When the bots begin building themselves
Last week, Anthropic (the private company behind the chatbot Claude, and currently the largest single position in the Fund), sent shockwaves through equity indices and private credit markets when it released a Generative AI tool that can automate legal work such as contract review, compliance workflows, and legal briefings. Claude’s coding tools, enabling ordinary non-technical people to develop custom software in minutes, had already been capturing much attention. The combination of these two releases sent stocks of software, data, and analytics companies plummeting amid fears over the impact of AI on their business models… thus wiping more than $1 Trillion in market value off the exchanges in just a few days.
Two days later Anthropic and OpenAI released their most recent models, Opus 4.6 and GPT 5.3 Codex respectively, marking an inflection point in the blistering speed of AI tool development. What’s profound about these releases is that predecessor models were instrumental in creating them. We are entering the phase where, according to Anthropic CEO Dario Amodei, “the current generation of AI builds the next”. Think about that for a minute. The bots are starting to build the new bots autonomously. Many believe that an Artificial Intelligence explosion is inevitable, the likes of which most of the world population isn’t prepared for. While no one knows exactly what the future holds, generative AI will gradually start to replace some degree of human cognitive work. AI may bring many benefits to mankind, but it may also dramatically alter the landscape of the job market. I believe that no job that can be done on a computer is safe from AI disruption in the near to medium term.
Those that have not used recent premium-tier versions of the frontier models may take a skeptical view. The active models of merely months ago were rife with errors and hallucinations. The most recent premium frontier releases, however, have dramatically cut down on inaccuracies and are even exhibiting judgement and intuition that would have been unfathomable only a year ago. This feedback loop of AI building the next generation, which is then smarter and builds the next even faster, which is then smarter… is inflecting higher and gaining steam as the pace of innovation grows exponentially.
These developments, which can be both exciting and horrifying, further increase my sense of urgency to gain equity exposure to leading AI companies to which the value of these advances will likely accrue. If the robots are coming for white-collar jobs, I want to own the robots. Virtually all of the leading-edge AI pure-play companies are still private and thus cannot be accessed by everyday investors in the public markets. These companies would have already been public a generation ago… but numerous forces have driven a great growth migration away from public markets, making private market access and participation imperative.
Market/Macro Themes and Outlook
Asset returns are typically driven by the interplay of 2 forces: economic growth (or contraction) and inflation (or disinflation/deflation). Currently we appear to be in an inflationary boom. Barring unforeseen exogenous shocks, the following conditions inspire confidence that a material downturn in the business cycle is unlikely this year:
Headwinds:
· Trade War – Effects are lessening
· Job growth (lack thereof) due to immigration policies – Mild effects
· Defaulted student loan collections restarted in May, after interest & payments were paused for 3.5 years during COVID – Mild effects
Tailwinds:
· Big Beautiful Bill’s accelerated depreciation to increase Capex, which is expected to increase GDP by 0.9% according to CBO
· Cheap Oil – Acts like a tax cut
· Depreciating Dollar
· AI Revolution driving earnings growth and encouraging significant capital expenditures
Headwinds have been weakening while tailwinds have been strengthening. Expectations of a recession in 2026 are low while inflation is expected to remain elevated, at least partially driven by a global regime change that has been underway for the past decade. From the 1980s through the early 2010s, globalization was in full effect with a focus on optimizing for manufacturing and supply chain efficiency. Outsourcing production to lowest-cost locales and just-in-time processing ruled the day. Today, we face a different dynamic as geopolitics have taken a multipolar and nationalist turn, with focus shifting to optimizing for supply chain and resource certainty rather than efficiency. Globalization has thus reversed, compounding inflationary pressures fueled by the COVID-era stimulus debt explosion.
As the Fund’s current asset allocation would suggest, these conditions and dynamics promote a preference for real assets, such as natural resources and infrastructure, while drawing interest away from long-duration fixed-rate debt.
Alpenglow Total Return, LP – Recent Performance and Current Exposure
In its first two months since public debut, Alpenglow Total Return LP delivered performance of +6.24% and +2.26% (Dec ’25 and Jan ’26, respectively) for share class P, net of fees and expenses. There are no share class A investors to date. This very limited track record should not be viewed as a prediction of future performance.
Despite being carried at conservative valuations, ~27% of the current portfolio is comprised of equity in cutting-edge private growth tech, most of which are central to the AI revolution. While investing in late-stage venture and growth equity undoubtedly carries risk, I believe that the current risk/return profiles of our holdings are very attractive.
As mentioned previously, Anthropic a has grown to be the Fund’s largest single holding at ~11%. In my view, it is the most compelling company to own today as it quickly closes the gap on OpenAI’s pole position in the AI arms race. Its rapid ascent to dominance is attributable to its leading coding and software development tools, lack of boardroom & legal drama, commitment to AI safety, a more lucrative focus on enterprise applications rather than “freemium” consumer offerings, and more conservative Cap-Ex discipline relative to OpenAI. Revenues, through partnerships with some of the largest enterprises on the globe, are currently growing by >1,000% annually. Just yesterday it announced the successful completion of an over-subscribed $30 Billion fundraise at a $380 Bln valuation, up from its September Series F that valued it at $183 Billion. The company has suddenly captured the investment world’s attention, and I am very optimistic about its future prospects.
OpenAI takes up ~3% of the portfolio as we’ve been harvesting healthy gains and rotating into a better risk/reward profile in Anthropic.
The remaining ~13% of the “private growth tech” allocation is collectively comprised of companies such as SpaceX/X/Xai, Databricks, Stripe, Perplexity, Anduril, Mercor, Rippling, Replit, Nscale, WIZ, Upgrade, Lambda, Turo, Dialpad, Monzo, Bolt, and Fanatics, among others.
Outside of Private Growth Tech, the rest of the Fund is allocated conservatively across Private Equity, Private Credit, Hedge Funds, Real Assets in natural resources and infrastructure, a tactically managed Global ETF mix, and a ~15% cash position to fund future capital calls. While most of these allocations would likely put readers to sleep if I elaborated, but their risk and return profiles are very attractive.
A recent addition to the portfolio includes a commitment to a Private Equity holding company comprised of B2B essential service businesses that I expect will be relatively unaffected by economic downturns. It currently kicks off a ~20% annual distribution yield and has achieved a 30%+ total return during its 8-year operating history. I expect this to be a core holding going forward as it shines in terms of risk, return, and cash flow while lacking the complexity of a draw-down structure. Without using leverage, it acquires boring but durable family-run businesses at low multiples (~4X on average) from retiring owner/operators. It then improves operational efficiency, grows top and bottom-line, and holds and operates them indefinitely while paying out operating profits to investors quarterly.
Another portfolio addition is a structured finance business that offers convertible debt financing to small and micro-cap companies, delivering asymmetric and uncorrelated return drivers. In its 16-year operating history, it has never had a down year and has compounded capital at 28% during that stretch.