Some Sunday Reading: A Look into the Future!

One thing gold has not kept up with!
Gold is up 15 times in sterling terms since January 1982. UK prices (CPI) are up around 5.5 times, but the cheapest adult ticket at Twickenham today is £79, up 32x! If I had invested my £2.50 in the S&P 500 instead of going to Twickenham, It would be worth £600 today (with dividends reinvested)!
I was lucky enough to be in the South Stand with a great view from just behind the posts. Can anyone remember what I had a great view of (what the game was famous for)? Answer at the bottom of the page.
An excellent thought piece from Citrini is circulating. “What if our AI bullishness continues to be right...and what if that’s actually bearish?” It attempts to predict AI’s impact on employment, companies, the economy, markets, and life in general over the next two years. In its view, things will be very different. It’s written as a research note dated June 2028.
You can read it HERE.
If that’s too long, you can read this summary that my AI Personal Assistant produced for me:
“The 2028 Global Intelligence Crisis” (Citrini and Alap Shah, 22 February 2026) is a deliberately bleak scenario, not a forecast. It’s written as a memo from June 2028 and asks what happens if the AI bulls are right. AI agents make office work far more productive, so companies cut staff and spend the savings on more AI. Output keeps rising, but household spending, which is about 70% of the economy, shrinks. Software and every “middleman” business gets crushed. That spreads to defaults in private credit, stress at life insurers, and worries about mortgages in tech cities. By mid-2028, unemployment is over 10%, and the S&P is about 38% below its October 2026 high, while the companies that own the computing power, such as Nvidia, TSMC and the hyperscalers (the big cloud providers), keep booming.
Who benefits, in this scenario? AI chips and the factories that make them: Nvidia, TSMC, Taiwan and Korea.- Hyperscalers and the AI labs that own the technology.- Power and data-centre building.- Cheap payment systems such as stablecoins, with Visa better placed than Mastercard.- Healthcare and skilled-trade jobs, which hold up better than office work.
Who gets hit? Software sold on subscription, and the private-equity deals built on it.- Consultancies and IT outsourcing, India above all.- Middlemen: travel booking, delivery apps, insurance renewals, financial and tax advice, routine legal work and estate agents.- Card networks and consumer lenders, such as Mastercard, Amex and Capital One.- Spending that depends on well-paid office workers.- Private credit and alternative asset managers.- Housing in tech cities.
On Friday, I attended the MoneyWeek annual investment day.

