Category: Investment Portfolio

Concentrated investing is not the same as careless investing. It is a decision to let a small number of ideas have a meaningful effect on your result, while accepting that being wrong will hurt more.

I run a concentrated main portfolio. That is a personal choice, not a template for anyone else. I am trying to own fewer businesses that I understand deeply rather than a larger list that gives the appearance of safety without real conviction.

What concentrated investing means to me

A concentrated portfolio is not defined by a magic number of holdings. It is defined by whether the biggest positions can materially change the outcome. In my case, Rocket Lab became a position large enough that it could move the main ISA meaningfully in either direction.

I wrote about why I made Rocket Lab roughly 50% of my portfolio because a decision that large should be documented before its outcome is obvious.

The case for fewer, higher-conviction positions

The potential benefit is focus. With fewer businesses to follow, I can spend more time understanding the thesis, the risks and the evidence that would change my mind. That does not guarantee good outcomes. It just makes the decision process more visible.

There is also a difference between diversification and simply owning more tickers. If I cannot explain why I own something, what would invalidate the thesis and what role it plays in the portfolio, the extra holding may not be reducing the risk that matters.

The risks are real

Concentration magnifies mistakes. A missed execution target, financing need, valuation reset or wider market drawdown can hurt far more when a position is large. It can also make ordinary volatility emotionally difficult to sit through.

That is why I do not describe concentration as safe. The point is not to pretend the risk away. The point is to decide whether the potential upside is worth the risk, then keep testing that decision honestly.

My checks before adding to a position

• Can I explain the business and the core thesis in plain English?
• What would make me reduce or sell the position?
• Am I adding because the thesis improved, or because I want a falling price to stop hurting?
• Would I still be comfortable owning it if the share price fell further?
• Is the position size consistent with the risk I am actually taking?

Document the decision before the result

The most useful discipline is recording why I made a decision before I know whether it works. That creates a more honest record than rewriting a story after a gain, or pretending a loss was unforeseeable.

My monthly portfolio reviews are part of that process. They show what changed, where conviction increased and where I could be wrong. The live Portfolio page shows the current IBKR ISA snapshot separately from the Trading 212 challenge.

This is a personal investing framework, not financial advice. A concentrated portfolio can create large losses as well as large gains.

July was a month of making the portfolio more honest about where my conviction actually sits.

This review uses my IBKR ISA only. It is the account represented by the main Portfolio page. My Trading 212 holdings are part of a separate £15k High Growth Challenge and are not included in the figures or commentary below.

The figures are from my IBKR prior-business-day-close snapshot for 31 July 2026. They are a dated record, not live prices or a recommendation for anyone to copy my allocation.

Portfolio snapshot

At the 31 July close, the IBKR ISA was worth £194,538.84, including £49.84 of cash. The invested value was £194,489.00.

The separate Trading 212 challenge was worth £12,132.13 at its 31 July snapshot, including £136.74 cash. I am keeping it separate because the two accounts use different reporting timings and serve different purposes.

The main portfolio remains deliberately growth-oriented and concentrated. That is not a claim that concentration is safe. It is an acknowledgement that my returns, good or bad, will be shaped heavily by a relatively small number of positions.

Biggest changes in July

The biggest decision was increasing my Rocket Lab position from 1,000 to 2,000 shares. To fund that change, I sold Apple, Alphabet, NVIDIA and SoFi, and reduced Uber in the IBKR ISA. I wrote separately about why I made Rocket Lab roughly 50% of my portfolio, including the concentration, financing and execution risks I am accepting.

I also increased Micron from eight to 10 shares and Surf Air Mobility from 10,000 to 13,800 shares. These are material changes to a concentrated portfolio, so I want them recorded rather than quietly absorbed into an evergreen holdings page.

No quantity changes were identified in the Trading 212 challenge between its late-July reporting snapshots. Its three holdings at month-end were Uber, Amkor and AST SpaceMobile.

Current main holdings at month-end

At the 31 July IBKR close, the core positions included:

• Rocket Lab: 2,000 shares at $64.95.
• Palantir: 250 shares at $123.06.
• AbCellera: 3,001 shares at $5.71.
• BETA Technologies: 1,000 shares at $18.93.
• ServiceNow: 150 shares at $111.23.
• Surf Air Mobility: 13,800 shares at $0.76.

Those prices are the values in the dated broker snapshot. The live Portfolio page is the better place for the latest IBKR ISA holdings and indicative GBP values.

