Bob Tannahill, Head of Asset Management, Channel Islands, writes this week’s update.
The defining choice facing equity investors today appears uncomfortably simple: embrace artificial intelligence and accept the accompanying risks or avoid them and be prepared to underperform.
This dilemma reaches far beyond the US technology sector. AI-related demand is influencing semiconductor manufacturers, data-centre operators, utilities, energy producers and also the companies supplying the physical infrastructure needed to support the buildout. At the same time, a relatively small group of businesses has become increasingly important to the performance of major equity indices. From the US to Japan, the AI trade runs deep and has even recently leached into bond indices as AI-related debt issuance grows. As a result, investors may have considerably more exposure to the AI theme than they realise.
The risks are not difficult to identify. Market concentration is elevated, valuations in parts of the technology sector leave limited room for disappointment and much depends on assumptions about future profitability. Nobody yet knows how far the performance of large language models can be scaled, how quickly AI will become commercially useful across different industries, or how the eventual economics will be divided between model developers, infrastructure providers and their customers.
Avoiding the theme entirely, however, creates a different risk. AI investment is supporting genuine revenues and profits today. If the technology delivers a sustained improvement in productivity and expands the economic pie, investors who recognised every danger but captured none of the upside may eventually look just as foolish as those who invested without restraint.
The experience of the “Magnificent Seven” stocks also offers an important reminder. A powerful profit cycle does not necessarily have to end in a dramatic collapse that vindicates every sceptic. Market leadership can instead broaden gradually as the early winners hand the baton to the companies building, supplying and eventually using the technology. Titan Wealth’s recent investment commentary has similarly highlighted an opportunity extending beyond AI developers into semiconductors, memory, networking, power infrastructure and other parts of the supporting ecosystem.
So, must investors be either all in or all out? We believe there is a middle ground.
In our multi-manager portfolios, this means maintaining actively managed and carefully sized exposure to the AI buildout. The objective is to participate in the profits being generated without allowing one theme to dominate portfolio risk. Strong performance should also create opportunities to take profits and recycle capital into other potential sources of return. This approach reflects our wider portfolio process of making tactical allocation decisions, combining funds and investment themes, and using structured risk oversight and feedback.
The other side of that approach is to own assets that could benefit when market leadership changes. Healthcare is one candidate. The sector has been deeply out of favour but offers comparatively predictable earninnavigatgs and more attractive valuations. If investors retreat from highly valued AI beneficiaries, healthcare could provide a natural destination for some of that capital.
Software may offer another opportunity. If AI becomes cheaper and more widely available, attention may shift from the companies that financed the infrastructure towards those able to use it most effectively. The eventual winners may include businesses that apply inexpensive AI to improve productivity, deepen customer relationships or strengthen existing competitive advantages.
Diversification must also extend beyond equities. Insurance-linked securities, defensive fixed income and absolute-return strategies can provide a secondary line of defence because their return drivers are less directly dependent on equity markets or the success of the AI cycle. This is not a guarantee against losses, but it can reduce reliance on any single economic outcome.
So far in 2026, this balanced approach has worked. Technology and AI-related holdings have contributed strongly, while regional equities and diversifying strategies have helped broaden the portfolios. We have taken profits selectively rather than abandoning the theme, consistent with our view that position sizing matters as much as identifying the right investment trend.
As always in markets, executing this strategy is easier than done. AI may fulfil its promise, disappoint spectacularly, or follow a more complicated path containing elements of both. Investors do not need to make a single heroic forecast. They do, however, need portfolios capable of participating if AI succeeds and remaining resilient during periods when enthusiasm periodically falters.
