Horizon Capital
I have launched a new investment fund, Horizon Capital, focused on the indispensable companies in artificial intelligence, energy, and life sciences. These companies offer sustainable, long-term opportunities because they address critical bottlenecks in essential industries. Essentially, they are the “arms dealers” as the world goes to war. Regardless of the winners, everyone needs the bullets from these companies.
Indispensable companies in AI, energy, and life sciences. The arms dealers and critical bottlenecks.
I invest in companies at the bottlenecks to critical industries. These companies solve essential problems that the most valuable and high-growth sectors require. The most attractive businesses combine broad demand with a capability that is difficult to replace. They supply the industry, resolve a constraint, and retain enough of the value created to reinvest at attractive returns. Their competitive positions are sustainable and profitable.
A good analogy is to think of the components of this portfolio as “the arms dealers to the most important wars.” It doesn’t matter who wins; everyone must buy their bullets. In addition, critical bottlenecks explain why the customer cannot easily bypass the supplier. An arms dealer is essentially the addressable market for all competitors. Critical bottlenecks identify the market sector and its sustainability. These create the most defensible and sustainable profits.
More frequent and intense volatility.
Research has shown that the most attractive investments tend to have the most significant drawdowns over time. In fact, stocks that have increased in value by 50x in the last 30 years had drawdowns between 50% and 70% at some point. A portfolio that outperforms the market must be able to withstand these kinds of drawdowns, and, given economic, geopolitical, and technological uncertainty, and the interaction of all these factors being magnified in today’s media, I expect this portfolio will have moments of excessive drawdowns. Investor patience and a long-term perspective are required.
The Horizon
As I have described in my book The Ten Year Horizon is critical to keep a long-term view. The waves will batter the boat sideways and even backward, but that’s no time to change direction and lose sight of the horizon. The portfolio focuses on this long-term perspective and can withstand the volatility we expect.
Perhaps the single biggest mistake any investor makes is selling a good investment too soon. People panic and sell at the wrong time, especially during a market correction or meaningful drawdown. However, these corrections and drawdowns are to be expected because that is a fundamental characteristic of public equity markets.
Portfolio Composition
Horizon Capital Full Report and Portfolio
The most important industries with the highest growth and profit potential, but also with identifiable and sustainable arms dealers and bottlenecks, are within artificial intelligence, energy, and life sciences. The portfolio holds 18 companies across these industries, allocating 62% to AI and semiconductor infrastructure, 24% to energy infrastructure and services, and 14% to life sciences. Individual positions range from 3% to 10%. These are strategic equity weights for a ten-year horizon, subject to valuation and the evidence supporting each thesis.
Of course, this portfolio is not set in stone, and there are fundamental assumptions as to why these companies are the most attractive. I do not look at market conditions to determine whether or not to exit the position. Instead, I look at the underlying assumptions that have led me to own that position and wait it accordingly. If any of those assumptions change, the weighting may change, or I may exit the position. I do this analysis daily, although modifications to the portfolio tend to be infrequent based on this fundamental perspective.
From investment principles to portfolio construction
In my book, Investment Principles: Strategies for an Irrational World , I examine the discipline required to make investment decisions under uncertainty. This portfolio applies that discipline to industries where technological progress depends on physical capacity, accumulated knowledge, and institutional execution. A large market is not sufficient. The business must capture value, convert earnings into cash, and justify the price paid.
In addition to “The New Ten-Year Horizon,” I developed a thorough treatment in my book, The New Fire: Artificial Intelligence, Geopolitics, and a Transformed World, that analyzes the interconnections and interdependencies underpinning technological development, including geopolitics, capital markets, and other factors.
Decisions are being made today about where to build data centers, how to structure semiconductor supply chains, which critical mineral deposits to develop, investment decisions, national security decisions, and bets on future technological architecture. These are all critical components of investment decisions, portfolio construction, and sustainable performance.