Peter Frankopan kicked it off. My able AI Assistant produced this summary; some thought-provoking points. I particularly liked his view on copper and other commodities given my positioning.
SUMMARYThe speaker is Peter Frankopan, Professor of Global History at Oxford. His core message is that we’re in an age of risk. Leaders with little accountability, such as Trump and Putin, make big miscalculations and then double down. Europe has drifted for 30 years without clear goals. China has quietly done well out of the wars in Ukraine and the Middle East. Meanwhile, a powerful El Niño will add climate shocks over the next 9-12 months, as other risks already pile up. His advice to investors is to build resilience now, as Joseph did in the Bible by storing grain in the years of plenty.
THE TEN MAIN POINTS
1. Power without accountability leads to bad decisions. Leaders with autocratic or impulsive streaks double down on mistakes. He says Putin’s planned “short war” has cost over a million Russian casualties. Trump campaigned against foreign interventions and on cutting the cost of living, and has, in his words, "managed to do both” the opposite way, including a “catastrophic” intervention in the Middle East.
2. The midterms will slow Trump down. He expects two years of congressional investigations. He doesn’t think they’ll change the Ukraine or Iran situations, and doubts a Trump-Putin “grand bargain” will achieve anything.
3. Russia won’t settle. Putin turned down a generous offer that reportedly included US funding for Arctic oil, chemicals and minerals and a return to the G8 and G20. He sees the war as an existential struggle against the West, so expect a long new Cold War rather than a handshake.
4. After Putin, expect more of the same. Putin will stay until he can’t anymore, and there’s no successor in sight. Historically, you “never bet on a dove”. The people who wanted to work with the West have left Russia, and the state machinery rewards loyalty.
5. The West’s credibility has worn thin. Seen from Asia and the Gulf, the West looks reckless and unaccountable, with double standards on military intervention. The UN has been weakened, mostly by the US, and we expect too much of it.
6. Europe needs to wake up. It has been “under an anaesthetic”, with no goal beyond keeping things as they are. Trump has finally given Europe some backbone on defence and its place in the world. He also thinks the UK suffers from a shortage of capable people at the top, with too few leaders who understand science, maths and technology.
7. China has gained from both wars. It buys cheap, sanctioned Russian and Iranian energy. Every ship hit in the Gulf is a win for its shipbuilders. It also gets to present itself as the responsible player. Its trade surplus is running at about $110-130bn a month, after $1.2tn last year.
8. China’s real worries are at home, not Taiwan. Its big problems are moving from manufacturing to consumer spending, local government debt, and its workforce and population, which may be smaller than official figures say. Its military activity near Taiwan was unusually low when US fleets were busy in the Middle East. That suggests either great caution, doubts about its own military, or that Taiwan is less of a flashpoint than people think.
9. A climate shock is coming over the next 9-12 months. The oceans absorbed about 23 zettajoules of heat in 2025, which he likens to 12 Hiroshima bombs every second. That brings a strong El Niño, with floods, broken supply chains, weaker crop yields, higher food, energy and insurance costs, and more inflation. India’s heat is already hurting crops, with all 50 of the world’s hottest cities in India in May. Climate emergencies also push governments towards more centralised and less democratic rule.
10. Points for investors:- Demand for commodities will stay strong, with the energy transition needing about 7m more tonnes of copper a year and Asia needing $1.7tn of infrastructure spending a year.- China trades at a discount for good reason, because the Communist Party can take a business model away overnight and put politics ahead of shareholders.- China is pushing open-source AI partly to break its dependence on US tech giants.- Above all, stay resilient so you aren’t caught at the wrong point in a funding or debt cycle.
Richard Clode of Bankers Investment Trust made a convincing argument that AI growth is a supply problem, not a demand problem. A shortage of data centres will keep returns high for those who have them or build them. SpaceX?
SUMMARY
Richard Clode, who co-manages Bankers Investment Trust at Janus Henderson, argued that last week’s sell-off in AI shares doesn’t change the story. Demand for AI is soaring, and the real limits are on supply: data centres, power and chips. Those limits should slow things down rather than derail them, and they also make it hard for the industry to overbuild. He thinks the big, cash-rich companies such as Amazon, Microsoft and Nvidia are reasonably valued and best placed to win. He avoids riskier, debt-funded players and sees sharp drops like this as chances to buy.
TEN MAIN POINTS
1. A far bigger market than IT budgets. AI is going after the whole market for knowledge work, which is worth trillions. Firms like law firms, accountants and banks will deliver their services through a mix of people and AI agents, rather than simply buying software.
2. The technology is moving fast. It has gone from ChatGPT to reasoning models, and then to agents, since late last year. On 6 February this year, AI agents produced more AI output than humans for the first time. Physical AI, such as robotaxis and robots, is the next wave.
3. No shortage of demand. Agents are using AI far faster than people do, which is like a company hiring thousands of extra staff overnight. Elon Musk claimed a data centre pays for itself in about a year today, and in about six months next year.
4. Supply is the real constraint. The midterm results spooked the market, especially Texas, where about a third of US data centres are being built, and the new Democratic governor wants to pause new permits. Sites already permitted for 2027–28 aren’t affected. He reads her stance as “build them properly”, with openness, union labour and clarity on electricity bills, rather than a ban.
5. Cheap Chinese and open-source models add to the market rather than undercut it. Simple jobs will use cheap models and demanding work will use the best ones. The hyperscalers (giant cloud providers such as Amazon, Microsoft, and Google) offer every model and every type of chip so that they can match each job to the right model at the right cost. He compared this to how Apple won smartphones by owning the whole system.
6. Making money faster than he’s ever seen. Anthropic is growing its revenue far faster than Google ever did, and it was profitable in the third quarter, years earlier than expected, ahead of its IPO in the next few weeks.
7. Spending first, cash later. It’s unfair to compare the early stage of AI with the late, cash-generating stage of the internet. Nvidia is producing huge free cash flow and returning roughly half of it to shareholders.
8. Borrowing is a real worry, but not yet a crisis. The biggest new spenders, Anthropic, OpenAI and SpaceX, need outside funding, and so-called circular financing is a fair concern. Most data centre debt, though, is secured on the chips themselves, and those bonds haven’t moved. He avoids high-risk “neoclouds” such as CoreWeave.
9. Valuations are reasonable for most of the big names. The exceptions are Tesla and SpaceX, which he doesn’t hold. He added that plenty of popular non-tech companies carry far more debt.
10. Q&A, on sell-offs, power and Bankers itself. Sell-offs of 20–30% come along most years, driven by forced selling from hedge funds and quant funds, and an investment trust’s permanent capital lets him buy instead. The power shortage is real but manageable: fixes include using spare grid capacity outside peak times, data centres generating their own power, more efficient technology and building in more remote places, and battery storage, which should also boost lithium demand. Bankers dates from 1888 and has raised its dividend 59 years in a row; it aims to grow capital and dividends by focusing on profits, cash flow and sensible valuations.








Answer: Erika Roe!






Comments