Nasdaq just tanked. What next?

We do not know whether the recent Nasdaq sell-off marks a pause, a correction or something more serious. A properly constructed portfolio should reduce the need to make that call today and importantly, the cost of getting it wrong. In this wire we’ll walk through how we approach this challenge.

AI is both an extraordinary opportunity and a portfolio risk

We believe AI is likely to change our lives, mostly for the better. But a powerful technology and a good investment are not always the same thing, particularly when expectations are already high.

Over the past year, technology and technology-related companies accounted for around 70% of global equity market returns. That does not prove AI is a bubble, but it does show that many portfolios have become increasingly dependent on one theme continuing to work.

Whenever one theme dominates a market, diversification tends to look unnecessary and we invite risk into our portfolios.

The objective is not to abandon AI exposure, but rather to balance this with other opportunities where investment outcomes are determined by a very different set of drivers.

Two questions for every diversifier

When we look for resilience, we apply two simple tests:

• Does the exposure genuinely diversify a risk already concentrated in the portfolio?
• Are we paying a sensible price for that protection?

The best diversifier can still be a poor investment if the protection is already fully priced. Conversely, an asset can be statistically cheap but offer little protection against the risk the portfolio actually owns. Cheap is not enough.

The most interesting opportunities sit where the risk driver is genuinely different and the starting valuation, yield or level of investor ownership leaves room for a useful payoff. Three current examples illustrate the framework.

1. Healthcare: diversify the earnings driver

Healthcare was left behind during the AI boom, but that is part of the charm. Demand for hospitals, medicines and medical services does not depend on another wave of data-centre capital expenditure. AI may improve drug discovery and diagnosis, but the sector's core earnings driver remains very different from semiconductors and cloud infrastructure.

We believe the combination of relatively consistent demand, more subdued expectations and recent resilience makes healthcare a useful way to reduce dependence on the AI investment cycle. We have been adding exposure to portfolios in recent months.

2. Duration: quit fighting the last war

The 2022 bond rout remains fresh in investors' minds. Bonds fell sharply as inflation surged and yields rose from exceptionally low starting levels. That experience has encouraged the view that bonds no longer diversify equities.

But 2026 is not 2021. With starting yields closer to 5% rather than 1%, investors receive materially more income plus duration has much greater capacity to produce capital gains if growth weakens and yields fall.

Expecting bonds to behave like they did in 2022 to a growth shock is, to us, like driving a WW1 tank into WW2. Don’t expect it to work. 

None of this means bonds will protect in every sell-off. A renewed inflation shock would still be difficult. Duration is held for a slowdown or disinflation scenario, not as a universal hedge. 

Last month we tactically reallocated capital across all portfolios, from equities to Australian bonds.

3. Yen: cheap skiing, potentially useful insurance

Anyone who has skied in Japan recently knows the yen is cheap*. Prices in US dollar terms remain unusually low across everyday measures, providing a better starting point than investors have seen for many years.

Valuation is only half the argument. For decades, the yen has been used as a funding currency for higher-yielding investments elsewhere. Investors borrow cheaply in yen and buy risk assets. During periods of funding stress or rapid deleveraging, those positions may be unwound and yen must be bought back, sometimes in a hurry.

This provides a potential source of diversification from global equities, although there are some complications from simply owning yen outright, not least the relatively high cost of holding it given the interest-rate differential. We have some indirect exposure in our portfolios but more direct exposure is an area we are exploring. It’s not as simple as buying Japanese equities. More to follow on this one.

Different protection for different seasons

Healthcare, duration and yen are not the only candidates. Selected smaller companies can benefit if market leadership broadens. Quality companies may be more resilient if earnings disappoint. Energy can help in a supply shock, while gold may respond to geopolitical or monetary-policy mistakes.

Resilience comes from not needing one forecast to be right

Some of these positions deliberately contradict one another. Energy may help if inflation or physical scarcity returns. Duration may help if growth collapses and inflation falls. Smaller companies could benefit from market broadening, while quality may prove more resilient if earnings disappoint.

That disagreement is not a flaw, it is the architecture of resilience.

We do not know what happens next to Nasdaq, AI or the global economy. Portfolio resilience does not come from getting that forecast right. It comes from owning a collection of assets that can respond to different failure modes, while remaining disciplined about the price paid for protection.