What I am watching

Rocket Lab remains the central position to watch, partly because of its size and partly because the investment case depends on execution continuing across several moving parts. The opportunity is why I own it. The concentration is why I need to keep looking for evidence that could challenge my view.

I am also watching whether the smaller positions continue to earn their place. Owning a company is not a permanent vote of confidence. I need to be able to explain what I am underwriting, what has changed and what would cause me to reassess.

For AbCellera in particular, the clinical proof points remain important. My latest AbCellera investment update sets out why the next programme milestones matter to the thesis.

Where I could be wrong

The obvious risk is concentration. A few disappointing outcomes at the same time would hurt this portfolio materially. That is especially true when growth companies can be volatile, funding needs can change and expectations move faster than the underlying businesses.

Another risk is confusing familiarity with understanding. Spending time researching a company can improve a thesis, but it can also make a view feel more comfortable than it deserves. The job is to keep testing the assumptions, not just to keep repeating them.

Finally, a positive month-end snapshot does not settle anything. I want the monthly record to show the decisions, the uncertainty and the areas where I could be wrong, rather than turn every portfolio move into a victory lap.

Final thoughts

July made the main portfolio simpler in one sense and riskier in another. I have fewer positions that matter, and more exposure to the ideas I believe in most. The trade-off is that I need to be more disciplined about reassessment when the screen is red as well as when it is green.

This is a personal portfolio record, not financial advice. The positions, prices and weights can change after the reporting date.

This is my current investment thesis, not financial advice. I own Uber in my main portfolio and the position can change.

My Uber thesis is increasingly less about whether ride-hailing works. That question has largely been answered. The more interesting question is whether Uber can turn its global marketplace into a durable free-cash-flow compounder while retaining a valuable position as autonomous fleets enter the market.

I see those as two connected parts of the same idea. The existing marketplace creates the demand, consumer relationships, driver and courier network, merchant base and dispatch capability. If Uber keeps improving the economics of that system, the cash flow can compound. If autonomous vehicles scale, Uber may be able to provide the marketplace layer without having to carry the full capital burden of building cars.

The base case is now cash generation

Uber’s latest quarterly filing shows why I think the business deserves to be viewed differently from its earlier growth-at-all-costs years. In Q1 2026, Uber generated $13.2 billion of revenue, $1.92 billion of operating income and $2.29 billion of free cash flow.

That free-cash-flow number matters more to me than a headline net-income figure in any one quarter. Net income can move with the valuation of investments and other non-operating items. The underlying marketplace is what I want to understand: how many people use it, how often they use it, whether both Mobility and Delivery keep growing, and how much incremental cash the platform can produce.

Uber reported 199 million monthly active platform consumers and 3.64 billion trips in Q1 2026. Gross bookings reached $53.7 billion, with Mobility and Delivery gross bookings up 20% and 23% respectively on a constant-currency basis. That is evidence that the marketplace is still expanding while the financial model becomes more disciplined.

Why the platform can compound

Uber has an advantage that is easy to take for granted because the app is familiar. It already sits between consumers, drivers, couriers and merchants in many cities. More activity can improve matching, availability and convenience. More services can give customers another reason to remain inside the same ecosystem.

That does not make the model immune to competition. It does mean there is a credible route for rides, delivery, subscriptions, advertising and business services to reinforce each other. The investment case is not based on one new product. It is based on a large, increasingly useful local-commerce network becoming more productive over time.

Autonomy is the upside option

Autonomous driving is the part of the story that could change the long-term economics, but I do not need it to justify the current holding. My base case is the existing marketplace and cash flow. Autonomy is upside if Uber becomes the place where riders discover, book and manage autonomous trips.

That role could be valuable because operating an autonomous fleet is not just a vehicle problem. It is also a demand problem, a dispatch problem, a customer-service problem and a local-market problem. Uber already has capabilities in those areas. Partnerships could let it benefit from autonomous supply without taking on the full manufacturing, sensor and fleet-financing risk itself.

The bull case is not that Uber has to build the winning autonomous car. It is that the winning fleets still need a large, trusted marketplace to find riders and keep vehicles productive.

