TheChek News Flow

The technology stocks propelling the S&P 500 have raced ahead while leaving much of the rest of the market behind.

It’s getting to be a silly season for certain stocks.

The technology stocks propelling the S&P 500 have raced ahead while leaving much of the rest of the market behind. Valuations on the priciest tech stocks based on profits and sales are approaching the peak values reached in late 2021, before the big Nasdaq selloff in 2022.

Companies such as Cadence Design Systems, Cloudflare, and Nvidia trade for nearly 20 times projected 2024 sales, a calculation based on market value divided by estimated revenue. Even Microsoft, the world’s largest company at $3 trillion, is valued at more than 10 times estimated sales in its current fiscal year ending in June.

Once upon a time, 10 times sales was viewed as pricey. As Scott McNealy, the former CEO of Sun Microsystems, said 20 years ago, “At 10 times revenue, to give you a 10-year payback, I have to pay you 100% of revenue for 10 straight years in dividends.”

That theoretical exercise isn’t really feasible, because companies have significant costs and can return only a fraction of revenue to investors annually. McNealy’s point was that at high valuations, investors are relying on long payback periods to justify their holdings.

That 10 times sales level now is widely breached as favored growth stocks regularly trade for close to 20 times sales. Some 10% of companies in the Russell 1000 index trade for 10 times sales or more.

It isn’t just tech companies. The five companies with the highest price/sales ratios in the S&P 500 index that have market values of $50 billion or more include Visa, Eli Lilly, and robotic-surgery leader Intuitive Surgical . Arm Holdings, Wingstop, Costco Wholesale, and WD-40 also trade at nosebleed multiples.

Lofty valuations haven’t kept many of these stocks from making huge gains. Nvidia has more than tripled in the past year despite a consistently high price/sales ratio as revenue and earnings have blown past estimates. The same has been true for Lilly as investors ratchet up their expectations for its diet drugs for the rest of the decade.

But the risk of owning highfliers was illustrated by the recent drop in Snowflake, the hot data-aggregation software company that has had one of the loftiest valuations of any sizable public company since it went public in 2020. Its stock is down over 25% to $163 since late February, when it released mildly disappointing financial guidance for 2024. When a stock trades for almost 20 times sales and for well over 100 times earnings, it’s vulnerable even to a small miss.

Some of these companies don’t look ridiculously expensive on other multiples. Nvidia, for one, trades for 36 times projected earnings in its January 2025 fiscal year. The discrepancy between the two multiples is due to shortages of its chips used to power artificial-intelligence applications, a situation that helped Nvidia’s net margins—net income divided by sales—hit 56% in the latest quarter. That trounces Apple’s 25% and the average company in the S&P 500’s 10%, and is closer to Visa and Mastercard, quasi-monopolies that have consistently had some of the highest net margins in the S&P 500. Visa trades at 28 times forward earnings.

The issues for Nvidia are its revenue growth—Wall Street expects a slowdown to 20% in 2025—and the sustainability of its margins, which have more than doubled in the past year. Those risks don’t seem to be reflected in the stock, which gained 10% during the past week to $900 after hitting another record high. The stock is up 80% this year and is valued at $2.3 trillion.

ARK Investment chief Cathie Wood warned about Nvidia this past week. In a letter posted on her firm’s website, she cited Nvidia’s guidance for “sequential” deceleration in growth, and falling lead times for its processing chips, to three to four months from eight to 11 months. “Without an explosion in software revenue to justify the overbuilding of GPU capacity, we would not be surprised to see a pause in spending, compounding a correction in excess inventories, particularly among the cloud customers that account for more than half of Nvidia’s data center sales,” she wrote.

Lilly, whose shares are up 150% in the past year to $765, is being valued on late 2020s earnings, given the growth potential for its GLP-1 diet drugs Mounjaro and Zepbound and its current inability to meet demand. Lilly is expected to earn about $12 a share this year, rising to $18 in 2025 and $30 a share in 2027, leaving it trading at about 25 projected 2027 earnings. Its reliance on diet drugs, however, makes it vulnerable to unexpected side effects from the medicines, other setbacks, and competitive threats.

Companies don’t need to be caught up in the AI or weight-loss hype to earn extreme valuations. Restaurants often carry above-market P/Es— McDonald’s trades for 23 times projected 2024 earnings—but Wingstop is in a class by itself. The chain is valued at nearly 20 times projected 2024 sales and over 100 times earnings. That’s a high valuation for a company that is expected to grow profits and revenue by about 25% in each of the next two years.

Costco Wholesale’s popularity with investors has never been higher, and its valuation shows it. The stock, at around $730, is up 50% in the past year and now trades for 45 times projected profits in its fiscal year ending in August. Costco is a great business, with a fanatically loyal customer base that renews memberships at a 90% annual rate and a “wide moat” due to rock-bottom prices. But the stock historically has traded at 35 times forward earnings. A P/E of 45 is a lot to pay for a company with low-double-digit projected earnings growth in the next two years. The stock fell 7% this past week after a mild earnings miss in its quarter that ended in mid-February. The company’s sales were up about 6% in the quarter and 5% in the first six months of its fiscal year—normally not the stuff of 45-multiple stocks.

