Any day now, the artificial-intelligence bubble will burst -- or inflate further. The Federal Reserve will hike rates, unless it cuts them. The Strait of Hormuz will close, or open, or close again. Any day now, the S&P 500 will break out of its two-month holding pattern, but will it soar or stumble? Yes.
What should investors do when the outlook is murky and an inflection point nears? In our midyear checks with the members of the Barron's Roundtable, all offered variations on the same advice: Run like the dickens from overowned, overvalued tech stocks and seek safety and hefty returns among the many promising companies -- in energy, materials, healthcare, and even lawn care -- whose shares have been left in the dust by the latest stampede into all things AI.
The Barron's Roundtable last met Jan. 5 in New York, before the Dow industrials topped 50,000, inflation hit 4%, and the guns of war boomed across the Middle East. In the midyear Roundtable, conducted by phone in the past three weeks, 10 of our panelists weighed in on this year's momentous macro changes and shared their favorite investment ideas -- 45 in all -- for the year's second half and beyond. (Mario Gabelli couldn't participate in the midyear Roundtable, but you can track the performance of his January 2026 and July 2025 picks on Barrons.com.)
Stash the tea leaves and crystal balls. Here is our pros' edited guide to the future.
Henry Ellenbogen
Barron's: Henry, stocks were down, then up, in the year's first half. Where to now for the market?
Henry Ellenbogen: The gains were highly concentrated in stocks associated with the AI buildout. Aerospace and defense stocks also showed some strength.
AI growth has been driven primarily by spending by Anthropic, and to a lesser extent OpenAI. People really want to invest in the LLM [large language model] layer, but these companies aren't public yet, so they are investing through proxies. The only publicly traded LLM, Gemini, is embedded in Alphabet's Google. Alphabet's stock has had a strong year. Other proxies include semiconductor and memory stocks, and stocks tied to the AI data-center buildout.
Anthropic became the leader among LLMs with its release of Claude Opus 4.5 at the end of November. Anthropic's revenue has shot up since then, from around $9 billion to more than $50 billion. It is incredible to see a company add that much revenue in only six months.
Anthropic has filed confidentially with the Securities and Exchange Commission for a proposed public offering of its stock. How will the company's initial public offering change investors' preferences and the market?
That is something I have spent a lot of time thinking about. If the LLMs' revenue continues to grow, and I think it will, two things will happen. No. 1, people will invest directly in these companies. No. 2, the market will become much more discerning about the duration of AI companies' cash flows.
For example, Nvidia's stock has significantly underperformed other AI-buildout stocks in the past year as the LLM companies have lessened their ties to the company and worked more with Broadcom, Google's TPUs [tensor processing units], and Amazon.com's AWS Graviton processors. If intelligence is as powerful as we think, it will start to solve some of the bottlenecks we have seen in memory, semis, and such. Thus, the rate of change in demand for these products won't be as strong as we saw in the year's first half.
Does that mean the stock market will end the year lower?
That depends to some degree on inflation. There is a good chance inflation is peaking now and interest rates will fall, which should lead to a broadening of the market, a positive. Also, the rate of change of inference, or intelligence, will continue to grow rapidly. We expect investors to embrace not only the LLM companies as they come public but other companies that use AI to gain market share, lower their cost structure, and generate longer-duration cash flows than the AI-buildout companies, many of which have nonrecurring revenue and will face technological obsolescence.
That means U.S. gross domestic product will also broaden out, another positive. The U.S. economy benefits from the country's AI leadership and our relative energy independence. For all the fears about AI's impact on white-collar work, there are more positives than people talk about. AI is unlocking productivity growth, which will curb inflation expectations. We are on the cusp of leveraging both AI adoption and energy independence to drive significant economic growth.
Which companies are using AI most successfully to increase their competitive advantage and lower costs?
I recommended DoorDash at the January Roundtable and it hasn't performed well, but we still really like it. After Anthropic's Claude Opus 4.5 release, the market became concerned about digital and software companies. Digital companies like DoorDash sold off as the market incorrectly labeled them as having terminal-value risk.
