In the third quarter, we saw a continuation of trends that began accelerating in the second half of Q2. Growing optimism surrounding the AI boom and the corresponding surge in electricity demand have helped power large- and mid-cap growth stocks to new highs regardless of business model quality or valuation. Honorable mention goes to continued chasing of speculative bets on crypto currencies where a business model is not required and where leverage is treated like vegetables, the more the better. If only our children would consume vegetables like crypto “investors” crave leverage. Meanwhile, corporate earnings broadly exceeded expectations, with profits for the S&P 500 Index up 10%, which is more than twice the growth forecast at the start of the quarter. 2025-2027 S&P 500 forecasted earnings are slightly lower than at the start of the year making the index more expensive following year-to-date gains.
Still, pockets of economic weakness persist, including in the residential housing market where several industry participants have recently called out deteriorating demand in new construction and remodeling activity. Even pockets of historical resilience, like replacing an air conditioning unit when it fails, are exhibiting abnormal demand weakness. Affordability is a persistent issue. Several negative economic datapoints on the job market developed in September, including disappointing job creation in August coupled with a larger-than-expected downward revision to non-farm payrolls for the prior year, dropping employment estimates by almost one million jobs.
As the chart below shows, changes in the trajectory of the job market, as evidenced by initial jobless claims, tend to persist for some time. A 20% spike in unemployment claims has foreshadowed recessions ahead coupled with a drop in equity valuations.
Source: FactSet Research Systems Inc. and The Leuthold Group, monthly data, 12/31/1968 to 8/29/2025 This chart shows the year over year percent change for initial jobless claims on a 12-month moving average verses the S&P 500 normalized price to earnings ratio. The S&P 500 Index is an index of 500 U.S. stocks chosen for market size, liquidity and industry group representation and is a widely used U.S. equity benchmark. Initial Unemployment (“Jobless”) Claims is a measure of the number of jobless claims filed by individuals seeking to receive state jobless benefits. Chart peaks for YoY % Change not shown are at a high 582% on 2/28/2021 and at a low of -75% on 3/31/2022. All indices are unmanaged. It is not possible to invest directly in an index. Past performance does not guarantee future results.
We’re not forecasting the timing of a recession, but history indicates that it’s unavoidable. When jobless claims start to rise, the trend generally doesn’t magically reverse course on its own. The Federal Reserve’s messaging suggests that board members are increasingly concerned about a further weakening of the labor market. Historically when the job market weakens, business valuations fall. For the moment, equities have been resilient. The Russell 3000® Index gained 8.18% in the quarter as market participants viewed “bad news as good news” in hopes that the Federal Reserve might be more aggressive in cutting rates down the road. We are mindful, though, that capital markets are a discounting mechanism and government debt is yielding 3.6-3.8% between six months and five years to maturity suggesting some further easing is already priced in. With inflation still tracking above 2%, significantly lower bond yields could signal a weakening economy.
Attribution Analysis
In the quarter, our Strategy gained 4.65%, trailing the Russell 3000® Value Index, which was up 5.63%. Stock selection was positive in 5 of the 11 sectors, led by Communication Services, Consumer Staples, and Information Technology.
Across the capitalization spectrum, our selection effect was generally positive as well, with a few exceptions including mid- to large-cap companies valued from $12.9 billion to $31.5 billion. That was largely owing to a steep selloff in shares of Centene, which was the biggest detractor to our performance in the quarter (a discussion of the stock can be found below).
In general, our process — which forces us to pay attention to downside risks along with a range of possible outcomes — hurt our performance against this current backdrop. So did our underweight to cyclical, high beta stocks. In the long-run, though, we are convinced that this process sets us up for success.
Portfolio Activity
We believe valuations are approaching historical extremes, and our process, driven by our 10 Principles of Value Investing™, causes us to pay close attention to the intrinsic value of a holding. As a result, during the quarter, we continued to reduce our exposure to businesses that appear to be fully valued.
We construct our portfolio from the bottom-up, based on individual company fundamentals rather than sectors or market-cap preference. However, as we’ve been looking for opportunities lately, we have generally found more in the mid- and large-cap space in the quarter, given the strong rally in small caps.
During the quarter, we initiated a position in the online retailer and cloud computing giant Amazon.com (AMZN). Until recently, Amazon traded at a meaningful premium to the S&P 500 Index and retail juggernaut peers. Management tends to invest capital in massive waves, which creates significant earnings noise. For example, AMZN doubled its retail distribution center footprint in 24 months following the COVID-19 pandemic, causing the company’s free cash flow and profit margins to drop. Today, however, the areas management is aggressively investing in — such as Amazon Web Services (AWS) and advertising— are high-margin businesses. That should lead to better long-term profitability despite likely short-term headwinds created by costs added ahead of incremental revenue.