Then there are oddball situations like WD-40, the lubricant maker. Its stock trades at around $250, or 50 times projected earnings in its fiscal year ending in August. This is high for a company that isn’t growing rapidly. Earnings this year are expected to be up about 7% to $5 a share, about equal to what it earned two years ago.

The award for most expensive stock, though, may go to Arm Holdings, the chip design company that went public in late 2023. It trades at 35 times sales based on projected revenue in its March 2025 fiscal year. The company has an attractive, capital-light chip design business and reported strong results for the December quarter. But its valuation is off the charts based both on sales and earnings, with the stock trading for 90 times projected earnings of $1.50 a share. And that earnings figure excludes sizable stock-based compensation. Arm’s thin float, at just 10% of its stock outstanding, may be contributing to the valuation.

Many of these companies, especially those in the tech sector, are even more expensive than they look because they continue to emphasize an earnings measure that excludes significant employee stock compensation. It is these earnings figures—and not those consistent with generally accepted accounting principles—that the analysts regularly cite, even though the companies do produce GAAP figures that properly include stock compensation.

Stock compensation is a real expense, as Berkshire Hathaway CEO Warren Buffett and others regularly point out. But tech companies continue to produce a blizzard of profit measures that exclude it, including adjusted net income, “free cash flow,” and adjusted earnings before interest, taxes, depreciation, and amortization.

A range of companies report these inflated, non-GAAP earnings measures, including Tesla, Nvidia, Salesforce, and Oracle. The gap between non-GAAP and GAAP earnings varies—it’s higher for software companies like Snowflake and lower for Nvidia. For Nvidia, the gap was around $1 a share last year, when its non-GAAP earnings were around $13 a share. Tesla had $3.12 a share in non-GAAP earnings last year, but profits including stock comp were $2.60 a share. Snowflake reported a $1 a share non-GAAP profit last year, but it was unprofitable on a GAAP basis.

It’s appropriate to make some adjustments to GAAP earnings, such as adding back noncash goodwill amortization—something Buffett supports—but not stock compensation, especially since most stock compensation is cash-like, restricted stock that is routinely monetized by employees. Not surprisingly, many tech companies are generous with stock compensation because it becomes a costless expense in non-GAAP earnings.

Tech bulls would acknowledge this issue but say that focusing on it would have caused investors to stay away from stocks like Amazon.com that have produced enormous gains. But not every tech company relies on non-GAAP earnings. Some, including Apple, Microsoft, Meta Platforms, and Alphabet, emphasize or only report GAAP earnings. This shows that accounting games aren’t needed to succeed in tech.

Investors should be careful about comparing the P/Es of Apple and other GAAP-focused companies to those that emphasize non-GAAP earnings, because the true P/Es of the non-GAAP companies are higher than they appear.

That’s a surefire way to look, well, silly.

With so many flavors to chose from, experts say it’s essential to understand how your needs and their skills match up

We’re all being deluged with news about how the latest generation of AI is transforming people’s lives, helping businesses be more productive, and even leading to layoffs. But that flood of information doesn’t help anyone answer the most basic question about these AIs: Which is best?

So I canvassed executives, engineers and researchers who are knee-deep in the process of applying the world’s most powerful AIs to real world problems, to find out what they have learned.

Their answers surprised me. There was plenty of practical advice about the relative strengths and weaknesses of AIs from Google, OpenAI, Anthropic and Meta. But the overall message was that the best AI for any task depends on both the user and the task. Their insights also offer a glimpse of where the entire field of AI is going.

In a way that wasn’t true even six months ago, companies can now either embrace the potential cost savings and productivity boost of generative AI—which some researchers believe is on the path to a “general” or humanlike AI—or risk losing out to competitors who will.

Today’s most powerful AIs aren’t something you can buy and run on your own computers. They’re only accessible through the cloud. This makes it easy to test them by feeding them documents, images and text, but also means that businesses have limited ability to alter their behavior.

Testing these AIs is more like hiring an employee than just buying a piece of software off the shelf, says Mark Daley, chief AI officer of Western University in Ontario.

“People expect the chatbot to work right out of the box, but you have to spend time trying them and see which of these will deliver, just like you do with an employee,” he adds.

Daley has found that all of the major large language models—including those from OpenAI, Anthropic, Google and Cohere, a startup that only offers its models to businesses—have their strengths and weaknesses. Which one to use depends on a person’s preferences and the task at hand, and it pays to experiment with them.

Other companies appear to be catching up with the capabilities of OpenAI, but OpenAI’s models remain, for now, the standard by which all others are judged. Earlier this week, Anthropic rolled out Claude 3, a new large language model which the company claims beats the gold standard GPT-4 on every benchmark.

