Inflation is stubbornly high, but not in home broadband service. There, telecom operators are making fiberoptic inroads in a business long dominated by cable television companies. The result has been discounting: UBS reports that prices for various broadband service tiers are down 7% to 14% for fiber over the past year, and 17% to 27% for cable.

For telecom, this is a good deal. Wireless customers will hop from one carrier to another to score lower monthly bills or free phones, but they rarely change, or even think about, the main internet connection coming into their homes. By bundling the two together at a discount, telecoms can hold down churn in their wireless accounts, much the way cable has long done with broadband and pay television. For cable, however, this telecom insurgency looks like a second death. First, streaming companies came for pay TV. Now, telecoms are taking share in broadband.

The first death wasn't so painful. In late 2015, when Walt Disney disclosed that ESPN subscriptions had fallen by seven million over two years, investors made up their minds. Streaming wasn't just a way for studios to pick up extra money licensing their catalogs to Netflix and others; it was an existential threat to the cable TV bundle. Shares of networks and cable carriers tumbled for weeks. But cable had a backup plan: raise stand-alone broadband prices for customers who "cut the cord" on cable TV.

Cable TV amounts to reselling programming. After cable companies pay studios carriage fees for their cable networks, and retransmission fees for their local broadcasts, gross profit margins are left at 30% to 40%. With broadband, once the lines are installed, costs to add new customers are minimal, and gross margins can reach 70% to 90%. Cable companies faced with losing a $120 a month TV-plus-internet customer could simply raise the cost of internet service from $50 inside the bundle to $80 as a stand-alone.

The Covid-19 pandemic made fast home internet service a must, and cable companies added millions of new customers and raised prices. Share prices for Comcast and Charter Communications peaked in 2021. Telecom companies at the time were spending massive sums to build out their 5G service networks. But by 2023, with the move to 5G largely paid for, telecom spending fell and cash flow swelled. The obvious choice for putting these funds to work was laying more fiber, both for cellular backhaul, which offloads data traffic and makes networks faster, and home internet connections.

The result has been unpleasant for cable shareholders. Comcast-which spun off its cable networks and digital ventures such as Versant early this year and still has the NBC broadcast network, Peacock streaming, and Universal studios and theme parks-has lost nearly 60% in stock market value over the past five years. Charter, a cable pure play, has lost 84%.

Deep-value investors might be tempted. Charter now goes for just three times forward earnings projections, the lowest in the S&P 500, and Comcast sells for six times earnings, the sixth-lowest. But the outlook is daunting. Fiber broadband, which is generally superior to cable on speed and stability, still has only about a 22% market share, compared with 56% for cable. But fiber availability has expanded to 65% of households, and is expected to top 90% by 2030, spurring further fiber share gains.

Two lesser competitors loom, too. Fixed wireless service, whereby telecoms use spare cellular spectrum to offer home internet connections, is no match for fiber on speed but has nonetheless grabbed a quick 14% share of broadband thanks to aggressive telecom bundling. And Starlink satellite service, with a 4% share, is pricey but is cutting into cable in rural markets.

For cable, the best defense might be an upgrade path called Docsis 4.0, which allows for fiberlike speeds on existing copper wires. Since cable was built for television-a downloaded service-its upload speeds are slow. Docsis 4.0 can achieve the upload/download symmetry of fiber with speeds that are good enough for most users. To retain customers in the meantime, cable operators are offering cut-rate mobile phone service, using connections bought from Verizon and others, and bundling streaming services with broadband.

For Comcast in particular, the Versant spinoff this year was supposed to quarantine the most challenged part of the business-cable networks-and leave behind a core of healthy growers. The question now is whether its broadband business is weak enough to warrant spinning off, too. Investors might want to hold off on cable altogether. UBS says the best pick among cable and telecom companies is AT&T. It started later than Verizon in fiber broadband but has far surpassed it in availability, including through the acquisition earlier this year of the fiber business of Lumen Technologies, which provides plenty of room for bundling and growth.

Let's end with a quick word on fitness, banking, and robot revenge. Earlier this month, Barron's highlighted eight stocks that can grow without much help from artificial intelligence. Among them was a catch-the-falling-knife pick, Planet Fitness, already down more than 50% this year on disappointing sign-ups. It fell more this past week on fears that Muse, a new personal AI assistant from Meta Platforms, takes the effort out of canceling memberships that consumers had forgotten about. Another pick, Charles Schwab, is off to a weak start, too, amid concern that Muse will nudge its customers to move their low-interest deposits into something that pays better.

Wall Street is still trying to get a handle on the extent of its Muse risk. When Life Time, a chain of posh gyms, fell along with Planet Fitness, the analyst at Jefferies, who likes both stocks, pointed out that while Planet Fitness members pay as little as $10 to $15 a month, Life Time members pay $245. "Nobody forgets this bill," he wrote.