What could challenge the thesis

• Regulation and worker classification: changes to the status or cost of drivers and couriers can materially affect marketplace economics.
• Competitive incentives: ride-hailing and delivery are competitive markets, and price competition can pressure margins.
• Autonomy economics: fleet partners may capture more of the value than expected, or autonomous deployment may take longer than the market expects.
• Consumer demand: travel, delivery and local-commerce spending are not immune to economic weakness.
• Execution across services: a broader platform only helps if it improves retention and economics rather than adding complexity.

My current view

I own UBER because it is becoming a stronger cash-generative marketplace while keeping an interesting option on autonomous mobility. The current business does not need a robotic future to work. But if autonomous fleets arrive at meaningful scale, Uber’s demand and dispatch layer could become more valuable, not less.

I will watch free cash flow, marketplace engagement, Mobility and Delivery growth, competitive incentives and the structure of autonomous partnerships. The key is to keep the base case and the optionality separate. A good existing business should not need a speculative future to make the investment work.

For the wider context, see my current portfolio. The figures cited above come from Uber’s Q1 2026 Form 10-Q.

This is a portfolio-decision update, not a recommendation to copy my allocation. I have previously written about why I find Rocket Lab interesting as an investment. This post explains why I have now made it my largest position.

I have sold a number of smaller positions and bought another 1,000 Rocket Lab shares. That takes my holding to 2,000 shares, worth roughly £100,000 at the time of writing and representing about 50% of my portfolio.

This is not a routine rebalance. It is the clearest expression of my current conviction, and it comes with a level of concentration risk that I need to acknowledge plainly.

Why I acted during the sell-off

Rocket Lab shares have fallen sharply. I cannot know precisely what is driving any individual day’s move. One material issue investors are assessing is the proposed Iridium transaction, the financing required to complete it and the potential dilution for existing Rocket Lab shareholders. Those are legitimate concerns, not noise to dismiss because I like the company.

The June 2026 merger agreement gives Iridium shareholders $27.00 in cash plus Rocket Lab stock. It also contemplates a 364-day senior secured bridge facility of up to $3.6 billion. A transaction of that size changes the financial conversation around Rocket Lab, and the market is right to scrutinise it.

My decision was not based on the view that these concerns do not matter. It was based on the view that the sell-off gave me a better entry point into a business whose long-term operating opportunity still looks unusually strong.

What I am underwriting

I see Rocket Lab as more than a launch company. Electron launch, spacecraft systems, components, national-security work and the development of Neutron create several ways for the company to compound if execution remains strong.

The Space Systems business is a major part of the thesis

The part of the business I think can be underestimated is Space Systems. Rocket Lab does not only build hardware for its own missions. Its filings describe a business that designs and manufactures many of the components and subsystems used in its launch vehicles and spacecraft, while also selling spacecraft components into the wider merchant market.

That matters because it creates revenue opportunities beyond Rocket Lab's own launch cadence. Solar cells, composite structures, separation systems, flight software, reaction wheels, spacecraft buses and other mission-critical components can be supplied to other satellite and space manufacturers. The long-term opportunity is not simply to launch more rockets. It is to participate in more of the value chain when other companies build spacecraft too.

I also like the acquisition strategy behind that ambition. Rocket Lab has been buying capabilities that can deepen vertical integration and solve difficult supply constraints. The completed Mynaric acquisition added laser optical communications technology and strengthened the company’s position in launch services, spacecraft manufacturing and satellite components. The proposed Motiv acquisition would bring solar-array drive assemblies, precision mechanisms and robotics in house, which the company describes as addressing a critical gap in its vertical-integration strategy.

That does not mean every acquisition will work. Integration, capital allocation and execution risk are real, particularly while Rocket Lab is also developing Neutron. But I see a deliberate pattern: buy strategic products and capabilities, reduce dependence on constrained suppliers, use them inside Rocket Lab's own missions and retain the ability to sell into the broader space market.

The proposed Iridium transaction takes that strategy another step. It would combine Rocket Lab's launch and spacecraft manufacturing capability with a global satellite communications network and spectrum assets. The financing and dilution questions are therefore not a side issue. They are the price and risk of trying to build a more fully integrated space company.

The most important point is that execution continues to produce evidence. In July, Rocket Lab announced full mission success for VICTUS HAZE, an end-to-end responsive-space mission for the U.S. Space Force. It also announced a full-duration hot-fire test of the Archimedes Vacuum engine, a meaningful development on the path toward Neutron.

Neither milestone removes the risk around Neutron, future funding or the Iridium deal. They do reinforce why I remain interested. The company is trying to become an integrated space business with launch, spacecraft and services capability, rather than a one-dimensional bet on a single rocket.