“We are using OpenAI like crazy,” says Brad Schneider, chief executive of Nomad Data, a company that helps large companies use AI. Nomad Data uses OpenAI to digest, summarize and search within huge libraries of documents, such as legal briefs, court cases and insurance claims. The company’s clients also include private-equity firms who might have only a week to digest thousands of documents about a company they’re about to acquire.

After trying all of the most-capable large-language models, Schneider’s company found that none are as good as OpenAI for these kinds of document-processing tasks. Previous versions of Anthropic’s Claude and current versions of Google’s Gemini both hallucinated too often, he found. (In AI, ‘hallucination’ is a term for when a chatbot makes up false information.)

Google senior vice president Prabhakar Raghavan recently wrote that hallucination is a challenge common to all large language models, but that “this is something that we’re constantly working on improving.” Anthropic President Daniela Amodei has said that it is “very, very hard” to get the hallucination rate in such models to zero. The company has said that its latest model is twice as likely as its previous one to answer questions accurately, and that eliminating all hallucinations can make models hesitant to answer questions they would otherwise get right.

In addition to accuracy, the other two big considerations are speed and cost, says Eric Olson, chief executive of Consensus, a scientific search engine.

On a search engine, users expect a response within seconds. Because Consensus pairs its search results with summaries of scientific papers made by GPT-4, the company needs those summaries to be generated nearly instantaneously.

For Olson’s purposes, this means the only truly suitable model is OpenAI’s GPT-4 “turbo,” which can get a user a response within 1.5 seconds, versus twice as long with regular GPT-4. Google’s Gemini and Anthropic’s Claude are also slower than OpenAI’s models, he adds.

That said, this kind of performance comes at a cost. OpenAI and its competitors charge business users of their systems by the token—in essence, by the word—to process their requests.

“We have cases where one question someone asks can cost $50,” says Schneider. That could happen if, for example, someone asks a specific question about a collection of 5,000 legal documents, because the number of calls to OpenAI’s systems could be in the tens of thousands.

While OpenAI and Anthropic duke it out for the title of most capable large language model, Google has been a laggard by many benchmarks.

But one advantage for Google and its customers is that its models have the ability to ingest huge volumes of data in each query. That’s something OpenAI currently doesn’t offer, and Anthropic does for only a small group of customers.

“Gemini 1.5 allows a million tokens of context, and that is an absolute game-changer,” says Daley. “You can feed in 10 textbooks of material, and it can synthesize across that, not perfectly, but better than a human could do given 35 seconds.”

But What About Microsoft?

Microsoft has a couple of challenges in its rollout of AI. First is that, despite its tie-up with OpenAI, the company is in some ways a reseller of OpenAI’s services—which businesses can also buy directly from OpenAI.

To be clear, Microsoft is providing a platform for a lot of different AI models, which it offers through its Azure cloud service. Microsoft also has a partnership with Mistral, for example, and offers Meta’s open source Llama model.

“With Azure AI we are bringing the most comprehensive selection of high-performing open and frontier models to customers on the world’s most trusted cloud platform,” says Eric Boyd, vice president of AI platform at Microsoft.

Amazon’s cloud services have a similar strategy, and the company has partnered with Anthropic.

When OpenAI adds a new feature, there is a meaningful delay before it becomes available in Microsoft’s version of those models, says Schneider. Microsoft’s version of GPT 4 also seems to be capacity constrained in a way that OpenAI’s is not, leading to stricter limits on how many tokens a minute businesses can buy from it, he adds.

For many specialized applications of generative AI, companies may want to build and train their own AI—or pay someone else to do it for them, says Petr Baudis, chief AI architect at Prague-based Rossum. Rossum automates the processing of invoices for companies, using a variety of AIs which its research team built themselves.

Training your own large-ish language model might sound like an impossible task, but with the rapid development of open source models like Meta’s Llama, it’s the sort of thing that even a small team can accomplish.

Everyone I spoke to for this piece said that open source large language models, which are rapidly becoming more capable, can be operated at a fraction of the cost of accessing OpenAI and Google’s models. There are a couple of reasons for this. The primary one is that these models are much smaller, and therefore require less power to run. The second is that since they can run on a company’s own servers, they cut out the middleman of the big AI companies and their margins.

Custom, open source AIs can also outperform the biggest large language models if they’re trained on the right data, and asked to do a sufficiently narrow task—such as the invoice processing service that Rossum provides.

Generative AI is a technology evolving at a rate not seen since the go-go days of the early internet itself. Anthropic’s release of a model that seems every bit as capable as OpenAI’s, despite having a smaller team and having been founded much more recently, suggests that large language models may become commodities. At that point, the only thing that will matter will be which company can offer the fastest response at the lowest price.

The beneficiary of that fierce competition will be companies large and small, which could see significant increases in the productivity of their employees. Those gains will come at a fraction of the cost of paying humans to do the same knowledge work. The implications for the future of white collar jobs are obvious—and worrisome.