To remind readers, DoorDash has a 65% share of the food-delivery market in the U.S., and has been growing share consistently, as it did in this year's first quarter. It has been broadening its delivery services to include grocery and retail, which has strengthened its relationship with its customers. Over 60% of DoorDash orders come from the top 20% of past subscribers, which is good from a stability and compounding perspective. While growth has remained strong in the U.S. this year, DoorDash has meaningfully accelerated the growth rate at Deliveroo, the European food-delivery leader that it acquired last year, to 20%.
What concerns the market about DoorDash?
One concern is profit-margin expansion. DoorDash announced at the end of last year that it would invest close to $500 million in its technology platform to drive faster growth and more advertising. We think it is on track to do that, and margins could expand significantly over the next several years.
The market has also been concerned about the disintermediation of digital businesses by AI. But my focus is on the fact that it costs DoorDash about $4 less to deliver an order today than when we first invested in it as a private company in 2020. That's because of the physical moat created by its Dasher [delivery person] network, which compounds with density.
Some of the biggest chains and retailers, including Domino's Pizza and Dollar General, are outsourcing delivery to DoorDash. Grocery is reaching an inflection point as younger consumers start to embrace it. And the international business, which doesn't make any money, is turning around faster than we expected.
The new tech platform goes live in the coming months. We expect earnings growth, which has been in the low 20% range, to accelerate. We see DoorDash earning more than $10 a share in 2028, and growing by more than 30% a year. And we see the stock, now around $188, approaching $300.
How about a new name?
There has been a focus this year on megacap IPOs, starting with SpaceX. We are excited about Bending Spoons, which we invested in first in 2023. The stock came public on July 1. As a private company, Bending Spoons distinguished itself by leveraging superior human capital and culture and a superior technology platform to buy modestly growing but undermanaged businesses at attractive prices.
When we first got involved, the company had just acquired Evernote, a note-taking app. It generated about $90 million of revenue and was break-even. Within two years, Bending Spoons drove revenue to $130 million, and achieved an Ebitda [earnings before interest, taxes, depreciation, and amortization] margin of about 60%, in line with the average across its businesses. When we first invested, Bending Spoons was making under $500,000 of Ebitda per Spooner, or employee. Today Ebitda is more than $1 million per Spooner. That speaks to the investment the company is making in the technology businesses it buys.
Bending Spoons is centralized. Evernote has fewer than 20 people at the application level. We believe Bending Spoons' revenue and Ebitda per employee will continue to compound, allowing the company to drive better organic growth and strong operating leverage. The market's concern about software companies should allow management to buy higher-quality companies that fit its model at attractive prices.
Bending Spoons came public at $29 a share, traded up to nearly $44, and now sells for around $35. What sort of upside do you see for the stock?
Based on our estimates, the stock came public at a low-teens multiple of free cash flow. Between organic and acquired growth, we think free cash flow can compound in the 30% range for the next three years. This implies a price target of around $45 to $55, based on 13 to 15 times 2028 free cash flow.
Flipping to the physical economy, Ferguson Enterprises has many parallels with XPO, which I recommended in January. It is a distributor of residential and commercial building products. It is a durable-growth business and a leader in its end markets. It is a scale leader in a scale business, and it is led by an excellent CEO, Kevin Murphy.
Since Murphy took over the North America business in 2017, Ferguson has compounded value at 18% a year, versus the S&P 500's 13%.
Ferguson's customers value availability, convenience, and price, in that order. Scale matters because it enables better supplier margins and higher inventory availability. The key for us is that Ferguson is investing in technology, which allows it to be more efficient, and it can reinvest its savings in better service and better prices to drive scale, much like XPO. Revenue has outgrown its industry in both up and down housing markets, which we like. We expect it to outperform in an up market as the economy broadens out.
(MORE TO FOLLOW) Dow Jones Newswires
July 10, 2026 14:09 ET (18:09 GMT)
Copyright (c) 2026 Dow Jones & Company, Inc.
Comments