Q2 2026 was a strong quarter for the Modern Value Investing Portfolio, which returned 35%. After a weak Q1 (-20%), the portfolio is now up approximately 7.5% for the first six months of 2026. That still trails my self-chosen benchmark, the MSCI World, which is up 12.5% year-to-date.The strong Q2 performance did not come entirely as a surprise. In my 2026 Q1 Portfolio and Performance review I wrote:Modern Value Investing is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.After updating my models, I am extremely optimistic about future return expectations. In full transparency, I believe the current setup is among the most attractive over the last 10 years, more attractive than at the bottom of last April’s tech selloff, and perhaps even comparable to the bottom of the Covid selloff in March 2020. That does not mean stocks cannot go lower from here, and I will certainly be wrong on some positions. But for the positions where I am right, I believe the upside is enormous at current valuations.Some of my largest positions saw significant gains in Q2. Position sizes going into Q2 are shown in brackets:SK Hynix (11% position): +196%Oscar Health (15% position): +181%Root Insurance (5% position): +39%Novo Nordisk (9% position): +37%WeightWatchers (11% position): +19%During Q1, I used the weakness in many of these stocks to add meaningfully. In Q2, I did the opposite. I trimmed several of the biggest winners and used the proceeds to diversify into new opportunities.Here are my 20 positions going into 2026 Q3:The portfolio now holds 20 positions going into Q3 2026, up from 10 positions at the start of Q2. This is probably the most diversified the Modern Value Investing Portfolio has been in the last decade.Part of this shift reflects simple valuation discipline. After their enormous moves, SK Hynix and Oscar Health are mathematically no longer as attractive as they were three months ago. They may still be attractive, but the expected return profile has changed.At the same time, I have identified a growing number of new opportunities. In Q2 alone, I wrote up Booking Holdings, Nvidia, and Charter Communications. Several additional positions were initiated during the quarter, some of which I plan to write about over the coming months.In other words, the portfolio has moved from a smaller number of exceptionally mispriced opportunities to a broader set of highly attractive risk/reward situations. Today, I would classify my top four positions — Shift4, WeightWatchers, Bumble, and Charter — as the most outstanding opportunities in the portfolio.The move toward diversification is also a response to a more uncertain economic environment, especially because of the likely impact of AI. I will write more about this in a future post, but it is already shaping how I think about portfolio construction.At a high level, one part of the portfolio is positioned to benefit from the buildout of the AI economy. Another part is invested in businesses that I believe are unlikely to be structurally disrupted by AI, and in some cases may even benefit indirectly from it.Let’s begin with the AI buildout part of the portfolio.AI Chips — 9%SK Hynix — 5%Nvidia — 4%These are direct beneficiaries of AI infrastructure spending. Nvidia remains central to accelerated computing, while SK Hynix benefits from the rapid growth in high-bandwidth memory, one of the most important bottlenecks in AI training and inference.AI Datacenters / Cloud Computing — 15%Microsoft — 4%Amazon — 4%Alphabet — 4%Nebius — 3%Cloud platforms should benefit from rising demand for AI compute, storage, model deployment, and enterprise AI applications. Microsoft, Amazon, and Alphabet have the scale, customer relationships, and capital to capture much of this demand, while Nebius offers a more focused and higher-upside infrastructure exposure.Chinese AI Companies — 7%Tencent / Prosus — 3%Alibaba — 3%Tencent and Alibaba are among the most important technology platforms in China. Both have large user ecosystems, cloud or AI capabilities, and the potential to embed AI into advertising, gaming, commerce, enterprise services, and productivity tools.Digital Infrastructure — 10%Charter — 6%Meta — 4%AI will not only require chips and datacenters. It will also require connectivity, distribution, attention, and digital infrastructure. Charter is a less obvious but important beneficiary through broadband demand and data usage, while Meta combines massive distribution with one of the strongest AI research and advertising platforms in the world. Meta may also gain a new monetization path if recent reports prove correct that the company is exploring ways to rent out excess AI compute capacity through a cloud infrastructure offering.In total, about 41% of the portfolio is exposed to companies that I believe can benefit directly or indirectly from the AI buildout.The other major part of the portfolio is invested in companies that I believe are unlikely to be structurally disrupted by AI. Importantly, many of them are not “ex-growth” defensive holdings. They are exposed to their own powerful megatrends, including GLP-1 adoption, digital payments, online travel, housing, and health care. In other words, the portfolio is not only an AI buildout portfolio. It is a collection of businesses that I believe can compound value across several independent long-term themes.GLP-1 / Weight Health — 14%WeightWatchers — 8%Novo Nordisk — 6%Another 14% of the portfolio is exposed to the GLP-1 and weight-health theme.I have written about both Novo Nordisk and WeightWatchers before, so I will keep this brief here. I believe the GLP-1 supercycle is still in its early innings, and the arrival of oral GLP-1 pills could further expand adoption by reducing the friction