Many experts thought high interest rates would break the economy and inflation couldn’t be tamed

In mid-2022, JPMorgan Chief Executive Jamie Dimon warned that a “hurricane” was about to hit the U.S. economy.

It could be “a minor one or superstorm Sandy,” Dimon said at a financial conference in New York. “You’d better brace yourself.”

Last year, Bridgewater Associates founder Ray Dalio predicted a “debt crisis” after earlier anticipating a “perfect storm” of economic pain.

High-profile investors and economists including DoubleLine’s Jeffrey Gundlach and Rosenberg Research’s David Rosenberg were just as fearful, with Gundlach last March saying a recession would come “in a few months.” Early last year, economists predicted a 61% chance of recession in 2023.

The experts were way off. They underestimated the impact of government stimulus and the resilience of consumers and businesses. And they were too skeptical of the Federal Reserve’s ability to push inflation lower without sparking a recession. The economy continues to grow at a steady clip. Inflation is getting closer to the Fed’s goal of 2%, unemployment remains near a half-century low and the stock market is near record highs.

“I was bearish on the economy,” Dalio said in an interview. “I got it wrong.”

Said Dimon, also in an interview: “I would have thought some of the fiscal stimulus would have worn off by now.”

Where we thought we were heading

At the time, the downbeat predictions made a lot of sense. Between March 2022 and last summer, the Fed raised interest rates 11 times to what is now a 23-year high.

In the past, surging rates have preceded recessions. Other telltale recession signs showed up, too. The yield curve inverted and remains that way, meaning yields on short-term U.S. Treasury bonds are higher than those on long-term bonds. The pace of GDP growth dropped, and inflation soared. Even Fed Chair Jerome Powell said there was only a narrow path to bring down inflation without pushing the U.S. economy into recession—the hoped-for “soft landing.”

Where we are now

So what happened? Many people were relatively unscathed by the higher rates—for example, those who locked in low-rate mortgages before the Fed’s rate increase or own their homes outright.

It’s also the case that U.S. politicians were much faster to act than in decades past to aid the economy. Several rounds of pandemic-related relief payments and other spending programs left many consumers and companies with extra cash, which fueled continued spending. Consumer spending accounts for roughly 70% of the U.S. economy.

“I got it wrong because, ordinarily, when you raise interest rates it curtails private-sector demand and asset prices and slows things down, but that didn’t happen,” Dalio said. “There was a historic transfer of wealth: The balance sheets of the private sector improved a lot and the balance sheet of the government deteriorated a lot.”

A spokesman for Dimon said the JPMorgan chief gives a range of possibilities and weighs them, but doesn’t make predictions.

At the same time, a resilient stock market and bonds with higher yields have produced a positive “wealth effect” that encouraged consumers to spend rather than retrench.

It’s also true that interest rates aren’t that high, historically speaking: They just feel that way because the U.S. went through an extended period of superlow rates.

Ed Yardeni, an economist who has been more bullish than most over the past two years, attributes some of the mistakes of the stars to excessive fear of higher rates.

“The economy can function just fine where interest rates are,” he said.

In hindsight, some indicators, such as the inverted yield curve, did accurately anticipate severe problems, such as last year’s regional banking crisis. But the Fed did a good job putting out fires in banking, and the troubles didn’t spread to other sectors.

Not out of the woods

The probability of a soft landing would appear to have improved, with the Fed signaling a willingness to cut rates this year. Economists recently pegged the chance of recession within the next year at 39%, down from the 61% a year ago.

But some of those who predicted tough times say difficulties are still on their way. Gundlach recently predicted a recession this year and a drop to 3200 for the S&P 500, which closed Friday at about 5100. Gundlach didn’t respond to requests for comment.

For his part, Rosenberg said gross domestic income, adjusted for inflation, has been flat over the past year, the job market is weaker than it appears and personal income isn’t keeping up with spending. In an interview, he said the economy’s troubles will be clear later this year.

Dalio says key questions are whether productivity will continue to improve, as businesses embrace artificial intelligence, and whether all the debt the government has piled up—and the resulting debt payments that will have to be made—will come back to haunt the economy. The U.S. government is expected to pay an additional $1.1 trillion in interest over the coming decade, according to the Congressional Budget Office’s latest estimates.

As for Dimon, he says he remains worried about serious issues facing the economy, including inflation remaining too high, geopolitical upheaval, and rising government spending and debt. He says so-called stagflation, or slow growth with unhealthy inflation, is a concern.

“I wouldn’t count my eggs yet,” Dimon said.

Plus, US student athletes unionise

Concerns about the health effects from PFAS are mounting, and efforts to ban the chemicals are gaining traction. These companies are profiting from the cleanup.

When Teflon cookware hit the market in the early 1960s, it was hailed as a modern chemical marvel and a big win for home cooks: Eggs slid effortlessly off “happy pans”; cleanup dispensed with the drudgery of scrubbing.