The portfolio understands the critical connections among technology, energy, biology, and the constraints on useful performance and profitability. AI needs manufacturable chips, reliable electricity, and efficient data movement. Biological discovery still requires clinical evidence, reliable production, and patient access. Those dependencies form the foundation of my research and require a deep understanding of these interconnections and where sustainable opportunity lies.
The discipline behind the allocation
The portfolio concentrates on capabilities that competing businesses need to turn investment into operating results. Manufacturing and design support computation. Power equipment makes capacity usable. Life sciences suppliers help translate discovery into reliable medicines. Each position and weighting in the portfolio must justify both its economic role and its share of capital.
Suppliers are not irreplaceable. Technologies and services can be obviated. Innovation and transformation is rapid and aggressive. The research behind this portfolio tries to capture the long-term view and sustainability of all these companies, understand how they capture their strategic advantage, and, more importantly, how it may be eroded. We try to stay ahead of this process, but it is dynamic and requires constant review.
The selection criteria and weighting framework
Necessary for the customer
I begin with the customer’s operating process. What fails, slows, or becomes more expensive if a capability disappears? A component can represent a small share of cost and still determine whether the entire system works. Process control protects manufacturing yield. Electrical protection enables safe operation. Drug containment protects the integrity of a medicine. Essentially, the company can be a small cost component to its customers but add disproportionate value. These are the best-positioned companies.
Difficult to substitute
A temporary shortage may attract capital and attention, but I am interested in durable bottlenecks that survive volatility, not temporary opportunities. The distinction depends on qualification time, accumulated engineering knowledge, embedded workflows, and the cost of failure. I test the advantage against credible alternatives, including redesign, internal development, and architectural changes. The relevant comparison is the customer’s total cost and risk of switching. Switching costs can be substantial, but not insurmountable. These need to be monitored constantly.
Able to retain economic value
An essential supplier can still be a poor investment. Customers may capture most productivity gains, competition may depress prices, or expansion may consume all incremental cash. I look for evidence that growth improves cash per share after maintenance investment, working capital, financing, and dilution. A backlog is valuable only when contracts produce acceptable margins and cash.
Broad enough to survive
The arms dealer test asks whether the business can prosper when industry leadership changes. TSMC serves competing chip designers. Research tools support competing scientific programs. Yet customer breadth does not remove a common spending cycle. NVIDIA and AMD supply AI developers while competing directly at the processor layer. Competitive interactions and shared customers exist, along with potential new competitors. NVIDIA and AMD face potential competition from enhancements at Intel, as well as internal developments at Google, Amazon, and Meta. It is a dynamic, valuable, and volatile sector. The winners will capture substantial value on an ongoing basis, but this needs to be monitored closely and reassessed constantly.
Quality and price remain separate decisions
The essential decision for any investment has three factors:
- Which companies are attractive?
- At what price?
- How much should I own?
This has been the essential process to determine the components and weights of this portfolio.
These weights define the intended fully invested equity portfolio. Capital should be committed as prices and evidence justify it. A higher strategic weight does not increase intrinsic value. Each holding requires a current valuation before establishing a purchase threshold. The portfolio represents my perspective on long-term ownership, with sufficient diversification but enough concentration in the most important industries and the most valuable and critical components of those industries.
For each holding, I would estimate cash available to owners after maintenance investment, working capital needs, dilution, and financing costs. The base case should use sustainable margins, realistic reinvestment requirements, and a terminal valuation that does not depend on perpetual enthusiasm. Downside cases should test weaker spending, lower margins, competition, and delayed execution.
Memory requires normalized cycle earnings. Life sciences suppliers require scrutiny of recurring demand, qualification and capacity utilization. Marvell requires attention to the announced transaction and the combined capital structure. Energy equipment and services require scrutiny of backlog conversion, contract terms, and customer spending.
What would change the decision?