Today, the company is spending big on AI infrastructure, which should allow it to sell more computing power via AWS while also making the company more efficient via process automation. Historically, Amazon was seen as more capital-intensive than large-tech competitors including Alphabet (Google), Microsoft, Oracle, and Meta (Facebook). Today, AMZN is the most capital efficient operator as measured by capital expenditures relative to sales. For example, a former holding is investing so aggressively that the company is free cash flow negative for the first time in the company’s public history. Yet the market has rewarded these investments with a much higher valuation.
When considering the value of its high-margin businesses like AWS, ads, and subscriptions, Amazon appears to be undervalued. In fact, based on our research and evaluation, we believe shareholders are getting AMZN’s dominate e-commerce business for free, providing a margin of safety. The retail business provides significant free cash flow that can be utilized for investments in higher margin opportunities. Most businesses require capital to grow, but Amazon’s retail business operates with negative working capital meaning Amazon gets paid before they pay suppliers. As a result, the faster retail grows, the more cash available for attractive long-term investments.
The stock currently trades at a 15% discount to the S&P 500 and an even greater discount to Walmart based on enterprise value to EBITDA.
Alphabet (GOOGL) is another member of the Magnificent 7, but one that has traded at a meaningful discount to the broader market on a debt-adjusted basis since we initiated a position during the depths of COVID-19. We purchased additional shares in 2022 and again this April when, in our opinion, shares were meaningfully undervalued. On almost every metric, GOOGL is a more attractive business than the typical business we analyze.
The stock, which was the top contributor to our Strategy’s performance in the third quarter, rallied on another strong earnings report that provided further evidence that Alphabet is a winner — not a laggard — when it comes to AI, contrary to prevailing views in early 2024. For example, the market has been concerned that competing AI chatbots would cannibalize demand for search, yet Alphabet’s integration of AI-assist into search has stimulated incremental query demand. In addition, GOOGL’s standalone AI assistant, Gemini, was the most downloaded app on iOS in August, overtaking ChatGPT for the first time since the competitor launched in 2022. Given Google’s dominate share of the search engine market and a growing recognition of the company’s AI capabilities, investors are becoming comfortable that GOOGL’s moat remains wide.
In September, Federal Judge Amit Mehta ruled on the Google search monopoly case providing remedies that were less punitive than market expectations. While we had no way of knowing how Judge Mehta would rule, we believed a near worst-case scenario was already priced into the stock. Rather than be forced to divest assets or be banned from compensating for third-party distribution (i.e. paying Apple for default search placement), the judge ruled Alphabet would only be restricted from “exclusive” placement deals. While Alphabet must now share proprietary data with competitors, they will do so on normal commercial terms.
We expect capital spending to grow around $30 billion to approximately $82 billion in FY25 and R&D to rise around $5 billion to $56 billion. Rather than dilute profitability to fund $35 billion of additional investments, growth in operating profit (EBIT) should cover most of the added costs. As evidence, the company’s trailing twelve months per-share free cash flow rose 9% in the second quarter despite a 50% increase in capital investments.
Since we added to our position earlier this year, GOOGL’s valuation has increased noticeably. The stock now trades at 14.4X Enterprise Value/EBITDA, up from around 10X in April as the stock has returned more than 27% year to date.
Our worst performing holding in the quarter was Centene (CNC). Shares of the managed care company suffered from a policy-induced risk pool shift in the ACA Healthcare Exchanges that negatively impacted profit margins and caused a pre-announced earnings cut, nearly a month ahead of the scheduled earnings release date. Frankly, we were evaluating a position size reduction given the policy changes but were surprised by the announcement timing. Recognizing the short-tailed nature of the liabilities (allowing faster adjustments than other forms of insurance), we waited for management to formally articulate their plans to address the issue in late July. We came away encouraged by aggressive pricing actions meant to reflect recent developments and restore margins. In September, the company provided a positive update on the action plan that helped shares begin to recover from July’s sharp selloff. Meanwhile, political pressure has intensified from both sides of the aisle to address the looming health insurance affordability crisis. Any policy improvement would further supplement the actions taken by CNC, which assume no relief. The stock trades at 11x 2026 earnings estimates which look increasingly beatable based on the recent update.