Teflon pans are still sold, but one of the original chemicals is long gone due to concerns about its health effects. Today, the big money isn’t in making those “forever chemicals,” a broad class of substances known as polyfluorinated substances, or PFAS. Rather, it’s in cleaning them up—an effort that could cost more than $200 billion.

Manufactured by companies such as DuPont spinoff Chemours and 3M, PFAS are a class of thousands of chemicals that have been used for decades in consumer and industrial goods. Dubbed forever chemicals for their remarkable stability, they can resist water, stains, grease, and heat. That has made them ideal for food packaging in everything from pizza boxes to microwave popcorn bags. They’re also used in thousands of products from carpeting to firefighting foam.

Now, concerns about their health effects are mounting, and efforts to ban the chemicals are gaining traction.

Europe has taken the lead in proposing a blanket ban on PFAS, made up of chains of carbon and fluorine linked together. The U.S. is also making progress. The Food and Drug Administration recently announced that grease-proofing PFAS in food packaging is being phased out. PFAS manufacturers such as 3M are also phasing out the chemicals. And new rules are expected soon from the Environmental Protection Agency to vastly reduce PFAS in drinking water.

Yet companies still face massive liability claims. While DuPont, 3M, and other manufacturers have already paid out billions in settlements, a wave of corporate liability continues to build. Some landfill operators and wastewater companies could also be on the hook for cleanups and liability as new federal rules ramp up.

Settlements could reach hundreds of billions, putting PFAS in range of the $206 billion tobacco settlement of the 1990s, according to a recent report by BofA Securities analyst Dimple Gosai. “PFAS is a risk everyone needs to hear about now,” she says.

A handful of companies, meanwhile, are profiting as demand rises for testing, monitoring, and remediation. Companies that could benefit, according to analysts, include Aecom, Montrose Environmental Group, Republic Services, and Xylem.

Phaseout Push

While efforts to ban PFAS have percolated for years, they are gathering steam due to mounting concerns about the chemicals’ toxicity. The substances don’t break down, accumulating in soil and water, and they have been found in the bloodstreams of nearly every American. Long-term exposure has been linked to some cancers, including kidney and testicular cancer, fertility and thyroid problems, and other health effects.

“Manufacturers used carbon-fluorine chemistry because it was so stable,” says Tarun Anumol, director of global environment and food markets at Agilent Technologies, a laboratory technologies company that makes instruments for measuring and analyzing the chemicals. “It is coming back to haunt us.”

Manufacturers have already paid out billions in liability claims over the chemicals. 3M, DuPont, and two other companies agreed to class-action settlements of more than $11 billion last year.

But some states and municipalities are seeking to reopen the settlements, while other lawsuits go forth. A class action filed in September by Dutch lawyers against DuPont and Chemours accuses them of decades of pollution at a plant south of Rotterdam. Late last year, the city of Wausau, Wis., filed a suit against 15 manufacturers and 61 insurance companies alleging contamination of the groundwater.

“Unlike asbestos and tobacco, PFAS touches almost every segment of humanity; it is in all of us, it’s even in the polar bears,” says attorney Michael London with law firm Douglas & London, who is leading several of the class action cases.

3M said the 2023 settlement wasn’t an admission of liability but CEO Mike Roman called the agreement “an important step forward” for the company. DuPont and its spinoffs Chemours and Corteva have said they “deny the allegations in the underlying litigation.”

Some types of PFAS were eliminated decades ago, but manufacturers are now phasing out more. 3M says it will stop making PFAS by the end of 2025 and recently told investors that manufacturing volumes are down 20%, citing “very good progress.” DuPont says it has phased out long-chain PFAS and replaced PFAS in firefighting foams with alternatives.

Other companies are getting rid of the chemicals too. Restaurant Brands International, which owns Burger King, Popeyes, and Tim Hortons, said it would ban PFAS in food packaging globally by 2025. Outdoor gear retailer Patagonia has pledged to eliminate PFAS across its entire product line by 2025.

Completely eliminating PFAS isn’t likely, however, because they remain essential to some types of manufacturing. Cutting PFAS out of microchip production, for example, would require manufacturers to come up with a whole new class of chemicals and revamp production processes.

Contaminated drinking water remains a problem. Nearly half of the nation’s tap water is estimated to have one or more types of PFAS. Contamination has come from decadeslong use of PFAS in firefighting foam. The chemicals have also leached into water supplies from landfills, wastewater-treatment plants, and manufacturing sites.

The European Union is now considering a ban on the production and import ation of around 10,000 PFAS across the bloc, with a possible vote in 2025. In the U.S., the EPA has proposed rules that would require water utilities to filter out certain PFAS. Ten states have established enforceable stands on PFAS in drinking water, while 13 more have adopted guidance levels, according to Safer States, an environmental advocacy group.

The EPA is also expected to designate PFAS variants, PFOA and PFOS, as “hazardous substances” under the Comprehensive Environmental Response, Compensation, and Liability Act, or Cercla, informally known as the Superfund bill, sometime this year. This would give the agency more authority and tools to hold polluters accountable for cleaning up contamination and recover costs from responsible parties. In February, the agency proposed adding nine PFAS to its list of “hazardous constituents.” The rule is in a comment period through April 8.

