The third quarter of 2025 was not unfamiliar, with several common themes in markets reemerging. Perhaps the most relevant for this shareholder letter is that active managers are, as a group, having an especially rough time of late. A chart published by Bloomberg citing Jefferies data shows that only 22% of active managers were outperforming their index through September of this year, which would represent the worst showing in decades if it holds through year end.
One of the longer-term causes of underperformance is basic math around competition — if active managers charge fees to ply their craft while passively managed peers charge almost no fees but get the benefit of efficiently priced securities, then active funds in the aggregate receive the market’s return less their higher fees, resulting in a lower net return for active fund shareholders than for passive.
This competition creates near-term incentives to not fall behind and to therefore “limit risk” by reducing the extent to which a strategy can deviate from its benchmark; it also creates an incentive to buy what’s been hot, and when those two combine, it can lead to extremes. We believe a lot of the current market concentration could be attributable to this dynamic. Indeed, a recent research piece from Empirical Research Partners notes that, “In the past six months the relative returns produced by a price momentum strategy have amounted to nearly +20 percentage points, making this one of the best runs in more than seven decades.”
In previous notes, we have warned that unprecedented index concentration and potentially stretched valuations among trending companies benefitting from artificial intelligence could lead to heartache among index adherents, a possibility that has not materialized. A financial news commentator recently flagged that Nvidia’s market cap of $4.5 trillion was higher than the market capitalizations of all but a few countries’ entire stock markets. Of course, this has no relevance to the intrinsic value of Nvidia, nor should this statistic have any bearing on whether Nvidia is likely to outperform or underperform the market in the future. However, here’s something that might: a study by Bain found that the revenues required to sustainably support projected AI capex investment would need to reach $2 trillion annually by 2030, which is more than the combined annual revenue of Microsoft, Alphabet, Meta, Nvidia, Apple and Amazon; the WSJ also notes it is more than five times the size of the entire global subscription software market. Bain still projects a stubborn $800 billion shortfall even if we assume all of the following: 1) 100% of all on-premise IT moves to the cloud, 2) AI reduces sales/marketing/customer support costs by 20%, and 3) R&D costs fall by 20% thanks to AI. Many are more optimistic than Bain, but few have yet to quantify or contextualize the aggregate economic demand offsets required to justify today’s investment frenzy.
Yes, AI will change the world in ways we probably cannot yet imagine. So too would airplanes, telephone lines, railroads, cars and the internet; all experienced a bust and consolidation on the heels of an investment rush. None of this changes our approach, which is to construct concentrated portfolios with attractively valued companies with minimal regard to the benchmark, which will lead to results that sometimes differ materially from our benchmarks, for both better and worse.
As always, we remain the largest investors in all of our public funds, and we appreciate your partnership and welcome any questions or comments.
Bill Miller IV, CFA CMT
Miller Value Partners
@billfour
Miller Income Fund
The Miller Income Fund (LMCLX) gained 4.67% during the third quarter, outperforming the ICE BofA High Yield Index’s 2.42% total return. (View current performance here). The fund exited its positions in Herbalife’s 12.25% 2029 corporate bonds, which approached our fair value estimate, and Easterly Government Properties (DEA), which revealed itself to be a less thoughtful capital steward than we had previously underwritten. The fund initiated a position in five new securities during the quarter. Cal-Maine Foods (CALM) is the country’s largest egg producer, which has benefitted from tight supply conditions following recent breakouts of Highly Pathogenic Avian Influenza (HPAI), trading at less than 4x FY25 EPS, a compelling entry point if conditions don’t normalize as dramatically as the market currently expects. Ituran Location & Control (ITRN) operates a near-monopoly stolen vehicle recovery service in Israel with expansion efforts underway in the larger South American market, with a highly aligned management team, robust returns on capital, and a nearly 6% dividend yield. Similar to last quarter, we initiated a position in Strategy’s 9% perpetual preferred equity, which yielded more than 10% as of quarter-end, a compelling return given the issue’s seniority to other preferred securities in the capital stack and significant asset coverage from Strategy’s ownership of 640K bitcoin as of 10/5/25, worth ~$79B. United Parcel Services (UPS), the world’s largest package delivery and logistics provider, offers a nearly 8% dividend yield as the stock trades at a forward (FY26) P/E multiple of 11.6x, representing a 47% discount to the S&P 500’s multiple, nearly the widest discount in the stock’s history. Lastly, Upbound Group (UPBD), formerly known as Rent-a-Center, trades at less than 5x FY26 EPS estimates and yields nearly 7%, despite potential deregulation tailwinds and a growing online business.
MVPA
Miller Value Partners Appreciation Fund (MVPA) returned 0.53% (market price) in the third quarter, underperforming the S&P 500’s 8.12% gain. (View current performance here). The fund initiated four new positions during the quarter, including the aforementioned UPS. Bumble (BMBL)’s founder, Whitney Wolfe Herd, recently returned as CEO to turnaround the company’s disappointing execution the past few years, with the stock trading at a nearly 20% trailing twelve-month (TTM) free cash flow (FCF) yield. Resideo Technologies (REZI) is a leading provider of energy usage and comfort technologies for residential and commercial properties, operating in a large addressable market with strong tailwinds for growth, yet trades at a forward (FY26) EV/EBITDA multiple of less than 11x. SharkNinja (SN) is a household appliance company with an intense focus on innovative products that solve consumer problems, allowing it to consistently gain market share against competitors in a growing $120B global addressable market, while also delivering peer-leading returns on capital, which are not reflected in a forward (FY26) EV/EBITDA multiple of 12x. The fund exited Centene (CNC) as ongoing Medicaid redeterminations and regulatory headwinds created too much uncertainty about the normalized earnings power of the company.
MVPL
Miller Value Partners Leverage Fund (MVPL) returned 14.64% (market price) in the third quarter, outperforming the S&P 500’s 8.12% gain.(View current performance here). The fund remained in a leveraged position throughout the entire quarter.