phongphan5922/iStock via Getty Images
The Great Divide: Why the Stock Market is So Concentrated and Where We Are Finding Opportunities
In the third quarter, the market continued and extended its recovery from the Liberation Day crash. Increasingly, the “market” as we have come to speak of it, has been dominated by a handful of mega cap companies. Consequently, the equal-weighted S&P 500 (NYSE: RSP (RSP)) is a far more accurate proxy for how the median stock is performing, and the divergence between RSP and SPY (SPY) shows just how distorted the cap-weighted “market” has become. This dynamic was highlighted in a recent Wall Street Journal column by Spencer Jakab titled “Stocks are Up–But Maybe Not Your Stocks.” This performance divide reflects the S&P 500’s extreme concentration within a small subset of the largest companies, as seen below:1
Magnificent Performance of equal and capitalization-weighted S&P 500 funds since 2023
Source: FactSet
The S&P 500 (SPY) is experiencing a historic level of concentration; the top 10 companies now account for over 40% of the index’s total weight:2
Exhibit 28: Concentration of S&P 500 market cap and earnings in the 10 largest index constituents
Source: Compustat, IBES, FactSet, Goldman Sachs Global Investment Research. Earnings reflect consensus forward 12-month estimates.
This level of concentration is unusual and makes historical comparisons of metrics, such as Price-to-Earnings (P/E) ratios, less relevant than they typically would be. This also creates performance challenges for active investment strategies, which often focus on smaller and mid-cap stocks and tend to underweight the market’s largest names.
Tensions in The Market: Speculation and Ai Focus
Alongside this concentration, we are observing two concurrent forces creating unique market tensions:
1. High Retail Participation and Speculation: Retail speculation is now strong enough to bend valuation logic as pockets of the market are trading at levels that have little to do with fundamentals and everything to do with narrative. Generally, stocks with faster revenue growth should justify a higher Enterprise Value to Sales ratio (EV/S). While most stocks cluster around a line of best fit, there are notable exceptions, such as Palantir (PLTR), whose high multiple is difficult to rationally justify.3
LTM) on the Y-axis versus Total Revenues CAGR (2Y TTM) on the X-axis. The plot shows a positive correlation, with a line of best fit (y = 0.2150x + 4.1079, R-squared = 0.1401). Most stocks cluster near the line, but Palantir (PLTR) is a significant outlier, showing a very high EV/Sales ratio (around 110) despite moderate revenue growth (around 40% CAGR).” contenteditable=”false” width=”640″ height=”311″ loading=”lazy” srcset=”https://static.seekingalpha.com/uploads/2025/12/11/542689-17654965732153776_origin.jpg?io=w640 640w,https://static.seekingalpha.com/uploads/2025/12/11/542689-17654965732153776_origin.jpg?io=w480 480w,https://static.seekingalpha.com/uploads/2025/12/11/542689-17654965732153776_origin.jpg?io=w320 320w,https://static.seekingalpha.com/uploads/2025/12/11/542689-17654965732153776_origin.jpg?io=w240 240w” sizes=”(max-width: 767px) calc(100vw – 36px), (max-width: 1023px) calc(100vw – 180px), 552px”>
2. AI Dominance: Artificial Intelligence (AI) is “sucking the air out of everything else.” Even outside the top 10 stocks, many market winners have a connection to AI. This is evident in the top 10 holdings of the Russell 2000 small-cap index, which include AI infrastructure and quantum computing stocks. Several of these smaller companies lack revenue, and those that do have sales boast some of the highest EV/S ratios in their sectors and the market at large.4
I Shares Russell 2000 ETF (IWM) – Holdings As of Sep 30 2025
Click to enlarge
We believe these forces create an incredible opportunity, even among the market darlings.
Healthcare remains one of the most mispriced sectors in the market, and the valuation gap continues to widen in investors’ favor. We have been vocal on this point for several commentaries and will share this chart from @lhamtil on Twitter to hammer home the point:5
Source: Ken French database; L. Hamtil’s calculations
Beyond healthcare, we do own three of the Mag7 names in size. Lumping the Mag7 together is analytically flawed as the underlying businesses diverge sharply in valuation, growth, cyclicality, profitability, and strategic positioning. Our largest holding is Alphabet (GOOGL) (formerly known as Google) and is in the Mag7, but jokingly has been called the “Lag7” at times.