Why I sold smaller positions

I did not make this change because diversification is bad. Diversification is often sensible, especially when conviction is low or an investor cannot follow every holding closely enough.

For me, the question became whether several smaller positions offered a better prospective return than putting that capital into the company I understand best and feel most committed to following. My answer, at this point in time, was no.

That is a personal judgement, not a universal rule. A concentrated portfolio can be emotionally difficult and financially punishing when the thesis is wrong. I am not presenting this as a safe allocation. I am presenting it as my most honest expression of conviction.

The uncomfortable part of a 50% position

A 50% position changes the emotional reality of investing. A difficult week in Rocket Lab will have a meaningful effect on my portfolio. I will not be able to pretend that volatility is irrelevant, and I should not allow price movement alone to dictate the next decision.

The responsibility is to follow the facts, not defend the purchase. If operational execution weakens, Neutron slips materially, the economics or financing of the Iridium transaction deteriorate, or the overall investment case changes, I need to reassess without anchoring to the size of the position.

What would strengthen or weaken my conviction

• Strengthen: continued Electron reliability, further national-security execution, tangible Neutron progress and disciplined financing of the Iridium transaction.
• Weaken: material Neutron delays, operating setbacks, a financing outcome that damages the long-term economics for existing shareholders, or evidence that the acquired assets cannot be integrated as expected.

My view now

I have chosen to be concentrated because I believe Rocket Lab has one of the strongest long-term opportunities in the public space market. That does not make it a certainty. It makes it the position I am most willing to study, hold accountable and accept real volatility in.

The next job is not to celebrate the size of the position. It is to keep testing the reasons I bought it.

Sources: Rocket Lab 2025 Form 10-K; Rocket Lab Q1 2026 Form 10-Q; Rocket Lab’s 29 June 2026 Form 8-K on the proposed Iridium transaction; Rocket Lab’s VICTUS HAZE mission update; Rocket Lab’s Archimedes Vacuum engine test update.

This is personal investing commentary, not financial advice. I own Rocket Lab shares and may buy or sell without updating this post immediately.

View my Portfolio · Read my original Rocket Lab thesis · Browse all RKLB writing

This is an update to my original 2025 AbCellera investment thesis. I am keeping that piece as a record of the original case. This update is about what has changed and what I am watching now.

My ABCL thesis is becoming more specific. The question is no longer only whether AbCellera has an interesting AI-powered antibody-discovery platform. The more important question is whether the company can convert that platform into clinical evidence that changes how investors value the business.

That makes the next set of milestones meaningful. AbCellera expects topline Phase 2 data for ABCL635 in the third quarter of 2026, Phase 1 data for ABCL575 in the fourth quarter, and an IND or CTA submission for ABCL688 in 2027. None of those outcomes is guaranteed. But they give the investment case real events against which it can be tested.

The platform now has to earn clinical credibility

The original attraction of AbCellera was the possibility that its discovery platform, data and integrated capabilities could improve the route from biological target to antibody candidate. That remains the foundation of the story. Platform claims are easier to make than they are to prove, though.

Clinical progress is where the abstraction becomes more concrete. ABCL635 is in a Phase 1/2 study, with the company expecting topline Phase 2 data in the third quarter of 2026. AbCellera has said the Phase 1 portion supported moving into Phase 2 after doses were well tolerated, with no observed liver toxicity or serious adverse events in the interim assessment. That is not proof of commercial success, but it is a more useful signal than a slide deck about platform potential.

ABCL575 is expected to produce Phase 1 topline data in the fourth quarter of 2026. Management has also stated that it currently has no plans to develop that programme beyond Phase 1. I read that as a reminder that not every programme has to become a blockbuster for the platform to create value. It also means I should not give ABCL575 more weight in the thesis than the evidence justifies.

There is more than one shot on goal

What still appeals to me is that AbCellera is not purely a single-asset biotech story. The company has partner programmes and royalty interests alongside its own internal pipeline. ABCL688, an autoimmune candidate from its GPCR and ion-channel platform, is expected to move toward an IND or CTA submission in 2027.

That does not eliminate risk. Partner activity, milestone revenue and royalty economics can be difficult to forecast. The value of multiple shots on goal is not that every one succeeds. It is that the company has several routes through which its discovery and development capabilities could create value over time.