I would reduce a position when a credible substitute weakens the economics, incremental returns deteriorate, cash generation fails to follow reported growth, or price implies outcomes the evidence cannot support. Quarterly review should revise the thesis where necessary. Price movement alone is not proof that the business improved or deteriorated.
Most importantly, my academic research and industry participation as a senior advisor to technology, life sciences, and energy companies enable me to see trends before the general market or media identify them. This enables more thorough, ahead-of-the-curve work on any trend that might impact the long-term value of any component of the portfolio. This enables me to revise holdings based on my proprietary research, network, and industry knowledge.
Risk and the underlying research framework
Test common dependencies
The principal risk is a simultaneous reduction in AI capital spending. Several holdings depend on the same customers and manufacturing chain. Owning more layers of that chain does not make their cash flows independent. A disruption in Taiwan, export restrictions, or a financing contraction could affect multiple positions together.
| Illustrative scenario | Assumed category declines | Portfolio effect |
| AI investment contraction | AI 45%; energy 35%; life sciences 20% | −39.1% |
| Severe Taiwan disruption | AI 60%; energy 20%; life sciences 10% | −43.4% |
| Broad valuation compression | All categories 25% | −25.0% |
These are arithmetic sensitivity checks using the 62% / 24% / 14% allocation. They are not forecasts, probabilities, or maximum loss estimates. They exclude rebalancing and recovery. Company outcomes could differ sharply within each category.
My point is that I am aware of the magnitude of these risks and am sensitive to impacts of economic, geopolitical, and industry trends. With this awareness, we can manage risk more effectively, but we can never eliminate volatility. This is the world we live in, and to ignore it would be foolish and naïve.
As I mentioned at the outset, stocks with the highest long-term returns have significant drawdowns. Patience and fortitude are essential. Also, market volatility will continue to be more frequent and intense. Recognizing the signal versus the noise is essential.
What is a real long-term market signal that must be acted upon, and what is noise that will resolve itself without further action? These are the critical components of the decisions made regarding the portfolio and any risk management modification or change in perspective about any holdings or sector.
My academic and intellectual framework
I have written eight books Amazon Author’s Page and published over 200 articles Nicholas Mitsakos Articles focusing on investment principles, technologies and transformation, geopolitical uncertainty, decision-making, and related topics and analysis. I lecture at several leading academic institutions and constantly revise my research and analysis. This informs my investment strategy and portfolio construction.
More specifically, my book, “Investment Principles: Strategies for an Irrational World,” says to distinguish a business’s quality from the attractiveness of its price, preserve room for error, and revise decisions when evidence changes. The arms dealer analogy is useful when it leads to specific questions about customers, substitutes, and cash returns.
In my book, “The New Ten Year Horizon,” I place these decisions in a wider system. Technology depends on energy, materials, manufacturing, institutions, and capital. A constraint can migrate as one problem is solved. The portfolio must follow that migration without assuming that today’s scarcity will remain tomorrow’s advantage.
Another of my books, What? So What? Now What?: Uncertainty, Transformation, and Upheaval informs how scenarios and incomplete knowledge are treated. A model makes assumptions visible; it does not remove uncertainty. My article, The Computable Molecule, extends that reasoning into biology, where discovery, clinical evidence and commercial adoption follow different schedules.
These are examples of the foundation of my analysis and recommendations. More examples appear in my writing, research, and lectures.
and the Horizon Capital website
My research and professional perspective
My investing, entrepreneurship, academic, and board service place operating consequences at the center of this analysis. My writing and graduate-level teaching connect those experiences to explicit frameworks for uncertainty, capital allocation, and technological change. Across these settings, I return to the same question: what must be true for a promising company to produce growing and sustainable value?
I’ve developed an investment process that crosses disciplinary boundaries and understands their interconnection. Engineering establishes feasibility. Customer behavior tests utility. Financial analysis tests value capture. Valuation determines the prospective return. Each stage requires its own evidence.