Source: AI by Gemini
“No one will pay good money to get from Berlin to Potsdam in one hour when he can ride his horse there in one day for free.”
– King William I of Prussia, 1864
Portfolio recovers from poor July
The Loftus Peak portfolio has responded positively in August as a trickle of highly significant data leaked into the public domain.
Most notable: there are strong suggestions that Anthropic’s revenue is annualising at over US$100 billion a year, based on recent leaked numbers, which doubtless come from the company’s initial public offering documentation. This suggests Anthropic will come to market at a value of between US$2 trillion – US$3 trillion, more than SpaceX and even more than Saudi Oil giant Aramco when it was listed in 2019. At US$2 trillion, Anthropic would rank within the top ten largest companies in the world.
At this stage, it appears that Anthropic is opening up a convincing lead over OpenAI, the company that Microsoft formed an early, but as it turns out not a durable, partnership in 2022.
But it’s the capital expenditure numbers that are truly staggering – they are the largest ever observed in investment markets in history, on par with the roll-out of railroads in the US and Europe in US dollar-adjusted terms. The investments these companies are making to effectively distribute AI services to users, both in enterprise and consumer markets, are scheduled to top US$4.6 trillion cumulatively from 2025 to 2030 for the participating hyperscalers Google, Amazon and Microsoft, according to Bloomberg estimates (including adjusting for Microsoft’s treatment of this capex as a finance lease).
To put these numbers into perspective, they are larger than the annual GDP of each country in the world with the exception of the US, China and Germany.
But these numbers should be seen in the context of cashflows of these same hyperscalers which is forecast to be US$5.3 trillion over the period, according to Bloomberg estimates.
AI and the industrial revolution
The industrial revolution resulted in machines which replaced brawn, producing in an afternoon what a team of labourers would do in a week. The invention of the steam engine, the weaving mill and the industrial furnace lifted outputs by 1000’s of times those of the artisan class. It would have been a grave error of judgement to minimise the impact the industrial machine has had on the human race.
But importantly, the benefits of the industrial age tools took decades to filter through society, and indeed many industries were ruined by these technologies in the mean time.
There seems little doubt that we are in another – possibly the largest – revolution to date: bigger than electricity, the internet and possibly even the railroads.
Reasons this revolution may still have room to grow include a revenue base for AI over US$100 billion and growing, just a few years from launch, and the fact that the major players are well-capitalised. In the case of railroads, it took decades to recoup the investment, while the dot-com bubble took a decade – but then delivered virtually everything which had been promised.
The increasing exposure to biotech as technology leads to treatment breakthroughs
Meanwhile, outside of AI, there are some very significant disruptive developments, especially in biotechnology, which now makes up more than 10% of the portfolios managed by Loftus Peak.
Investors will have likely noted the returns from Eli Lilly, a company in which we invested not simply because of the weight loss drugs it produces, but because of the collateral benefits these drugs have on the health of the heart, vascular system, liver and kidneys, to name just the most important.
Less well known are the portfolio’s two cancer company positions. Moderna, a company which Loftus Peak portfolios entered earlier this year, last week announced the results of a successful Phase 3 trial showing five year recurrence-free survival in certain patients who had melanomas.
Importantly, it was Australian scientists, including Australians of the Year Professor Georgina Young AO and the late Professor Richard Scolyer AO, that played a crucial role in advancing Moderna’s cancer vaccine breakthrough. Their contribution included clinical trial leadership, foundational oncology research and strategic manufacturing partnerships.
The stock rose +174% on the night of the announcement. Obviously, the revenue implications may be significant, but at the same time there was excitement because the unprecedented treatment prevented the cancer from metastasising, often a silent killer.
Biotechnology now makes up around 10% of the portfolios managed by Loftus Peak.
Intismeran, which is the vaccine’s name, works by using a sample of an individual patient’s own tumor to target unique cancer mutations (neoantigens). It does this using messenger RNA (which is related to DNA) which instructs the body’s cells to produce those specific neoantigens, training the patient’s own T-cells to recognise and attack the cancer.
Moderna is hoping to develop the experimental vaccine technology to treat different types and stages of cancer, such as non-small cell lung cancer, which is part of the reason why the share price jumped so much.
Separately, Loftus Peak is also an investor in Revolution Medicines, which hit the headlines a few weeks ago with an effective treatment for pancreatic cancer, previously considered “undruggable”.
The method of action here is novel. Cancer cells are notoriously smooth in shape, so finding an antigen to bind to them is very hard. Revolution Medicine’s approach was to find a substance, Cyclophilin, which naturally binds to the cancer, and attach its drug, Daraxonrasib.
Both these drugs, and indeed the weight loss treatments of Eli Lilly are not of themselves products of artificial intelligence. But they are here because of the explosion of processing power, enabling scientific teams to do in days what previously took years.
This work, indeed all the work, which Loftus Peak undertakes is designed to unearth the real value of the companies which are at the forefront of these developments, and has underpinned outperformance for more than ten years.
Share this Post