The balance sheet buys time, not certainty

At 31 March 2026, AbCellera reported $504.7 million in cash, cash equivalents and marketable securities. That is a meaningful resource for a company investing in internal programmes and infrastructure. It gives management time to pursue the upcoming clinical milestones without making the next quarter the only thing that matters.

It should not be treated as a free pass. Cash, equivalents and marketable securities were $533.8 million at the end of 2025, and AbCellera remained loss-making in the first quarter. Revenue increased to $8.3 million from $4.2 million in the prior-year quarter, but I view that cautiously because the company’s revenue can be milestone-driven and uneven.

What would strengthen my conviction

• ABCL635 produces credible Phase 2 data and the data supports a clear development path.
• ABCL575 delivers a clean Phase 1 readout, even if the strategic value is primarily platform validation.
• ABCL688 progresses toward its planned 2027 filing.
• The company demonstrates that partner activity and internal programmes can coexist without the platform becoming an expensive science project.

Where I could be wrong

• Early clinical signals may not translate into meaningful efficacy or commercial value.
• Clinical timelines can slip, data can disappoint and development costs can rise.
• AbCellera has no marketed proprietary drugs, so the internal-pipeline case remains unproven.
• Partner fees and milestones can make revenue volatile and difficult to model.

My view now

I still see ABCL as a high-risk, high-uncertainty holding. The reason I remain interested is not a claim that the platform has already won. It is that the next twelve to eighteen months should produce clearer evidence about whether AbCellera can turn an interesting discovery engine into a clinically validated and commercially relevant business.

That is the test. If the clinical milestones strengthen the case, the platform narrative becomes more credible. If they do not, I need to be willing to reassess the thesis rather than defend the original idea.

Sources: AbCellera Q1 2026 Form 10-Q; my original 2025 ABCL investment thesis.

This is my personal investing commentary, not financial advice. I may own shares in AbCellera and can change my view as new information emerges.

View my Portfolio · Browse all ABCL writing

This is my current investment thesis, not financial advice. I own BETA Technologies in my main portfolio and the position can change.

BETA Technologies is not a conventional aerospace investment. It is still early in commercialisation, still loss-making and still dependent on certification and execution. That is exactly why the opportunity is interesting to me. The company is trying to turn electric aviation from a promising concept into an operating system of aircraft, charging and manufacturing.

The part of the thesis I find most compelling is not simply vertical take-off aircraft. It is BETA’s ability to build relationships around the conventional take-off ALIA variant, where payload and practical operations may offer a clearer route into early commercial use. The charging ecosystem is equally important. Aircraft without dependable infrastructure are much harder to deploy at scale.

A more practical route into electric aviation

BETA is developing the ALIA in both conventional take-off and landing, or CTOL, and vertical take-off and landing, or VTOL, configurations. The conventional variant matters to me because it can address real logistics and regional-aviation use cases without asking operators to solve every part of the VTOL challenge on day one.

A useful payload and straightforward airport operations can make the first customer conversations more concrete. Rather than selling a distant vision of urban air mobility, BETA can work with operators that have existing routes, maintenance needs and infrastructure decisions to make now.

BETA’s own product site states a demonstrated range of 336 nautical miles, cargo capacity of 200 cubic feet and charge times of under one hour. These are company-reported figures, not guarantees of commercial economics. They do, however, show why the ALIA has a different feel from a purely conceptual eVTOL programme.

The charging ecosystem is part of the product

The aircraft is the visible part of the story. The more important question may be whether BETA can help customers operate a fleet. Its Charge Cube infrastructure, manufacturing capability and focus on deployment create the possibility of a more integrated proposition.

That matters because early electric aviation cannot rely on a fully built public network appearing by itself. Operators need aircraft, charging, maintenance, locations and a credible pathway through certification. If BETA can make those pieces work together, it may reduce the friction that slows adoption for everyone else.

The bull case is not simply that electric aircraft will exist. It is that BETA can make electric flight practical enough for real operators to introduce it before the wider market catches up.

Commercial and defence validation

Early relationships matter in a business like this. BETA’s website says it has an established customer base and more than 800 aircraft in backlog. I would not treat all of that as equivalent to firm, funded deliveries. The distinction between orders, options, deposits and delivered aircraft matters enormously.

There are more tangible points to watch. In March 2026, BETA announced that Surf Air Mobility had placed a firm order for 25 ALIA aircraft, with options for up to 75 more. The company also points to defence engagements and partner relationships. These do not remove the execution risk, but they are evidence that serious operators are engaging with the platform.