The Day Google Became an AI Stock
We have owned shares in Google (now Alphabet) since the very first day of 2012. At the time, the perceived threat from the mobile app world caused its P/E ratio to fall to a historic low. The current AI saga feels like “deja vu all over again”, with Alphabet’s trailing-twelve-month (TTM) P/E setting a new all-time low, even below the 2012 level.
Since initiating our position early in the firm’s history, Google has become our largest holding, even as the journey has included its share of difficult stretches. No one stretch has been harder to hold through than this latest stretch, as the stock faced dual overhangs from both the AI threat and the anti-trust action brought by the US government. The anticipation, and eventual arrival, of regulatory action weighed heavily on Google, both competitively and in the market. Investors feared drastic outcomes, including the loss of Apple’s default search placement or even a forced separation of Chrome from the core business. This tension ended on September 2, 2025, when Judge Amit Mehta handed down his ruling on the remedies in United States v. Google LLC.6
While some constraints were imposed, the worst-case scenarios were removed. Internally, we declared this “The Day Google Became an AI stock,” given how Google could largely continue business as usual and more importantly, how the company could absolve key personnel of time committed to this case and redirect the entire focus of the organization to winning in AI.
The AI Evolution of Search
The notion that Google’s job is to distribute clicks misstates the company’s purpose. Users want answers, not blue links, and Google has been conceptualized and in some respects, architected, for an answer-first world since inception. The very first version of Google in 1998 had an “I’m feeling lucky” button, designed to take users directly to an answer:
From day one, the company was moving toward “zero-click searches,” where the search engine provides the answer directly.
Constraints imposed by regulatory actions, beginning with the acquisition of ITA Software (the backbone of travel sites like Expedia and Kayak), required Google to out-license data and avoid suppressing competitor links. Without these restrictions, Google could have used that data to deliver a fully integrated, best-in-class travel experience directly within search. More recently, Yelp’s lawsuit against Google underscores the ongoing tension between preserving a fair and open internet in its role as the dominant search engine and optimizing the platform to give users the most seamless, high-quality answers, quickly and directly.7
That all changed with the launch of OpenAI. Not only did Google face a new rival, but the company seemingly faced an existential threat, with the value of search itself in peril. This threat offered Google cover from a fast-moving competitor to build a product geared at answers first and the cover broadened as multiple competitors stepped into the fold. Eventually, not only did Gemini (Google’s AI Model) become a powerful standalone product, but Google was also able to weave AI-driven answers into search itself.
The world increasingly insisted that answers, rather than a series of links, was the goal and arbiter of “winning” in AI. Judge Mehta acknowledged this directly in his ruling on the remedies in the case, declaring that “The emergence of GenAI changed the course of this case” given how experts perceive it as a “nascent competitive threat.”8
The Vertical Integration Advantage
Google’s long-term investment in AI and machine learning is deep. The key breakthrough that led to the creation of OpenAI—the paper “Attention is All You Need”—was written by the Google Brain team.9 Investments in AI, such as Google Brain/DeepMind and Waymo, were often “masked” by the company’s “Other Bets” structure. However, the strategic decision to shift DeepMind’s reporting to Alphabet’s corporate costs in Q1 2023 reflected the central role of GenAI to the company’s entire corporate strategy.
Another key investment is the Tensor Processing Unit (TPU), Google’s proprietary chip designed specifically for AI. With the launch of the 7th generation, a meaningful portion of Google’s products now run on the TPU. This gives Google an advantage: while competitors must wait for Nvidia (NVDA) GPUs, Google can order its own chips at a substantially cheaper price.
Google now has a distinct and accelerating advantage in AI. It is the only fully vertically integrated AI company, excelling in all five critical areas:
- Semiconductors—Proprietary TPU chips.
- Hardware—Owns its datacenters, ensuring needed capacity for itself and capacity to sell to others.