Billions of dollars are flowing for cleanups. Congress has set aside $9 billion to address PFAS-contaminated drinking water. Analysts say new regulations and federal funding along with consumer awareness and more lawsuits will boost demand for testing and remediation.

It will be a Herculean task. A recent study found that 57,412 sites across the U.S., including 519 major airports, are likely contaminated with PFAS. Global Infrastructure consulting firm Aecom pegs the U.S. remediation opportunity at around $200 billion and the global market at $250 billion. “Without a doubt, the regulatory setting provides the tailwind,” Lara Poloni, Aecom president, tells Barron’s.

Companies involved in cleanup efforts span three broad areas: testing and removing PFAS from drinking water; remediating soil and landfills; and providing consulting services, including assessing regulations and compliance risk. While there aren’t many pure-play companies, several firms provide testing, monitoring, and remediation services as part of a broader business.

On the water side, proposed EPA rules set the bar for PFAS incredibly low. The EPA wants a limit for PFOA and PFOS at four parts per trillion each in public drinking-water—equivalent to one drop in six Olympic-size pools, according to Agilent’s Anumol. That is far lower than the EPA’s maximum level for arsenic in drinking water at 10 parts per billion.

The American Water Works Association, an industry trade group, estimates that at least 5,000 water utilities will have to install advanced treatment systems or find new water sources, while an additional 2,500 will need costly overhauls or upgrades to filtration systems. According to a report commissioned by AWWA, water utilities will need to invest more than $50 billion in treatment technology over the next 20 years to comply with new PFAS standards. If PFOA and PFOS are designated hazardous substances, it could add $3.5 billion a year in disposal costs.

Deane Dray, a managing director at RBC Capital Markets, says Xylem is his top pick in the water sector. Last year the firm acquired Evoqua, a leader in remediation of emerging contaminants, including PFAS. Xylem has worked with municipalities across the country and offers a variety of PFAS treatment technologies, including granular activated carbon, or GAC, which is like a giant Brita filter for municipal systems, notes investment banking firm William Blair in an extensive report on forever chemicals published last year.

“PFAS is going to be one of the growth drivers for the company,” Dray says. “The floodgates will open with [water] utilities that are going to raise their hand and say ‘we are over the limits, we need help, we need remediation.’ ”

Xylem—a Barron’s stock pick in 2023—will be a winner, says Dray, because “there will be more demand over the near term than the industry can satisfy.” The company is expected to book sales of $8.5 billion in 2024 with $988 million in net income, up 20% from last year.

Shares trade around $128, up 12% this year. At a price/earnings ratio of 27 times 2025 earnings, it looks pricey and will have to live up to growth forecasts. Dray says the high multiple is warranted, partly because of a tailwind from PFAS services; he sees the stock reaching $145 over the next year.

On the waste removal side, there’s “an arms race” for companies to safely transport or destroy PFAS on contaminated sites, says Dray. After the chemicals have been separated from the water, the hazardous material has to go somewhere, and burning or burying is problematic. PFAS that are incinerated can rejoin air and burying may not stop them from leaching into groundwater; incinerating is energy-intensive and can release harmful chemicals into the environment.

One company with a promising PFAS destruction offering is privately held Aclarity. CEO Julie Bliss Mullen says the company has developed a very low-energy system that uses electrochemical oxidation to break PFAS bonds on-site, preventing the concentration and movement of PFAS from landfills and other sites. Dray says Aclarity is the farthest along with a scalable solution to destroy PFAS. “Even though it’s a private company, investors should be keeping an eye on where this company is going,” he adds.

One of its rivals for on-site destruction is Aecom, which has been tackling PFAS for more than two decades. Aecom’s technology, called De-Flouro, destroys the compounds in industrial waste and water through electrochemical oxidation. The company says PFAS water services account for 1% of revenue but over the next ten years it believes it will “benefit from $10 billion of PFAS gross revenue opportunity,” according to Poloni, with the vast majority of revenue coming from its consulting services.

Aecom’s revenue is estimated at $15.6 billion this year, and shares trade for 17 times 2025 earnings, a roughly 15% discount to the market, based on consensus estimate. The stock also looks inexpensive with a price to earnings-growth (PEG) ratio of 1. According to FactSet, analysts on average see the price rising to about $105 over the coming year, from recent prices around $90, for a gain of about 16%.

Another way to play the cleanup opportunity is Montrose Environmental. The company sells testing, monitoring, and environmental consulting services, including a business for PFAS. CEO Vijay Manthripragada says the firm has developed unique ways and patents to remove and concentrate PFAS waste. “We help prevent the need to actually transport, bury, or burn the waste. That’s been really compelling for a lot of our industrial clients,” he explains.

