Artificial intelligence is attracting record levels of investment. Gold is trading near all-time highs. Questions are being asked about the long-term dominance of the US dollar. Geopolitical tensions continue to simmer in Eastern Europe, the Middle East, and Asia.
For many investors, the temptation is to wait for certainty before committing capital. According to Allan Gray and offshore investment partner Orbis, that may be precisely the wrong response.
At Allan Gray’s annual The Times investment update in Cape Town on 28 May 2026, portfolio manager Kamal Govan and Orbis director Matthew Spencer (pictured) argued that uncertainty is not an obstacle to investing. It is a permanent feature of markets.
The challenge for investors is not predicting the future but positioning portfolios to withstand a range of possible outcomes.
AI is attracting unprecedented levels of capital
Few themes dominate markets as much as AI.
According to figures presented by Allan Gray, about US$930 billion has already been invested in data-centre infrastructure over the past six years. Adjusted for inflation, that exceeds the cost of the Interstate Highway System by roughly US$230bn, the US railroad expansion by about US$370bn, and the Apollo space programme by more than US$600bn.
Orbis estimates that AI-related capital expenditure in 2026 alone could reach about US$650bn.
The spending reflects the speed at which the technology is being adopted. Allan Gray noted that while landline telephones took roughly 80 years to reach mass adoption and smartphones around 12 years, AI has achieved similar levels of uptake in just two to three years.
Yet Spencer cautioned against assuming that investors should focus exclusively on AI.
“We’re looking at AI companies, but we’re also looking at mining companies, financial companies, Japanese payment companies, Japanese real estate companies,” he said in an interview after the event.
“When all eyes go to one part of the market, you get all that excitement, all the capital goes in, pushes the prices up, and meanwhile, in the less popular parts of the market, there’s a lot of bargains to be had.”
Getting back to AI, Spencer said the scale of the opportunity is clear. Less clear is who will ultimately emerge as the biggest winners.
Investors are debating whether the greatest value will accrue to technology giants such as Alphabet, Microsoft, and Amazon, specialised AI developers, semiconductor manufacturers, or businesses that have yet to emerge.
Spencer said history offers a useful lesson.
“We’re not smart enough to predict who’s going to be the winner,” he said.
Drawing a parallel with Johannesburg’s gold rush, he noted that many of the fortunes created during that era did not come from finding gold itself.
“Back in the gold rush in Joburg, the people who made all the money were the people who sold the picks and the shovels. Very few gold miners actually made money.”
Rather than trying to predict which company will dominate AI, Spencer said investors should also consider the businesses providing the infrastructure that makes the technology possible.
One example he gave is South Korea’s SK Hynix, one of only three companies globally capable of producing the high-bandwidth memory chips required for advanced AI systems.
Another opportunity lies in energy.
Spencer noted that by 2030, data centres could consume more electricity than Japan currently uses. Meeting that demand will require substantial investment in energy generation and infrastructure.
Natural gas is likely to play an important role because it can provide the reliable, large-scale power needed by energy-intensive data centres. Spencer cited US natural-gas producer EQT as one example of a company positioned to benefit from rising electricity demand linked to AI development.
Gold and the dollar are telling a different story
AI may dominate headlines, but it is not the only trend reshaping markets.
Gold has rallied strongly in recent years, supported by geopolitical tensions, rising government debt, and growing central-bank purchases.
Govan noted that central banks have increasingly been diversifying reserves away from traditional US-dollar assets, while concerns have also emerged about the growing use of financial sanctions and the “weaponisation” of the dollar.
Spencer said Orbis remains underweight the dollar and has increased exposure to alternatives such as the Japanese yen and sterling.
The argument is not that the dollar is about to lose its reserve-currency status. Rather, it is that investors may have become overly concentrated in a market that has dominated global returns for more than a decade.
Gold itself is only part of the story.
Spencer said Orbis has reduced some of its exposure to physical gold and increased its exposure to selected mining companies.
“We’ve actually sold out quite a bit of the metal, and we’ve entered into some very low-cost miners,” he said.
At current gold prices, some mining companies are generating strong margins that are not fully reflected in their share prices.
You don’t have to predict the future
Despite the wide-ranging discussion on AI, currencies, commodities, and geopolitics, both speakers repeatedly returned to the same point.
Forecasting is difficult. Markets are shaped by events that nobody can predict consistently, whether wars, elections, pandemics, or technological breakthroughs.
The good news, Spencer said, is that investors do not need to get those predictions right.
“You don’t have to be able to predict the future in order to build a thriving portfolio.”
Govan made a similar point by highlighting how global markets have navigated decades of crises.
Since 1969, the MSCI World Index has delivered annualised returns of about 8% in US dollars despite oil shocks, Black Monday, the dotcom crash, the global financial crisis, Covid-19, and multiple wars.
Referring to a long-term market chart, Govan noted that events which once dominated headlines often become barely visible over time. Covid-19, for example, appears little more than a blip despite the fear and uncertainty it created at the time.
His point was not that crises are unimportant. Rather, it was that investors often overestimate the long-term impact of short-term events and underestimate the market’s ability to adapt and recover.
For long-term investors, the bigger risk may be reacting emotionally to uncertainty rather than the uncertainty itself.
Seek low expectations
If predicting the future is not the answer, what is?
Spencer said investors should focus on two principles.
The first is to “seek low expectations”.
When expectations are high, prices are often high as well. Even strong companies can disappoint investors if too much future success has already been priced in. Conversely, where expectations are low, companies do not need perfect outcomes to generate attractive returns.