- Models—Google’s Gemini appears to be a top-tier model for many critical functions.
- Consumer—Dominance in search and Android provides built-in customer relationships and awareness at no incremental marketing spend.
- Enterprise—Google Cloud Platform (GCP) has a large and growing customer base and can package models with pure computing power.
Competitors check at most three of these boxes. Google’s dominance in all areas makes it the low-cost provider for each one individually. This is particularly valuable in Enterprise solutions, where bundling computing power with models is becoming essential. We believe the market perception of Google is shifting from a slow-moving business facing disruption to a dominant and advantaged leader in the technology that will power the next decade.
What’s (New) in Your Wallet? Capital One (COF)
We recently added shares of Capital One (NYSE: COF) to our portfolios. At one point, through our shares in ING (ING) years ago, we owned COF indirectly. This resulted from ING’s divestiture of ING Direct to Capital One, in exchange for cash and shares totaling just shy of 10% of the company. This acquisition gave Capital One a low-cost, digitally native (branchless) deposit base with a state-of-the-art user interface and consumer experience. Though we never owned COF directly, we watched it admirably for years. When COF announced their Discover (DFS) acquisition, we thought it was too good to be true and feared regulators could prevent the tie-up from happening, but we did our work as we watched to see what would happen. As luck would have it, regulators eventually approved the deal and we now own COF directly.
The acquisition of DFS is a transformative transaction that uniquely positions COF to generate substantial value, fundamentally redefine its competitive standing, and reshape the US payments landscape. By integrating DFS’s proprietary network and customer base, COF is shifting from being primarily a card issuer to a vertically integrated payments entity, akin to an American Express-like operation with scale rivaling the largest payments networks in the world.
The most immediate and transformative value driver of the acquisition is the ownership of the Discover (DFS) network, which allows COF to avoid regulatory constraints that cap interchange fees for most large banks. This network capability is a “game-changer” that provides an immediate boost to fee revenues by repapering and issuing debit cards to all existing COF customers.
Specifically, the acquisition allows COF to leverage a key exemption related to the Durbin Amendment, which typically caps debit interchange fees, by issuing debit cards on its owned network and negotiating rates directly with merchants. Experts estimate that this switch will result in a pickup of approximately 75 basis points (BPS) in debit interchange rates, leading to an implied weighted average rate of 1.25% on a $50 transaction, compared to the existing Durbin-capped rate of 0.49%. This ability to monetize its massive debit volumes is projected to deliver over $1.0 billion in network synergies by 2027. COF will invest considerably in its merchant network, especially abroad, so that its customer value can truly rival what the Visa (V) and Mastercard (MA) networks offer their own debit customers. As it currently stands, COF is well prepared to do so.
We think it is less likely, though possible that COF eventually pursues this strategy on the credit side of things. Even if they do not go all-in, as they are in debit, there are levers the company can pull to capture greater economics in the credit card network. As it stands, this acquisition elevates COF’s scale, making the combined entity the largest credit card issuer in the country and the second largest by transaction volume. Even if COF keeps the credit card setup as is, owning a network and added degrees of scale provides real leverage in future renewal negotiations with Visa and Mastercard.
Central to realizing this value is Capital One’s sophisticated technological infrastructure and operational expertise. COF has historically made substantial investments in its tech stack, designed specifically to grow through acquisition and convert new portfolios efficiently onto its platform. This technological playbook is crucial for integrating Discover’s systems, especially on the loan and asset management side, where integration is a core competency for COF. This integration capacity drives confidence in achieving significant operational benefits. COF has guided toward realizing approximately $1.5 billion in annual expense savings within two years, representing about 23% of Discover’s operating expenses.
Redefining Capital One and Enhancing Its Banking Franchise
The merger facilitates COF’s redefinition by broadening its consumer base and strengthening its banking foundation. Discover’s portfolio includes prime and student credit customers, complementing COF’s traditional strength in the “main street, new to credit” and subprime markets. This shift improves the pro forma loan book mix by directing it toward less consumer subprime risk. The combination with Discover’s lower-loss loan book is expected to act as a dampener in future regulatory stress tests. Furthermore, the consolidation eliminates a key top competitor in the main street credit card segment, reducing new customer acquisition costs.
