sluggish profit growth as signs AI has been biting into software industry profits for years. AI is exacerbating a slowdown caused by a pulling forward of demand during COVID - pandemic-era software investment was high to account for lockdowns and working from home, and slowed dramatically after the crisis abated.
Mr. Murphy believes selling is overdone, led by panicked generalist investors with little knowledge of industry fundamentals. He writes, “Our sense is that investor concerns have finally caught up with the rate-of-change vector for AI and likely begun to overshoot to the downside.” He further points to surveys of the chief information officers in charge of corporate software spending that have yet to uncover lower software vendor demand.
Jonathan Krinsky, a technically focused analyst at New York-based BTIG, has practical advice for investors looking to benefit from volatility in the software sector. In a Tuesday report he noted that the S&P 1500 Software index was more than 15 per cent below its 200-day moving average. As a further sign of potential overselling, the ratio of software stocks to semiconductors was fully 38 per cent lower than its 200-day moving average.
Mr. Krinsky believes that a sustainable bottom for stock prices in the sector will take some time. However, the North American Tech-Software iShares ETF (IGV-A) doesn’t have far to fall to hit its own 200-day moving average at US$80.60, and this would in his mind be a very attractive entry point.
I like this idea but more for a trade than an investment. Software still runs the business and consumer worlds and if Mr. Murphy is right and CIOs don’t intend to stop paying software vendors, the complete AI takeover of the software sector is still a ways off. I suspect that the expertise necessary to apply Claude and other LLMs to write corporate software is not yet broad-based enough to kill the sector.
That said, growth for software companies in the coming years faces a considerable and, in the end likely unbeatable, challenge from AI. This is why a bit of bottom fishing here is interesting as a trade, but I wouldn’t intend to hold software stocks in large weightings for the long term.
Investing ideas
Stock picks in the winning AI subsector
Another set of peripheral plays on AI, this time from Citi research, was published this week. The picks were limited to the U.S. definition of small and mid cap, above US$1-billion and under US$40-billion. In aggregate, Citi strategists found that earnings revisions, premium profit growth expectations, and PEG ratios were more attractive for AI-related stocks at smaller market caps.
Earnings trends are currently more profitable for AI enablers, companies like memory provider Sandisk Corp. (SNDK-Q) that are selling the equipment necessary to provide AI, than for adopters - non-technology companies adopting AI as part of business activities.
The new basket of stocks is heavily weighted in semiconductors and hardware with utilities and power infrastructure a prominent sub theme. Software companies make up 16 per cent of the equal-weighted basket, which is unfortunate in light of recent market action.
Hardware names in the basket include Sandisk Corp., Microchip Technology (MCHP-Q), Crown Castle (CCI-N), electronic design and test specialists Keysight Technologies Inc. (KEYS-N), Teradyne Inc. (TER-Q, that one keeps coming up), optical network equipment maker Ciena Corp. (CIEN-N), blast from the past Hewlett Packard Co (HPE-N), connectivity experts Astera Labs Inc. (ALAB-Q) and Lumentum Holdings (LITE-Q), another maker of optical equipment.
NRG Energy (NRG-N) and Nvent Electric PLC are included in the power portion of the basket. Symbiotic Inc. (SYM-Q), an automated logistics company, is also listed only with “High Risk” in big letters after it.
Gary Lockwood and Keir Dullea in 2001: A Space Odyssey
Diversions
Sci-Fi movies that should have won the Oscar
The MakeUseOf site celebrated the release of 2026 Oscar candidates by listing nine science fiction movies they believe should have got the nod for best picture instead of the actual winner. In most cases I agree.
The list starts in 1968 when the article’s author Dan Selcke believes 2001: A Space Odyssey should have won over Oliver!. The Kubrick film was intellectually interesting but dragged a bit for me. On the the hand, the Wizard of Oz and Singing in the Rain are the only musicals I like so we’ll agree here.
The next pick is Star Wars over Annie Hall in 1978. It’s a bad week to discuss Woody Allen so let’s just move on. Mr. Selcke also thinks Alien should have beat out Kramer vs Kramer in 1980. In terms of social importance he’s wrong but if we’re talking fun he’s exactly correct.
The Matrix over American Beauty in 2000 is an easy one, even more so because the star of the latter is as problematic as Mr. Allen. I don’t know what to say about the Wall-E over Slumdog Millionaire argument for 2009. Wall-E is great but it’s animated so I disagree on this one. (For animated, Wes Anderson’s Isle of Dogs is my top pick by the way).
Inception over The King’s Speech in 2011 is the next one. They’re so different I don’t want to judge and since it’s my newsletter no one can make me.
Interstellar over Birdman in 2015 is self-evidently correct in my opinion. The music alone in Interstellar increases its rewatchability. As for Fury Road over Spotlight in 2016 it’s the same thing as 1980 - whether you put fun over important or the other way around.
The last one is Dune part two over Anora. I loved Dune but that’s to be expected because I like the first Dune book more than Lord of the Rings. Haven’t seen Anora.
The essentials
Looking for our updates on market movers, analyst actions, stock technicals, insider trades and other daily, weekly and monthly insight? Click here to visit our Inside the Market page.
Globe Investor highlights
Economist David Rosenberg is abandoning his bullish view on the Canadian dollar and expecting more Bank of Canada rate cuts
Hedge fund trading of federal bonds poses financial stability risks, say BoC researchers
David Berman makes the argument that most investors are best served by a big bank ETF than than buying Canada’s big lenders directly. Meanwhile, RBC shares some advice on when to buy bank stocks
Thirty analysts and traders are polled by Reuters on their latest forecasts for gold and silver
The New York Times on how interpreting the cascade of today’s economic data has become trickier than ever
Quick hits
The five most searched investment terms in Canada in January were TSFA, artificial intelligence, gold, REIT and energy stocks. The most searched tickers were for Celestica, Nvidia, Shopify, Enbridge, Canadian Natural Resources, NextGen Energy, Granite REIT, Smartcentre REIT and Agnico Eagle Mines.
A Morgan Stanley strategy report indicated that U.S. self-directed investors were buying more U.S. stocks despite global underperformance - they are less sensitive to a weaker U.S. dollar. A reminder that we shouldn’t expect a complete collapse in the S&P 500.
Also from Morgan Stanley, an important change in valuation models for Microsoft, Oracle, Meta Platforms and Alphabet could see a collective US$650-billion in depreciation charges as they move from asset-light to data centre-heavy business models.
See our full earnings and economic calendar here