PFAS remediation services should make up about 12% of Montrose’s $700 million in projected revenue this year. Tim Mulrooney, lead analyst for global services at William Blair, sees it as a major growth driver, expecting it to be “several multiples” of current revenue in a few years. “My investment thesis is simple,” he explains. “Montrose is small. The PFAS market is going to be very, very big. And Montrose has a better mousetrap for certain applications.”

One opportunity lies in a push to eliminate heavily contaminated sites where PFAS could be as high as 2,000 or 3,000 parts per trillion, says Mulrooney. That’s where the company’s technology “really shines,” he says, “and where I think Montrose has a leg-up against many competitors.”

Another fan of the stock is Randy Gwirtzman, portfolio manager of the small-cap Baron Discovery mutual fund. Montrose is a “one-stop shop” because it covers the full spectrum of services, he says. “Montrose fit the mold of what we like: a strong management team, strong organic revenue growth in a fragmented industry that can be consolidated, leading to higher overall growth,” says Gwirtzman.

Waste removal could see higher demand, benefiting Republic Services. The company mainly does recycling and solid waste removal, but it also owns five hazardous-waste facilities that could benefit from PFAS cleanup. It acquired an environmental solutions company, US Ecology, in 2022. “That put us from the back seat into the driver’s seat in terms of taking care of [PFAS],” says Jon Vander Ark, Republic’s CEO.

Republic doesn’t break out PFAS services as a revenue segment, and it remains a small part of the firm’s estimated $16 billion in sales this year. But Vander Ark says there’s a “good pipeline in 2024,” fueled by corporate demand for cleanups ahead of new EPA rules, and some Defense Department work. “This year it will be comfortably in the nine figures and we think over time it could be a billion-plus dollar opportunity,” he tells Barron’s.

Other growth drivers for Republic include renewable gas projects, environmental services, and digital platforms to improve productivity and margins.

It all makes for an expensive stock at 31 times estimated earnings of $5.99 per share in 2024. And Republic has already notched big gains, including a 43% rise last year. But the stock has long been pricey while managing to outperform the S&P 500 by an average 4% over the last five years. Helping get rid of forever chemicals could keep Republic shares moving up.

Farmers across Europe have been sticking it to the authorities over the past month, sometimes literally. Agrarian protesters showered police with liquid manure and eggs outside European Union headquarters in Brussels last week.

Farmers taking grievances to the streets is nothing new, in Europe or elsewhere. But this wave threatens pillars of EU environmental and foreign policy. It could help far-right parties grab a quarter of the European Parliament in elections this June, on current polling.

“The EU’s Green Deal will have to change to survive,” says Eoin Drea, senior researcher at the Wilfried Martens Centre for European Studies. “Centrist parties don’t want to lose the rural vote like the Democrats did in America.”

Agriculture is an underrecognized contributor to pollution and climate change around the world. In Europe, the sector emits nearly two-thirds as much carbon as transportation, Drea says.

Von der Leyen, hoping the next European parliament will hand her a second term, has backpedaled fast. Last month she proposed freezing the pesticide reduction plan. Her commission opened exemptions to the 4% fallow requirement. “Our farmers deserve to be listened to,” she told legislators.

“We have seen a U-turn away from an ambitious Green Deal,” laments Ineke Maes, an adviser to the Belgian Better Environment Federation.

The Green Deal is one focus of European farmers’ complaints. The other is cheap imports. The EU’s long-pending free trade deal with South America’s Mercosur bloc, which includes agri-powerhouses Brazil and Argentina, is a likely casualty.

“It is impossible to conclude talks in these conditions,” an adviser to French President Emmanuel Macron told reporters in late January.

A more sensitive subject is Ukraine. The EU waived numerous tariffs on the onetime breadbasket of Europe after Russia’s invasion two years ago. That brought cut-price flows of grain and poultry, particularly irritating farmers in former East Bloc countries that most strongly support Kyiv politically.

That’s led Polish Prime Minister Donald Tusk, among other leaders, into a delicate dance. “The interests of Ukrainian agro-holdings cannot override the interests of Europe and our farmers,” he declared after meeting Polish agrarians on Feb. 29. Literally in the same breath he added: “This has nothing to do with Ukraine’s security.”

Rumors of Green Europe’s death should not be exaggerated. The U.S. spews 50% more carbon dioxide relative to gross domestic product than Germany, more than twice as much as France, by United Nations figures.

Flying manure and heated elections don’t create the best environment for hashing out the complex way forward, though. ‘Some of the things we need to compromise about are getting lost in rhetoric,” says Justin Zahra, director of the EU agricultural program at the Environmental Defense Fund.

I’m changing my Bitcoin rating to Incognito Unclear, which means I have no idea where the price is headed next, and it’s doing so well despite my skepticism that I’m wearing a paper bag over my head to hide my shame from the crypto tycoons.

Other assets are shining, too. Gold recently hit a new high. U.S. stocks have reached the point in the cycle where Bank of America calls the S&P 500 “egregiously expensive,” while predicting 8% more upside by year’s end. (It thinks average 10-year returns from here will stink, however.)