On the banking front, the combined institution would rise from 9th to 6th among US banks by deposits. Discover’s own historical strength in online savings combines nicely with COF’s online bank that emerged from the ING Direct acquisition over a decade ago.
In preparation for the merger, both COF and DFS accumulated large excess capital balances, resulting in substantial capacity for capital returns to shareholders. To that end, COF recently increased its dividend by 33% and announced a new $16 billion share repurchase program. We expect at least $10b in repurchase activity (or over 7.5% of market cap at today’s valuation) each of the next two years, with capacity to either continue repurchases at an aggressive rate or to grow the dividend, depending on how cheap or expensive shares trade.
The translation of these strategic advantages and capital return into shareholder value makes COF an attractive investment opportunity today. The DFS acquisition is more than a simple consolidation; it is a strategic metamorphosis. Taken together, this strategic vertical integration transforms COF into a major payments player and the company is now one of the most important domestic entities in this key space within the financial sector.
The traditional driver of value for banks is tethered to the company’s balance sheet and its ability to profitably underwrite loans. Banks are heavily regulated and as such, it is challenging to earn meaningfully in excess of bank cost of capital, effectively constraining the ability to earn very high ROEs in the space. Some banks, including COF have excelled at earning ROEs consistently above their cost of capital, but few have earned ROEs above 20%. A thriving payments franchise with its return profile driven by network effects, technology and intangible assets that do not exist on a balance sheet can earn ROEs well above 20% and now Capital One owns a scaled, rapidly growing piece of this market.
This convergence of network economics and operational efficiency acts like a flywheel, spinning off immense cash flow that can be continuously recycled to investors through accelerated capital returns, while growth in net income accelerates. We expect the combined entity to earn $25 or more in EPS come 2027, versus 2024’s $14 in EPS. Most importantly, we expect the bank’s Return on Tangible Common Equity (ROTCE) to exceed 20% by 2027, justifying a meaningfully higher multiple in the market than today’s 12x trailing and 10x forward P/E.10 Look at a group of comparable P/Es on this year’s earnings:
Click to enlarge
The average peer trades right around a 15x P/E. The ROE advantaged group that have earnings streams untethered to the balance sheet trade at nearly 17.5x. At $25 in 2027 EPS, the stock is simply mispriced: even a pedestrian 15x multiple implies ~70% upside, and a hybrid bank / payment-network multiple implies a double.
Thank you for continuing to place your confidence in us. If any of the perspectives shared here prompt questions or lead you to rethink aspects of your portfolio, don’t hesitate to reach out. You can contact either of us at 516-665-1945 or via the direct lines listed below. Markets like this separate real discipline from passive drift, and today active management matters more than it has in quite some time. We’re excited at the opportunities in front of us.
Jason Gilbert, CPA/PFS, CFF, CGMA, Managing Partner, President
Elliot Turner, CFA, Managing Partner, CIO
Footnotes
- https://www.wsj.com/livecoverage/stock-market-today-dow-sp-500-nasdaq-10-30-2025/
- https://x.com/GlobalMktObserv/status/1982854891010715991
- Comprehensive financial data analysis – Koyfin
- iShares Russell 2000 ETF | IWM | US Class
- https://x.com/lhamtil/status/1971259035887526067?t=XS6vKz6HKkXIj4wdxONgjw&s=19
- https://www.cnbc.com/2025/09/02/google-antitrust-search-ruling.html
- Yelp sues Google, alleging a search engine monopoly that promotes its own reviews | CNN Business
- https://cases.justia.com/federal/district-courts/district-of-columbia/dcdce/1:2020cv03010/223205/1436/0.pdf?ts=1756893273
- https://research.google/pubs/attention-is-all-you-need/ and How Google gave the key breakthrough technology for ChatGPT to OpenAI
- Pinegap – AI Platform for Equity Research
Original Post
Editor’s Note: The summary bullets for this article were chosen by Seeking Alpha editors.