The financial reality

This remains a venture-style aerospace investment, even as a public company. In Q1 2026, BETA reported revenue of $10.1 million, a net loss of $122.3 million and cash of $1.59 billion. That cash balance provides time, but the company still needs to deliver against certification, production and customer milestones before it can demonstrate a self-sustaining commercial model.

What could break the thesis

• Certification delay: timelines can slip, and aviation approvals are not a formality.
• Production execution: building a few aircraft and manufacturing reliably at scale are different challenges.
• Order quality: backlog, options and partnership announcements must turn into funded deliveries.
• Cash burn and dilution: commercialisation takes time and capital.
• Infrastructure adoption: charging only becomes a moat if customers actually deploy and use it.

My current view

I own BETA because it appears to be building a practical entry point into electric aviation, not just an aircraft for a future market. The CTOL variant, useful payload, customer relationships and charging infrastructure are the pieces I think could matter most.

I will be watching certification progress, firm deliveries, operator utilisation, charging deployments and the cash runway. This is a high-risk position. The upside depends on BETA proving that the aircraft, infrastructure and customer economics can work together in the real world.

For the wider context, see my current portfolio. The financial figures cited above come from BETA’s Q1 2026 Form 10-Q. Product specifications and partnership information are from BETA Technologies.

This is my current investment thesis, not financial advice. I own NVIDIA in my main portfolio and the position can change.

The reason I continue to own NVIDIA is not simply that AI is a large theme. It is that NVIDIA has repeatedly shown an ability to turn a technological lead into an operating machine: products that customers want, supply that has to be coordinated at extraordinary scale, and earnings power that follows when the company executes.

That is the heart of my bull case. The next few years will not be decided by whether AI matters. They will be decided by which businesses can keep delivering the computing, networking and systems customers need as the buildout becomes more demanding.

Execution at a scale that is hard to ignore

NVIDIA’s latest quarterly filing illustrates the scale of the current opportunity. For the quarter ended 26 April 2026, the company reported revenue of $81.6 billion, up 85.2% year on year. Net income was $58.3 billion, compared with $18.8 billion in the comparable prior-year quarter.

Numbers at that level can make a thesis look obvious after the event. They are not. The reason they matter to me is that they reflect execution across a difficult chain: silicon design, manufacturing capacity, memory, networking, systems integration, customer deployment and software support. It is one thing to make a fast chip. It is another to help customers build and operate AI infrastructure at enormous scale.

More than a GPU supplier

I do not see NVIDIA purely as a seller of individual GPUs. The company’s position is strengthened by the wider stack around accelerated computing: systems, high-speed networking and the CUDA software ecosystem that many developers and organisations already use.

That does not mean customers cannot change suppliers. They can, and they will keep trying to improve economics. But moving a serious AI workload is not always as simple as comparing the headline price of two chips. Software tools, developer familiarity, model performance, networking and the operational cost of changing a production environment all matter.

The bull case is not that NVIDIA wins every AI workload forever. It is that its scale and execution keep it central to the most important workloads long enough for the earnings power to compound.

The architecture cycle is part of the thesis

One of the reasons I find NVIDIA interesting is the cadence of the platform cycle. Customers are not just making a one-off purchase. They are trying to build capacity for models, inference, agents and applications that are changing quickly. If NVIDIA can keep making each generation valuable enough to justify upgrades, the opportunity is larger than a single hardware refresh.

The company’s supply commitments show how much planning this requires. As of 26 April 2026, NVIDIA disclosed $119 billion of supplier commitments, with most expected to be paid during fiscal 2027. That is not proof of future demand, and it increases the importance of execution. It does show the physical scale behind the AI buildout.

What could challenge the thesis

• Demand digestion: customers may pause after a period of exceptionally heavy infrastructure spending.
• Competition and custom silicon: hyperscalers, AMD and other suppliers have strong incentives to reduce dependency and improve their own economics.
• Export controls and China: policy can affect which products can be sold and where future growth comes from.
• Infrastructure constraints: data-centre capacity, power and customer financing all matter. Demand for chips alone is not enough if the wider buildout stalls.
• Expectations: a company producing exceptional results can still be a poor investment if the valuation assumes too much perfection.

