Disruption and the price of patience

Markets are working through two related but distinct shocks. The first is straightforward and sentiment-driven: chipmakers and the industrial suppliers of the AI build-out have seen share prices climb sharply, with the market largely assuming the trend would continue rather than expecting the recalibration that usually follows a big wave of spending. Some of that recalibration has since come through, and semiconductors have eased back from their highs.

The second shock has run the other way. When investors start worrying about the long run, they sell first and sort later, and that’s roughly what happened to capital-light software and information-services businesses tagged as AI-exposed. Some probably deserved it. But many were caught in the crossfire, marked down alongside the genuinely vulnerable because the market hadn’t had time to tell them apart. That sorting has begun, and parts of software have recovered some ground, though it is still early.

For quality investors, this split has been a real headwind – quality hasn't lagged momentum by this much since the dot-com era. The interesting question isn't which style is in favour this year. It's whether the advantages that used to make a business hard to compete with – its data, its place in a customer's workflow, its network effects, the cost of switching away – still hold. We don't think they have broken, for most businesses, and both history and the outlook back that up.

A familiar pattern

Every disruption cycle produces a similar argument – that the incumbents are finished. We've had this argument before, more than once, over the nearly two decades we've managed the
Ninety One Global Franchise Fund.

In the 2010s it was the consumer space, traditionally a comfortable home for high-quality businesses with strong brands and distribution. E-commerce and digital advertising lowered the barriers that had protected those businesses for decades, and the slowest to respond saw their fundamentals erode. We reduced exposure to the companies least able to adapt, parts of food and household products among them, and reallocated toward technology businesses that were higher quality and more cash-generative. That wasn't drift. It was the same discipline, applied to a market that had moved.

Mobile was expected to end desktop software. The cloud was expected to end on-premise vendors. Each time, the pattern ran through the same stages: early euphoria, a sell-off in ‘legacy’ names, a reality check as displacing an incumbent proved harder than expected, and the incumbents that adapted regaining control by embedding the new technology into what they already owned. We made a similar call within the portfolio during the shift to digital advertising, exiting WPP as its scale advantage eroded under algorithmically optimised advertising from Google and Meta. We built a position in ASML instead, where extreme ultraviolet (EUV) lithography was emerging as the new source of moat. ASML's earnings have since compounded roughly 13-fold; WPP's have roughly halved (Figure 1). Microsoft shows a different route to the same outcome, adapting in place rather than being replaced, moving from per-PC licensing to subscriptions and now into an AI platform without losing its grip on the workflows it already owned. We believe the current AI cycle is following the same course and is sitting somewhere in that uncomfortable middle stage.

Figure 1: The evolution of quality

Share prices follow earnings

Index: December 2015 = 100, log scale

Source: Bloomberg, Ninety One, 31 December 2015 to 31 August 2026. Forward earnings per share (EPS) is Bloomberg consensus (BEST_EPS): ASML in EUR, WPP in GBP, both rebased to 100, log scale. For illustrative purposes only – not a recommendation to buy or sell any particular security.

The thread running through all three cycles is the same. Disruption isn't new; it's how markets renew themselves, and we've navigated it successfully more than once. What doesn't survive is a moat that stops where it was five years ago. A competitive advantage that isn't being actively extended is, by definition, allowing competitors to catch up. The businesses worth owning aren't the ones that once built a wide moat; they're the ones still digging it.

Expenditure scrutiny

The same market that has rewarded AI infrastructure spending is starting to look more closely at how it's financed. Cloud companies used to fund this kind of expansion out of their own cash flow; increasingly, they're borrowing instead, which raises the stakes if the payoff takes longer than hoped. Memory chipmakers like Micron and SK Hynix are especially exposed: shortages let them charge premium prices today, but that kind of pricing power has historically faded once new supply came online. That's spending to ride a boom, not spending to build something that outlasts it.

Not all heavy spending carries the same risk. Contemporary Amperex Technology (CATL), the world's leading maker of lithium-ion batteries for EVs and energy storage, is more capital intensive than our typical holdings. The business sits at the centre of two long-term growth trends and has built a position on manufacturing scale, supply chain control and hard-to-replicate technology. It also spends less per unit of capacity than its peers and funds its own growth rather than relying on debt. That's spending that entrenches an advantage. Spending to keep pace with a boom is a different kind of risk, and the portfolio remains less capital intensive than the market, even as expensive, capital-hungry tech names take over more of the index.

What this means for the portfolio

We don't dismiss the idea that businesses can be disrupted and lose the edge that once protected them. Some caught up in the current sell-off have real cause for concern, and some will undoubtedly be disrupted. But what separates a company from the narrative around it is how it adapts, not how convincingly it talks about the threat it faces. On that measure, many of today's supposed losers have a stronger defence than the market is pricing in.

Distinguishing between the sustainable and the unsustainable is the real work of quality investing. The differentiated insight is rarely in spotting that a quality attribute exists; it's in judging how durable it is, which must be a forward-looking assessment. What we look for hasn't changed: a hard-to-replicate competitive advantage, a dominant position in a growing market, sustainable cash flow and disciplined capital allocation, alongside the low cyclicality and balance sheet strength that provide resilience through the cycle.

It's worth noting, in passing, that passive flows do the opposite almost by construction, following whatever is driving the index, which today is an unusually narrow and expensive slice of the market. An active approach is forced to ask the harder question: whether today's leadership can still justify tomorrow's price.

We believe many of the advantages that predate the current disruption narrative still hold, but businesses need to keep adapting. Our track record through previous cycles suggests the underlying discipline can travel well across environments that look nothing alike.

Markets evolve. The opportunities within them evolve too, and so does the portfolio built to capture them. Figure 2 shows that evolution playing out over nearly two decades, as leadership shifts from a staples-heavy market toward technology, communication services and financials – several cycles of disruption, condensed into one chart.

Figure 2: Sector breakdown since inception

The portfolio may change significantly over a short period of time.
Source: Ninety One, FactSet, 30 September 2026.

What hasn't changed underneath any of that rotation is what we're looking for: durable moats, real cash generation, and the discipline to keep extending an advantage rather than simply defending one. The portfolio's return on invested capital has climbed to a record high over the past decade, while its valuation relative to the market has fallen to a decade low.¹ By most measures we track, it is higher quality, growing earnings faster than at almost any point in its history, and more attractively valued. We don't believe that disconnect persists indefinitely, and the underlying earnings keep compounding while we wait, an opportunity we believe should reward patience.

Download PDF

¹ Sources: Ninety One and FactSet, 30 September 2026.