September Investment Views
Creative Destruction and Dodging Disruption
- New technologies drive economic progress
- Innovation can be disruptive for established businesses
- The market is continually assessing the winners and losers
Over the last two years, we have written extensively about Artificial Intelligence (AI). The theme of AI has been a key driver of markets since the launch of ChatGPT in late 2022. The data centre build-out is also showing up clearly in economic growth data. The AI opportunity helps to explain the apparent disconnect between a global economy that is muddling through a range of different challenges, and buoyant stock markets that continue to make new all-time highs.
A lot of AI coverage has focused on the potential for the new technology to drive efficiencies across a range of different industries. AI has the ability to simulate judgement by analysing data, recognising patterns, and generating creative content in word, graphical, and video format. We are at relatively early stages of this revolution, and the market has largely focussed on the early potential winners in the semiconductor, data centre, and Large Language Model parts of the process.
Many applications will be built on these new AI models, and large technology firms are spending huge amounts of money in the race to be one of the dominant AI platform providers. The market has been more circumspect when it comes to the software layer of AI. We have recently seen the market focus on the software industry and discriminate between companies which will be beneficiaries and those which will be disrupted by new competitors. AI’s ability to generate code makes it easier to develop new software applications. Incumbents in the sector enjoy healthy margins, so these new applications present a threat to their subscription-based revenue model.
Large software packages that have a very wide installed base in large companies are likely going to be difficult to dislodge, especially if they are integrated with existing processes and other systems. There is a thriving ecosystem of venture capital, concentrated in Silicon Valley, which is funding fast moving disruptive businesses. These present a threat to existing software applications without a competitive moat, which is a risk for investors. Leading research and advisory company Gartner has seen their stock price fall materially on worries that AI will compete with their data analytics and research generations capabilities.
Back in the early 1940s, economist Joseph Schumpeter coined the term “creative destruction”. He described this as a natural process in a capitalist economy where innovation constantly disrupts and replaces old industries, products, and ways of doing things. This process is a threat to established firms as it can make their products obsolete. Such disruption can be a painful adjustment, but Schumpeter observed that it was an essential fact of capitalism and the engine of economic progress. Creative destruction provides important context for investors when assessing which industries and companies will emerge as relative winners and losers from AI.
Historically, the winners in a new technological revolution are rarely the same as the previous technological revolution. Large companies often find it harder to innovate compared to smaller, more nimble businesses. So far, the AI revolution has bucked this trend, with the large technology companies being seen as winners from AI, as they provide the computing power and own, or have relationships with, companies that supply the underlying model platforms.
Alphabet is an interesting case study, as it was initially feared that AI products, such as ChatGPT were an existential threat to Google Search. For example, instead of users going to Google, they can use an AI application to search, analyse, then shortlist products and services. This could be much more efficient than reviewing numerous links individually provided by Google. However, Google has been quick to build AI functionality into Google Search, while also rolling out their own Gemini AI platform. The market has become increasingly comfortable that Google is innovating sufficiently to maintain their dominance by benefitting from AI rather than being disrupted by it.
AI start-ups are not merely aiming to disrupt existing technology businesses; they want to also go after other areas of company budgets. Replacing people in organisations would open huge new revenue opportunities, and even capital expenditure savings on buildings. For example, better chatbots could mean less need for call centres. The scale of the potential revenue opportunity helps explain why the market is comfortable with very high capital investment into AI.
We are still in the early stages of the AI revolution and there are a lot of unknowns in terms of beneficiaries and those who are negatively impacted. The market is comfortable that the current large technology companies will be winners in the new AI world, and this is understandable until evidence proves otherwise. Apple is a laggard when it comes to AI innovation, but they have an impressive hardware platform; people will still need these devices to run AI applications. There is a lot of political and geopolitical noise in the market right now, but under the surface, the theme of winners and losers from AI is an area where we are spending an increasing amount of time.
Fixed Income
Employment Cracks, Bond Market Rallies
- Fed cut odds surge; curve steepens on soft jobs.
- Powell flags labour concerns, lifting bonds and risk assets
- Dollar weakens; gold rallies, oil prices slide
Risk assets continued to grind higher in August but bond markets captured the clearest momentum as US policy expectations shifted sharply toward easing. The most consequential development came from the US labour market, where payrolls slowed dramatically. These cracks in the employment picture cemented expectations that the Federal Reserve will cut rates in September, with debate focused on whether the move would be 25 or 50 basis points. US Treasury markets responded decisively: two-year yields fell by more than 30 bps, its lowest since spring. The curve steepened meaningfully as short-end yields collapsed, reflecting imminent easing, while long yields held steadier amid persistent fiscal supply concerns. By month-end, the probability of a September Fed rate cut had surged from just 40% to nearly 90%, underscoring how quickly sentiment shifted. The Bureau of Labor Statistics’ annual benchmark employment revision, due in September, is expected to be particularly influential. Should those revisions confirm an even weaker underlying jobs market, this could prove decisive in the Fed’s policy deliberations for the remainder of 2025.