Biotechnology was one example.
Spencer said the sector has fallen out of favour as investors redirected capital towards AI-related opportunities. Yet some biotechnology companies are trading at valuations that appear to place little value on future drug-development pipelines.
He highlighted Danish biotechnology company Genmab, which has a long track record of developing successful cancer treatments and bringing products to market. According to Spencer, the market appears to be valuing the company largely on its existing medicines and cash flows, while assigning relatively little value to its next generation of potential treatments.
For investors, the attraction is not knowing which future drug will succeed. It is investing in a company that has repeatedly demonstrated an ability to develop winning therapies, while paying a price that does not appear to fully reflect that future potential.
China was another area where Orbis sees opportunity.
He highlighted sportswear manufacturer Anta, which has increased its share of China’s sportswear market from about 5% to roughly 25% over the past decade.
Anta’s home city of Jinjiang produces more than a billion pairs of shoes a year and has become one of the world’s largest footwear-manufacturing hubs. Through acquisitions, the company has also built a portfolio of globally recognised brands, including FILA, Salomon, Arc’teryx, and Wilson.
Yet concerns about tariffs, geopolitics, and China’s economy continue to weigh heavily on investor sentiment towards Chinese equities.
Another example came from Allan Gray’s Frontier Fund.
When Russia invaded Ukraine in 2022, investors feared neighbouring Georgia could be next. Georgian financial shares sold off sharply as markets priced in a worst-case scenario.
Instead, Georgia benefited from an influx of about 100 000 wealthy Russians and Belarusians, stronger banking activity and increased trade flows.
Several Georgian financial stocks subsequently delivered returns of between five and 12 times their value after initially falling by about 50%.
For Govan, the example illustrates how periods of extreme uncertainty can create opportunities when market expectations become disconnected from underlying fundamentals.
Closer to home, Govan pointed to South Africa’s retail sector as an example of how investors’ everyday experience can differ from what drives market returns.
Retailers such as Shoprite, Pick n Pay, Spar, Mr Price, The Foschini Group, Truworths, and Clicks have faced pressure from weak economic growth and strained consumers, resulting in significant underperformance relative to other parts of the market.
Yet despite being businesses that South Africans interact with almost daily, these retailers collectively account for only about 5% of the All Share Index.
By contrast, gold, and platinum shares now make up close to 30% of the index.
For Govan, the comparison highlights how market returns are often driven by factors that differ from investors’ day-to-day experience of the economy.
It also illustrates why investors need to look beyond prevailing sentiment and focus on where expectations, valuations, and fundamentals may have become disconnected.
Be properly diversified
The second principle, Spencer said, is to “get actually diversified”. That means owning businesses whose fortunes are not all tied to the same economic outcome, sector, or investment theme.
He argued that many portfolios appear diversified because they hold multiple companies or asset classes. In reality, however, those investments may all depend on the same underlying assumptions about economic growth, interest rates, technology adoption, or market sentiment.
True diversification means owning investments that can succeed under very different circumstances.
A portfolio heavily concentrated in US technology shares, for example, may contain many companies but still depend on a similar set of outcomes. If those assumptions prove wrong, all the holdings may come under pressure at the same time.
Instead, Spencer argued that investors should seek exposure to a broad range of opportunities whose returns are driven by different factors. That could include businesses linked to AI infrastructure, energy demand, biotechnology innovation, Chinese consumer spending, precious metals, or different currencies.
The objective is not simply to spread investments across sectors. It is to build a portfolio where different holdings can succeed for different reasons.
These investments do not all rely on the same forecast about the future. Some may benefit from faster economic growth. Others may benefit from higher commodity prices, technological change, shifts in consumer behaviour, or geopolitical developments.
The objective is to avoid building a portfolio that depends on one outcome, one sector or one investment theme being right.
A genuinely diversified portfolio allows different parts of the portfolio to succeed under different conditions. That way, investors do not need to predict which scenario will unfold – only that the future is unlikely to play out exactly as expected.
Where advisers matter most
Periods of uncertainty often tempt investors to move to cash and wait for clarity.
Spencer described that as “an absolutely horrible idea”. Even when cash earns interest, inflation continues to erode purchasing power.
Govan reinforced the point with two observations. The first was that an investment of US$100 in the MSCI World Index in 1969 would have grown to more than US$12 800 by March 2026 despite oil shocks, wars, terrorist attacks, the dotcom crash, the global financial crisis, and the Covid-19 pandemic.
Many of these events dominated headlines at the time and felt existential to investors. Yet on a long-term market chart, some barely register.
The second was that investment outcomes become increasingly predictable as time horizons lengthen. Between 1950 and 2025, one-year stock market returns ranged from a loss of 37% to a gain of 52%. Over rolling 20-year periods, however, annualised returns ranged from 6% to 18%, with no negative outcomes.
Together, they reinforce a familiar investment principle: time in the market matters more than trying to time the market.
At the same time, remaining fully invested is not always possible. Life happens. Retirement, medical emergencies, job losses, family responsibilities, and other unexpected events can all require access to capital.
Spencer argued that the solution is not to abandon long-term investing but to align investments with the time horizon for which the money is intended. Money that may be needed in the near future should be invested differently from money intended for retirement decades away.
He said that is where advisers become particularly important. Their role is not to predict the next market shock but to help clients build portfolios that can withstand uncertainty, remain diversified, and stay aligned with their long-term objectives.
Disclaimer: The information in this article does not constitute investment or financial planning advice.