My current view

I own NVDA because I think its scale, product execution and ability to turn an AI infrastructure cycle into earnings are still unusual. The growth is already visible. The investment question is whether the company can maintain its central role as customers become more sophisticated, competitors improve and the market starts to separate durable demand from short-term enthusiasm.

I will keep watching revenue quality, gross margins, customer spending behaviour, platform transitions and the willingness of customers to build around NVIDIA’s broader stack. The thesis is strong, but it is not a reason to ignore the price paid or the risk that this cycle eventually slows.

For the wider context, see my current portfolio. The financial figures cited above come from NVIDIA’s Q1 fiscal 2027 Form 10-Q.

This is my current investment thesis, not financial advice. I own Palantir in my main portfolio and the position can change.

My interest in Palantir is not simply that it is an AI company. There are plenty of companies that can claim that label. The question I find more interesting is whether Palantir’s Artificial Intelligence Platform, or AIP, can become durable commercial infrastructure inside large organisations.

That is the part of the bull case I care about most. If AIP is helping customers move from experimentation into real workflows, it has a chance to become more than an AI feature or a short-lived spending cycle. It could become a platform that sits underneath important decisions, operations and automation.

The commercial adoption question

Palantir’s latest filed quarterly results give the thesis some substance. For the three months ended 31 March 2026, the company reported total revenue of $1.63 billion, compared with $884 million in the comparable period a year earlier. Commercial revenue was $774 million, up from $397 million.

Those numbers alone do not prove a long-term moat. Fast growth can attract a lot of attention, and it can make any valuation look more forgiving than it really is. But they do suggest that the commercial story is no longer just an aspirational slide in an investor presentation.

For me, the key question is what customers are actually buying. If they are mainly paying for short pilots and AI enthusiasm, the revenue can fade when budgets tighten. If they are using AIP to connect data, people and workflows in ways that are difficult to remove, the economics could be very different.

Why AIP could matter

Palantir describes AIP as a platform that connects large language models to an organisation’s data, permissions, workflows, agents and governance. That matters because most companies do not need another chatbot. They need a way to use AI in real operating environments without losing control of their data or creating unmanageable risk.

The potential appeal is that Palantir already has the components around the model layer. Foundry helps organisations map and work with data, the Ontology gives that data operational context, and Apollo supports deployment across complex environments. If those pieces work together in practice, AIP can be more useful than a standalone model interface.

The bull case is not that every company will buy more AI software. It is that a smaller number of important customers will build Palantir into workflows they do not want to unwind.

What I would watch from here

• Commercial growth quality: whether growth remains strong as deployments mature, rather than relying on a small number of highly visible wins.
• Customer depth: whether customers expand from pilots into repeatable, material programmes.
• Operating leverage: whether revenue growth continues to translate into disciplined profitability and cash generation.
• Remaining performance obligations: the company reported $4.5 billion at the end of Q1 2026, which is useful context, but not the same thing as recognised revenue.

The risks I do not want to ignore

The valuation risk is obvious. A business priced for exceptional execution leaves little room for disappointment. Commercial growth slowing, customers choosing cheaper or more open alternatives, or the market deciding that AI budgets were pulled forward would all challenge the thesis.

I also keep an eye on stock-based compensation. Palantir recorded $202 million in stock-based compensation in Q1 2026. That does not invalidate the business, but it matters when judging how much of the reported progress ultimately belongs to shareholders.

There is also a genuine execution question. Palantir operates in competitive markets and has to prove that its platform can scale commercially without becoming too dependent on a narrow group of customers or on an unusually favourable AI spending environment.

My current view

I own PLTR because I think AIP has a credible path to becoming operational software rather than a superficial AI add-on. The commercial growth is encouraging, but the enduring thesis depends on adoption becoming deeper and stickier over time.

This is therefore a thesis I want to revisit through customer expansion, commercial durability, margins and the valuation the market is asking me to pay. I am not assuming the outcome. I am watching for evidence that the platform is becoming harder to replace as it becomes more embedded.

For context on how this fits into the wider picture, see my current portfolio. The financial figures cited above come from Palantir’s Q1 2026 Form 10-Q.

AbCellera Biologics (NASDAQ: ABCL) offers a captivating opportunity for investors intrigued by the intersection of biotechnology and artificial intelligence. While the company’s share price has ridden a bumpy road since its IPO, AbCellera’s innovative platform, robust pipeline, and strategic partnerships suggest that this “AI-native” drug discovery company may be approaching an inflection point. For those eager to invest in the next frontier of medicine, here’s why ABCL deserves a closer look in 2025.