of injectable treatment.Novo Nordisk remains one of the key pharmaceutical beneficiaries of this shift, while WeightWatchers offers a more leveraged consumer-platform angle: medical access, behavioral support, nutrition, community, and long-term weight maintenance around GLP-1 adoption.Insurance / Insurtech — 11%Root — 5%Oscar Health — 5%Another 11% of the portfolio is exposed to insurance and insurtech through Root and Oscar Health.I view both companies as unlikely to be structurally disrupted by AI. Insurance remains a regulated, balance-sheet-driven, data-intensive business where underwriting discipline, pricing accuracy, claims management, and distribution matter enormously.At the same time, both Root and Oscar should be in a good position to leverage AI because they were built with digital-first DNA. Root can apply AI and data science to pricing, underwriting, telematics, claims, and customer acquisition. Oscar can use AI to improve member navigation, care routing, administrative efficiency, claims processing, and the overall consumer experience in health insurance.In other words, these are not traditional insurers trying to retrofit technology onto legacy systems. They are technology-enabled insurers that may be able to use AI to widen their efficiency and underwriting advantages over time.Travel & Hospitality Infrastructure — 18%Shift4 — 14%Booking Holdings — 4%Another 18% of the portfolio is exposed to travel and hospitality infrastructure through Shift4 and Booking Holdings.I view both companies as unlikely to be structurally disrupted by AI. People will continue to travel, eat out, stay in hotels, attend events, and spend money on experiences. AI may change parts of the discovery and booking journey, but I do not believe it eliminates the need for scaled distribution, trusted platforms, payment infrastructure, and merchant relationships.Booking Holdings is the obvious travel platform in this bucket, with enormous scale, brand strength, and global demand aggregation. Shift4 is a payments and commerce infrastructure company with strong exposure to restaurants, hotels, stadiums, entertainment venues, and travel-related merchants.Both companies may also benefit indirectly from AI. Booking can use AI to improve search, personalization, customer service, and conversion. Shift4 can use technology and data to improve merchant acquisition, fraud prevention, payment optimization, and software-led monetization.This bucket gives the portfolio exposure to the long-term growth of travel, hospitality, experiences, and commerce infrastructure.Online Dating / Relationships — 8%Bumble — 8%Another 8% of the portfolio is invested in Bumble, which I last wrote about following its Q1 2026 earnings report.Online dating is another category that I believe is unlikely to disappear because of AI. People will continue to seek relationships, dates, friendships, and social connection. However, AI could meaningfully change how these platforms work — from profile creation and matching to conversation support, safety, moderation, and personalized coaching.Bumble has clearly struggled, but this is also why the opportunity exists. If management can use AI to improve the user experience, reduce friction, and reignite growth, I believe the current valuation leaves significant upside.Enterprise Software / AI Transition — 5%Salesforce — 5%Salesforce is a recently added 5% position and likely the subject of a future write-up, so I will not say too much here, except that on the surface, Salesforce looks like a potential AI loser, but I think the story may be more nuanced. Digital Banking / Consumer Finance — 3%SoFi — 3%Another 3% of the portfolio is invested in SoFi.SoFi is my exposure to the continued shift from legacy banking to digital-first consumer finance. The company has built a broad financial platform across lending, deposits, investing, credit cards, and financial services infrastructure. If SoFi continues to grow members, deepen relationships, and expand profitability, I believe it can become one of the long-term winners in online banking.Consumer Health / Functional Beverages — 3%Celsius Holdings — 3%Celsius is a smaller consumer-health position and fits loosely with the broader health and weight-management theme in the portfolio.Energy drinks remain a growing category, and Celsius has positioned itself around fitness, energy, and often zero-calorie consumption. That makes it an interesting adjacent beneficiary of a world in which consumers are increasingly focused on weight, health, protein intake, exercise, and metabolic wellness. ConclusionOverall, I believe the portfolio is now better balanced than it was three months ago. I have reduced exposure to some of the biggest Q2 winners, diversified into a broader set of attractive opportunities, and maintained meaningful concentration in the ideas where I see the best risk/reward. The portfolio is no longer just a bet on a handful of deeply mispriced stocks. It is now built around several independent long-term themes — AI infrastructure, GLP-1 adoption, digital insurance, travel and hospitality infrastructure, online relationships, digital banking, and consumer health. I will certainly be wrong on some of these positions, but I believe the current mix gives the portfolio a strong combination of upside, diversification, and exposure to some of the most important structural shifts in the economy.Disclaimer: I share my own thinking and positioning here for transparency and informational purposes. Nothing in this article should be taken as investment advice. Please do your own research and make your own decisions based on your individual circumstances.Modern Value Investing is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.