But Bitcoin has tripled over the past year, to a recent $67,000. I first wrote about it in June 2011 for something called SmartMoney.com, calling it the “top-performing money in the world.” It had climbed 200,000% in a year, to just over $10. If only I had sold everything and bet it on Bitcoin, I’d be pulling my superyacht up to Bezos’ right now to ask if he has any Grey Poupon. Instead, I continued plopping my pay into boring stocks and bonds and have achieved mere suburban comfort.

I’m too square to turn bullish now. But for those who are tempted by the new field of bitcoin ETFs—brands include Bitwise, iShares, Fidelity Investments, Invesco, and WisdomTree—or those who are wondering when to sell, it would be helpful if there were something more than future-of-money narratives to go by. How about math? The lack of cash flows makes that tricky, but there are creative workarounds. I reached out to a couple of Bitcoin analysts to talk about the models they use. I sniffed them over out of curiosity, not to endorse them. Let’s run through five.

J.P. Morgan has used this one. It treats Bitcoin like digital gold, because the two assets have things in common. Supply is limited; they’re fungible and divisible stores of value; they’re not under government control; and both are apparently durable—Bitcoin has outlived the publication where I first wrote about it. The value of all mined gold is estimated at close to $15 trillion, but much of that is jewelry, or else held by central banks. Gold held by private investors, including through bars, coins, and ETFs, is estimated at $3.3 trillion. Bitcoin’s market cap is around $1.3 trillion. If it’s as good as gold, maybe it should be trading at more than twice its current price.

JPM adjusts Bitcoin’s fair value downward for the fact that it has been more volatile than gold. Reda Farran, an analyst at financial information site Finimize, says that this adjustment is too harsh, because Bitcoin’s volatility has been falling. He also points out that Bitcoin has slower supply growth than gold, and in his opinion more financial utility, both of which warrant upward price adjustments. My main question: If Bitcoin is indeed as good as gold, wouldn’t that point to a substitution effect that would push gold’s price lower, not to record highs? Farran says the substitution effect is evident in ETF flows, but that central banks have been offsetting it by buying gold.

Sometimes analysts point to the cost of mining an ounce of gold as a peg for its price. This is possible with Bitcoin, which is “mined” using massive computing power, meaning that the marginal cost of production is tied to electricity prices, the efficiency of the hardware, and the rate at which miners are rewarded for their computations. This reward rate is engineered to halve every so often, slowing supply growth until it eventually halts, and as it turns out, the next halving is in April.

The anticipation of this, combined with the new demand from ETFs, is part of what has Bitcoin bulls so abuzz. One researcher who has studied using Bitcoin’s production cost as a valuation guide, Adam Hayes of Hebrew University in Jerusalem, predicted last year that it will cost about $75,000 to mine a Bitcoin after the halving. Finimize’s Farran says that production costs should be considered a price floor, not a fair value.

You know those Ibbotson charts that financial advisors point to that show the long-term performance of stocks, bonds, bills, and inflation? If they showed Bitcoin, too, the customers would say, “Give me some of that.” We should all say that, says Raphael Zagury, chief investment officer at Swan Bitcoin, which offers buying plans for the cryptocoin. He reckons that a 20% allocation is the ideal risk/return tradeoff, or as he puts it, “the optimal point on the efficient frontier.” This isn’t so much a pricing model as it is an argument for a higher price. But Zagury has two more models…

The world’s real estate is worth $320 trillion. Some portion of that reflects a “monetary premium,” or pure investment value beyond the practical value of the land and structures themselves. Call it 30%. There’s a chance, according to Zagury, that Bitcoin will become the world’s apex monetary asset, and devour this premium not just from real estate, but also from stocks, bonds, gold, silver, other cryptocurrencies, and fine art. Zagury has created a website, nakamotoportfolio.com, that allows users to enter their own assumptions. The default assumptions show, for example, a 5% probability that Bitcoin captures real estate’s 30% monetary premium over the next 20 years, and a 90% probability that it will grab other crypto’s 100% monetary premium over six years. Using those default assumptions, the calculator produces a fair value for Bitcoin of over $620,000.

A global collapse of so many financial assets at once brings to mind societal chaos on a Mad Max level. I’m not clear on how, in that world, internet coins will be the thing everyone covets. I mean, beef jerky, I could see. But then, I’m the guy who didn’t buy bitcoin at $10.

There are financial instruments that pay off in the event of a bond default. They’re called credit default swaps. There are CDS for Treasuries, even though the U.S. government can just make new money if it needs to. The presence of a Treasury CDS market implies some risk of default, however small. You can use the pricing in that market to calculate a default probability. If you assume that Bitcoin wins if Treasuries go kablooey, you can also use CDS pricing to calculate a fair price for Bitcoin. Zagury has. He gets between $75,000 and $100,000, “depending on your assumptions.”

Hope for interest-rate cuts has been replaced by stronger economic growth as fuel for stocks.