Steering portfolios through a fragmented world

Assumptions underpinning traditional portfolio construction are under pressure from a series of stagflationary shocks and as stock-bond correlation turns positive. The illiquidity premium in private markets has proved largely illusory, and geopolitical instability can reprice assets faster than any model can respond.

The investors who navigate this environment most successfully will be those who have planned for regime change rather than cyclical variation, and who have the governance infrastructure to deploy capital when others are forced to sell.

Bernard Wee, group head of markets and investment at the Monetary Authority of Singapore, told the Top1000funds.com Fiduciary Investors Symposium in Singapore that the authority views diversification as a mechanism to deal with “changes in regime”.

“It’s not that you don’t understand how assets perform in particular regimes, but that regimes can change very frequently,” Wee said. At the end of last year, MAS expected a regime of growth and falling inflation, but the regime switched quickly to one of high inflation. “and the assets performed exactly as you’d expect when the regime changes”.

“For us, we approach the question of diversification as diversifying across different regimes and understanding how different assets perform in different regimes,” he said.

“Some of the work that we’re doing this year [and] next year is identifying regimes better and understanding the impact of how you construct your portfolio to be resilient against different regimes.”

Modelling investment scenarios is a challenging task when circumstances can change on, literally, a single tweet. Franklin Templeton head of outsourced CIO Rich Nuzum said: “We’ve swung from risk-on/risk-off, which was the prevailing environment for 10-plus years, into a couple of big stagflationary shocks”.

“Liberation Day, the tariff announcements – stagflationary. The US de-immigration policy – a stagflationary shock. And now the conflict in and around Iran – stagflationary,” he said.

“That flips your stock-bond correlations to be positive, instead of giving you any diversification. And some of the stagflationary shocks are coming from policy announcements, so that’s not something you can model easily. You get a tweet and suddenly the covariance matrix flips on you.”

The factors driving a shift in the global order have not been gradual, they have landed through specific events that have forced investors to reprice risk almost overnight.

Wee said these sorts of shifts mean it is “very hard, and we are very reluctant” to pick winners between countries.

“Ten years ago, you might have picked Russia as an economy on the up and up, and that changed overnight in 2022,” he said.

“Ten years ago, India would have been one of the fragile five. As allocators, we are very reluctant to make big decisions that can change over the horizon of your portfolio.”

Nuzum said the Franklin Templeton team spent a week leading up to the World Economic Forum in Davos earlier this year “speculating whether the US would engage in a military conflict over Greenland”.

“When we hit Davos, [our] European contacts – whether clients, government, NGOs, journalists – were angry and afraid,” he said.

“Our Middle Eastern and Asian contacts were just afraid that the world order was going to be shattered by actual armed conflict over Greenland. When President Trump ruled out military action, there was a massive exhaling of breath. And we all went back to obsessing about AI.”

The force of AI

Alongside geopolitical uncertainty is a second force: the impact of AI as a dominant driver of equity market concentration which for investors that read the situation most accurately promises a significant long-term opportunity.

Wai Seng Wong, head of strategy at Malaysian sovereign fund Khazanah Nasional, said it’s important for asset owners to understand whether the impact of AI is “structural, cyclical, or are we facing a current cycle within a long-term structural change?”

“In our developed markets exposure, 30 per cent is AI or AI-adjacent,” Wong said.

“In emerging markets Asia, because of Taiwan, Korea, and now China, 40 per cent is AI-adjacent. I see it as a cycle within a structural change, because in 2025 roughly 40 per cent of returns came from the [Magnificent] Seven. The tail risk does seem to be broadening a bit, which is a relief from an allocator perspective.”

Nuzum said Franklin Templeton believes that “about a third of the global economy will get completely rewired” by AI.

“Global GDP last year was roughly $117 trillion. We think there’s $37 trillion worth of industry sectors that haven’t yet been disrupted and transformed by AI and digitalisation,” he said.

“And collectively, as a global economy, we’re still under-investing.”

Debt worries

If the impact of AI is potentially disinflationary, it is countered by issues pulling in the opposite directions, including increased sovereign bond issuance and mounting fiscal deficits and credit spreads.

Wee said it is “very easy to get worried about fiscals, very easy to get worried about supply [and] very easy to get worried about interest costs”.

“But there’s also supply coming out from the corporate sector, and that supply wasn’t there in the last few years, and it’s only starting to come out with Amazon – $40 billion two weeks ago – and Meta – $30 billion last year. That is going to add to supply in the overall bond market.”

Wee said the likely impact would be twofold: a steepening of the yield curve, and a widening of credit spreads. The one risk that neither bonds nor equities can diversify away is the discount rate.

“When the discount rate is rising, both asset classes will do poorly, which is exactly what happened in the 2022–23 hiking cycle. And that leads to a second question: how do you diversify your diversifiers?”

‘Prepare to act’

With traditional defensive assets potentially offering less reliable protection, managing downside risk has become more difficult. For these investors, the answer lies not in hedging but in staying liquid and being prepared to act when others cannot.

Nuzum said Franklin Templeton’s “biggest recommendation to clients is, if they possibly can, to have an opportunistic bucket in their risk budget and physical cash as dry powder to deploy when other people are forced sellers”.

“And one reason to have that is the policy shifts from not just the US, but potentially other countries, changing direction, announcing something that creates forced sellers.”

But Nuzum said it is not enough on its own to have cash on hand, “you have to have the governance in place to put the money to work when it looks like the real world has a big problem”.

“That’s a really good way to outperform over the long term, but it takes a lot of pre-planning and discipline,” he said.

Wong said the challenges of investing in an increasingly fragmented world are elevated by the peculiarities of its investment mandate as a sovereign institution with both developmental and commercial obligations.

He said in the past 12 months, it had become clear that private markets, for example, had not delivered the illiquidity premium expected.

“Not [in] the last 12 months only, we started seeing it before that, but in the last 12 months I think we’ve kind of accepted the fact that… a lot of what we got was actually just illiquidity itself,” he said.

“[That] is a real pressure to a portfolio that needs to generate not only return, but capital recycling into both our mandates.”