Inflation is high in the United States. The 10-year Treasury yield has reached its highest level since 2002. The Federal Reserve may raise interest rates as soon as this month. Despite these conditions, the S&P 500 remains within single digits of its all-time high. Expected profit growth has outpaced the effect of higher rates on prices. Beneath the index, however, the past week was more difficult. The S&P 500 slipped about 1%, while chip stocks fell about 9%. Each of the five stocks discussed fell more than 9%.

The author applies a screening filter to identify potential buys. A company must meet three conditions:

• Analysts expect its profits to grow at least 15% next year.
• Analysts have raised their profit forecasts since July.
• The stock price has not risen in line with those higher forecasts.

The author states that he retired at 46 after a career that generated billions of dollars in profit for companies including Target and Amazon, and that he applies the same methods to his own investments.

Interest Rates and Stock Prices

The author compares interest rates to gravity acting on asset prices. Low rates exert a weak pull. When a safe Treasury bond pays about 5.3% a year, a stock must offer a higher expected return to justify its risk. Higher rates therefore push stock prices down. The author notes that the pull has strengthened. On an earnings call, the chief executive of the largest U.S. bank was asked whether the 10-year yield could reach 6%. The author states that such a move cannot be ruled out.

Profits are offsetting this pressure. Since the end of June, expected profits for the S&P 500 have risen about 10%, while prices have risen about 3.5%. As a result, the market became cheaper relative to profits while it continued to climb.

Before this earnings season, 72 S&P 500 companies told investors to expect profits above forecasts. This is the highest number since fact sheets began tracking this data in 2006.

Investor Sentiment on AI

Steve Eisman, the investor known for The Big Short, has warned that the AI boom could end badly. He still owns Nvidia. He said, "Nvidia, God bless them, and I own the stock."

Profit Growth and Market Risk

Expected profit growth for the S&P 500 is forecast to fall to about 3% by the middle of next year. The author cites historical data showing that the market usually absorbs this slowdown. In eight of the last 10 instances when profit growth peaked, the S&P 500 was higher one year later. Significant declines occurred when profits began to fall.

On the days in the past year when the 10-year yield rose most sharply, the S&P 500 fell about 1%. Three AI hardware stocks in the analysis fell between about 2.4% and 3% on those days. The author states that AI hardware stocks move much more than the broader market. He suggests buying such stocks in stages and keeping the riskiest positions very small.

The author uses a "signal ladder" to rank each stock's position under his investment system. The rungs of the ladder change slowly, but a sharp price decline can move a stock through several rungs in one day. The author expects the readings to stabilize when the market does.

Stock Analysis

Vertiv

Vertiv manufactures power and cooling equipment for AI data centers. Its chief executive said, "Our global pipeline momentum remains very strong and we expect another year of robust orders growth."

• Earnings: Profit per share rose about 60% last quarter. The company raised its full-year outlook across all segments.
• Stock performance: The stock fell about 31% since mid-July and about 45% from its high. It dropped more than 13% on the day it reported results.
• Reasons for the decline: Sales were light last quarter because several large projects took longer than planned. Vertiv also made several acquisitions. The largest was a $1.4 billion deal for a company that produces power equipment located next to data centers. Many data center projects cannot receive power from the grid quickly enough.
• Forecasts: Analysts expect profits to grow about 36% next year. Forecasts have risen since July.
• Product cycle: Nvidia plans to change how power is delivered to server racks starting in 2027. Vertiv expects to release equipment for this design by the end of this year.
• Valuation: The PEG ratio is about 1.0, which the author describes as a fair price for the expected growth.
• Technicals: The price is below its 50-day and 200-day moving averages. The author notes that the further a price sits below these averages, the more compressed the trend becomes.
• Signal ladder: The price has dropped to the "buy a lot" level. The author's system agrees with this classification, but advises a small position because the trend has not turned.
• Next event: Vertiv reports on October 28. The author does not buy or sell within one to two weeks of an earnings report. He would consider waiting for the results and then investing gradually through dollar-cost averaging.

nVent

nVent provides electrical enclosures, power distribution, and cooling systems. These products keep power flowing and chips cool inside data centers.

• Sales and guidance: Sales rose about 53% last quarter. The company expects to sell more than $2 billion of data center equipment this year, more than double last year's level.
• Business mix: Data centers accounted for about 12% of the business when nVent became an independent company. They now account for close to 60%.
• Forecasts: Analysts have raised next year's profit forecast by about 14% since July. Profits are expected to grow about 27% next year.
• Stock performance: The stock is down about 9% over the same period.
• Competition: In July, Schneider Electric's cooling system was incorporated into the design of AMD's new AI rack. The author identifies competition as the principal risk.
• Acquisition and debt: nVent acquired a company that makes data center power equipment for about $1.75 billion. It financed a large part of the purchase with debt at just over 6% interest. Higher rates increase the cost of this debt.
• Signal ladder: The ladder rates the stock as "buy a lot," but the author's system lowers the rating to "buy." The stock remains expensive compared with its own history.
• Technicals: The price is about 9% below its 50-day moving average and just above its 200-day moving average.
• Next event: nVent reports at the end of the month. The author plans to wait for the report and may begin a small position afterward.