AbCellera’s Platform Advantage: Disrupting Drug Discovery With AI

At its core, AbCellera’s business model is built around a powerful AI-molecular discovery engine. Unlike traditional biotech firms that focus on a handful of experimental drugs, AbCellera leverages data-driven insights, automation, and proprietary algorithms to discover, design, and develop antibodies for a diverse range of diseases.

Three key platform strengths stand out:

• AbCellera’s platform can tailor molecular shapes to precise biological targets, making it possible to address highly complex or previously “undruggable” conditions.
• Partnerships with Big Pharma (like Eli Lilly and AbbVie) and government agencies generate milestone payments, royalties, and research revenue, creating a diversified income stream.
• As more drug candidates move through the clinic, the platform’s value compounds, establishing recurring revenue and a broader moat against competitors.

A good analogy here is comparing AbCellera to the “Palantir of biotech”—a company driving a major shift in how problems in medicine are solved by leveraging software and rapid data integration.

Pipeline Progress: Clinical Trials and Internal Programs

For 2025, ABCL’s narrative is pivoting from platform potential to clinical translation. The flagship internal drug candidates—ABCL575 (targeting the OX-40 pathway for immune-mediated diseases) and ABCL635 (addressing hormonal balance and menopausal symptoms)—are both entering clinical trials this year. Early-stage results could validate years of scientific development and unlock tangible value.

Other highlights:

• More than 20 internal discovery programs are progressing, giving AbCellera a robust internal pipeline beyond partnered efforts.
• The company aspires to submit at least two new molecules to clinical trials per year, aiming for a sustainable cadence of product launches.
• Success with these lead candidates could broaden AbCellera’s reach into autoimmune disorders, cancer, and complex metabolic conditions.

Building a GMP manufacturing facility in Vancouver means AbCellera can soon bring its drug candidates from lab bench to human trials even faster, bolstering the “design-to-delivery” promise.

Financials, Valuation, and Investor Considerations

AbCellera’s fiscal profile presents both upside and risk. While revenues dipped in 2024 due to a drop in COVID-related partnership milestones, analysts expect revenue growth to accelerate in 2025, with a forecasted 6.1% annualized growth rate—a healthy improvement from prior years’ declines. The company remains cash-rich, with $650 million on the balance sheet and an additional $200 million in committed government capital.

What investors should weigh:

• Analyst consensus rates ABCL as a Strong Buy, with a 12-month price target of $8 (roughly a 91% gain from recent levels).
• For the bold: Long-term models see bullish scenarios with multi-bagger potential by 2030 and beyond, highlighting the transformative prospects of AI-led drug discovery.
• The biggest risks: biotech is volatile, clinical trial failures or regulatory delays could impact sentiment, and milestone revenues remain lumpy.

Valuation now appears undemanding compared to the promise of its technology, and a single high-profile clinical win could rapidly change its earnings outlook.

Actionable Insights for ABCL Investors

• Keep a close eye on Phase 1 trial results for ABCL575 and ABCL635—early data will be pivotal for sentiment and stock direction.
• Monitor announcements regarding new partnerships or major milestone payments, which will bolster the revenue story.
• Watch for expansion in the internal pipeline, as more clinical trials signal the platform’s scalability.

Before You Go: Key Takeaways

AbCellera’s investment thesis hinges on the convergence of AI, drug discovery, and scalable innovation. Backed by a strong balance sheet, cutting-edge technology, and a maturing clinical pipeline, ABCL offers investors a levered bet on the next generation of biotech breakthroughs. Volatility and execution risks are par for the biotech course, but for those with patience and a long-term mindset, AbCellera’s journey toward redefining how new medicines are found and delivered may prove transformational. In short: ABCL is a stock for believers in the future of biotech—and one with the tools to shape it.

Investors should stay tuned for clinical trial milestones in 2025 and beyond; these could prove defining moments for both the company and its stock trajectory.

2026 update: I have published a separate ABCL investment update focused on the approaching clinical readouts. This 2025 thesis remains the record of my original view.

Explore More: My Current Investment Portfolio

Interested in exploring more investment ideas and tracking how these insights play out? Check out my personal investment portfolio, it offers a look at the stocks I’m currently following and investing in, including SOFI, Tesla, Rocket Lab, and other high-potential opportunities.