DBS has argued that the artificial intelligence trade is moving beyond the broad early-cycle enthusiasm that lifted nearly every name linked to the technology. In its latest view, the bank said investors are now separating companies that are merely spending on AI from those that are converting adoption into tangible business outcomes. The distinction matters because the market's next leg is likely to be driven less by narrative and more by proof: higher productivity, stronger margins and recurring revenue growth.
Selective AI Phase
DBS's stance reflects a wider maturing of the AI investment cycle. In the first phase, capital flowed aggressively into semiconductor makers, cloud providers, data-centre operators and software firms promising AI-enabled transformation. That phase rewarded scale, capacity expansion and early positioning. The current phase, by contrast, is about discipline. Investors are asking which companies can use AI to do more with less, rather than simply spend more to stay relevant.
The bank's preference for lower-capex AI adopters is especially notable. These are the firms that can integrate AI into existing workflows, customer service systems, analytics engines or internal operations without committing to large, ongoing infrastructure outlays. For such companies, AI can improve operating leverage: costs rise more slowly than output, and incremental revenue can fall more cleanly to the bottom line. That profile is increasingly attractive in a market that is more sensitive to returns on invested capital than to headline technology spending.
DBS's view also suggests that investors are becoming more selective about the quality of AI exposure. Companies with heavy capital expenditure requirements may still benefit from the theme, but they face a higher hurdle. If the spending does not translate into faster monetisation, better margins or a defensible competitive edge, the market may begin to discount the story. In other words, AI is no longer enough on its own; execution now determines valuation.
Monetisation Over Narratives
The bank's framework points to a crucial shift in how AI is being priced across sectors. In the early days, the market often treated AI as a broad beta trade, lifting everything from chip designers to enterprise software names. DBS is effectively warning that this indiscriminate approach is fading. The winners will be those able to show that AI is not just a cost centre or a strategic talking point, but a direct contributor to earnings quality.
This has implications for corporate strategy as well. Management teams are under pressure to demonstrate clear use cases, measurable efficiency gains and a path to scale. For investors, that means scrutinising whether AI initiatives are embedded in core operations or remain experimental. It also means watching for evidence that AI is supporting pricing power, customer retention or faster product development, rather than simply inflating technology budgets.
DBS's preference for adopters over spenders does not imply a retreat from the broader AI ecosystem. The bank said it continues to maintain exposure to infrastructure enablers, recognising that the build-out of AI still requires chips, servers, networking, cloud capacity and data-centre infrastructure. These businesses remain essential to the ecosystem and may continue to benefit from structural demand. But the bank's emphasis suggests that the most compelling risk-reward may now lie closer to the application layer, where AI can be translated into operating gains.
Infrastructure Still Matters
The message is not that infrastructure is losing relevance. Rather, it is that the market is becoming more discriminating about where the value accrues. Infrastructure enablers remain critical because AI workloads are compute-intensive and require sustained investment in physical and digital capacity. Yet as the cycle matures, investors are increasingly asking whether the strongest returns will come from the companies building the rails or from those using the rails to transform their businesses.
That distinction is particularly important in a macro environment where capital is no longer free and fiscal policy is under pressure in many economies. Companies that can demonstrate AI-led efficiency gains without materially expanding their balance sheets may be better positioned if financing conditions tighten or if investor tolerance for long-dated payoffs diminishes. DBS's preference therefore aligns with a broader market bias toward cash generation, capital discipline and visible earnings conversion.
For India and other growth markets, the implication is significant. AI adoption is likely to be judged not by the size of the technology bill, but by its contribution to productivity in sectors such as financial services, consumer platforms, logistics and enterprise software. The next phase of the AI trade may therefore reward firms that embed the technology quietly and effectively, rather than those that advertise the largest spending plans.
In that sense, DBS's call captures a turning point. The AI story is not ending; it is becoming more selective. The market is moving from enthusiasm for exposure to scrutiny of outcomes, and from capex-heavy ambition to operational proof.