Onto Innovation

Onto Innovation makes inspection tools that check AI chips and memory during stacking and packaging. The author compares these tools to a home inspector who checks each floor during construction. A single defective layer can ruin a chip that costs tens of thousands of dollars.

• Sales and backlog: Sales set a company record last quarter and rose about 35%. The order backlog exceeded $1 billion for the first time.
• Growth outlook: The inspection and packaging segment is expected to grow about 80% this year. One packaging customer alone ordered more than $200 million of inspection tools. Most of these units will not ship until 2027.
• Forecasts: Analysts have raised next year's profit forecast by about 18% since July. In the past month, 10 analysts raised their forecasts, and none lowered them.
• Stock performance: The stock is down about 26% since July.
• Valuation: Profits are expected to grow about 44% next year. The PEG ratio is about 0.6, which the author describes as below one, meaning the expected growth is available at a discount.
• Risks: A large share of revenue comes from memory manufacturers. A Chinese memory maker announced a new chip design that pushed down memory and chip equipment stocks. The $200 million order depends on a single customer. The stock moves about three times as much as the overall market.
• Signal ladder: The price has fallen below the lowest rung of the ladder. The author does not add to the position until he has reviewed the entire business.
• Technicals: The price is well below its 50-day and 200-day moving averages.
• Next event: Onto Innovation reports in early November. The author would wait for the report, keep the position small, and buy gradually through dollar-cost averaging on the way down. He states that no investor can know where the bottom is.

Comfort Systems USA

Comfort Systems USA builds the mechanical, electrical, and cooling systems for data centers, hospitals, and factories. The company also builds sections of buildings in its own plants and ships them to job sites for installation.

• Results: Sales grew about 50% last quarter. Profit per share nearly doubled.
• Backlog: Contracted work totals about $14 billion, up about 73% from a year earlier.
• Customer mix: Technology customers rose from about 43% of sales a year ago to almost 60%.
• Forecasts: Analysts have raised next year's profit forecast by about 11% since July. Of the analysts who changed their estimates in the past month, seven raised them.
• Stock performance: The stock fell about 16% since July and about 15% in the last five trading days.
• Earnings growth and valuation: Analysts expect profits to grow about 23% next year. The PEG ratio is about 1.3, which the author describes as a premium.
• Track record: The company has beaten Wall Street estimates for four consecutive quarters.
• Risks: Data center construction depends on the spending pace of large technology companies. Most of the company's contracts are fixed-price. If labor or materials cost more than planned, the additional costs reduce profit.
• Signal ladder: The rating has moved from "buy a lot" to "buy a little," a reduction of two rungs. The stock trades near the top of its historical valuation range, and the system's earnings estimates vary widely.
• Technicals: The price is about 11% below its 50-day moving average and about 8% below its 200-day moving average.
• Next event: The company reports in a few days. The author plans to wait for the results before building a position.

Innodata

Innodata trains and tests AI models for large technology companies. The author describes it as a tutor and examination board for AI systems. It is the smallest company in the group and, according to the author, the most volatile.

• Results: Sales rose about 58% last quarter, a record for the company.
• Forecasts: Profits are expected to grow about 27% next year.
• Stock performance: The stock is down about 15% since July and trades at about half its June peak.
• Growth initiatives: The company is building a lab to capture motion data for robots. Management sees an opportunity to test AI models for the U.S. government. The author states that neither initiative will appear in current financial results.
• Customer concentration: Two large technology customers account for about 71% of sales. The author identifies this concentration as the principal risk.
• Competitive context: Scale AI is a major competitor. Last year, Meta paid about $14 billion for roughly half of Scale AI and hired its chief executive to lead Meta's AI program. Google, Microsoft, and xAI were reported to be stepping back from Scale AI within days. The author states that companies are reluctant to train models with data from a firm partly owned by a rival. Innodata is not owned by any of these companies.
• Governance and capital: Innodata changed its chief executive at the end of September. It established a plan to sell up to $300 million of new stock. The author states that this plan limits price gains. The stock has moved 10% or more in a single day about 15 times this year.
• Signal ladder: The price has fallen below the lowest rung of the ladder. The author reviewed the business and states that it remains sound, but no clear entry point exists.
• Technicals: The price is about 10% below both its 50-day and 200-day moving averages.
• Next event: Innodata reports in early November. The author would wait for the report. He would place the stock only in the riskiest portion of an investment portfolio, keep the position very small, and add gradually over time.

Outlook

The author agrees with Warren Buffett that interest rates act like gravity on asset prices. If the Federal Reserve raises rates again at the end of the month, the pull on companies in this group would increase. According to the author, gravity dominates prices only when profits begin to fall. Analysts are currently raising forecasts for all five companies, and each will report results over the next several weeks.

The author's plan is to:

1. Wait for the upcoming earnings reports.
2. Buy in stages.
3. Keep positions in the riskiest stocks extremely small.