The Fed was not alone in leaning toward easier policy. The Reserve Bank of Australia and the Reserve Bank of New Zealand each lowered their policy rates by 25 bps, and the Bank of England trimmed its base rate by the same amount to 4.0% despite still lingering inflation risks. While the Fed remained anchored, the willingness of other developed market central banks to act added weight to the fixed income rally and underpinned broader risk appetite. Credit markets absorbed these moves with enthusiasm. Investment-grade corporates saw their busiest primary calendar in years, with nearly $100 billion in new issuance readily digested, reflecting confidence that easier monetary conditions will help contain default risks. Spreads tightened across both IG and HY, while agency mortgage-backed securities delivered standout returns as demand for spread product intensified.
Beyond the US, European yields also moved lower, though UK gilts lagged on fiscal concerns ahead of the November budget and French bonds were impacted by political instability. Meanwhile, Chinese assets attracted significant inflows, even as economic data remained weak, helping broader emerging markets record positive returns supported by a weaker dollar. The dollar’s decline - down 2% on a trade-weighted basis - lent further support to foreign borrowers and underpinned flows into risk assets. Gold rallied nearly 6% to fresh highs, while oil prices fell sharply, with WTI down over 7% as uneven global demand weighed. The renminbi, yen, euro, and pound all gained against the dollar, while the Canadian dollar also strengthened but to a lesser degree as the economic outlook weakened.
The annual Jackson Hole symposium reinforced the dovish sentiment. Fed Chair Powell relieved some political pressure, striking a more cautious tone, acknowledging that labour market weakness has become a pressing concern. While consumer spending and services activity have shown resilience, it is the slowing in US employment that increasingly appears to be providing the monetary policy offset, giving the Fed cover to move toward easing without losing credibility on inflation. Powell’s remarks were the final confirmation investors needed to price an imminent rate cut.
Portfolio positioning has shifted to include additional corporate credit exposure, reflecting the rising prospects of Federal Reserve easing over the next 12 months, with 125 bps of cuts now priced in. This, combined with still-buoyant nominal growth expectations for both this year and 2026, raises the risk that an overly aggressive policy response could fuel a risk-on bubble. To hedge against an unexpected slowdown in US growth, we remain overweight longer-dated bonds, which would likely rally in the event of macro weakness. This positioning also offers protection in the opposite scenario: if US growth rebounds, the Fed would be unlikely to deliver the full extent of cuts currently priced in, leaving long-duration exposure well supported. In addition, we see value in UK gilts, where speculation about an IMF bailout is far-fetched, and where the Bank of England is likely to reduce the scale of its bond sales to provide stability at the long end of the curve.
Equities
Software Slips with Technology Stocks Lagging the Broader Market
- Fifth consecutive month of positive returns
- Earnings beat as FX tailwinds and margins boost earnings
- Software lags as the death of software narrative takes hold
The MSCI World index recorded the fifth consecutive month of positive returns, adding 2.6% in August. The returns have slowed since the rally in May and June, with the summer months resulting in less volatility and more muted returns. The second quarter earnings season was significantly better than expected. The S&P 500 earnings grew by 11% year-on-year, surpassing the 4% growth projected by analysts. 58% of companies raised their 2025 guidance, double the number that did so in the first quarter. Part of the beat came from the weaker dollar, as international sales account for 28% of S&P 500 revenues. Companies also benefitted from more resilient margins. After “Liberation Day” analysts cut earnings projections due to tariffs, yet the impact of these new policies have not yet materialised. Despite all the upgrades to estimates, the S&P 500 earnings growth year-on-year is still expected to decelerate from 11% in the second quarter to 7% in the second half of 2025, reflecting the impact of tariffs. The real question is, will companies be able to pass on the higher costs to consumers or will they take the hit, resulting in margin compression. The answer is probably a combination of both.
The best performing sectors this month were Materials and Healthcare, returning 7.2% and 5.2% respectively. Despite the strong performance, Healthcare continues to be the worst performing sector year-to-date. We reduced our position earlier this year to neutral, however the consensus position is overweight, a trade that has disappointed investors for a number of years now.
The worst performing sectors in August were Utilities (down 0.5%), Information Technology (up 0.4%), and Industrials (up 0.6%). Industrials and IT are sectors that have benefitted from the build-out of data centres and AI. IT stumbled this month due to weakness in the software sector.
The death of software narrative took hold of the markets in August with a broad sell-off in some of the software names. The fear is that AI is making coding easier for companies to develop their own software and that AI led coding will eat the software pie. This narrative gained traction with the launch of ChatGPT-5 this month. The commoditising of programming has been a longstanding debate, with every advancement bringing us closer to automation. Investors appear to believe that software companies’ moats are vulnerable. Stretched valuations for some of these stocks leaves them exposed. Software names like Microsoft and Oracle are off the hook as investors are more focused on their growing cloud businesses, however the other software names were not spared.
Industry specialists disagree with this narrative, with some believing this disruption might actually be a positive, with companies increasing spending on software as the cost benefits multiply. The most logical pushback against the death of software narrative appears to be that the value in writing software is in the planning, design, and optimising. The actual writing of the software is where the least amount of value lies. At the end of the day, you still need smart software people to solve problems and optimise operations. We also came across a recent study that showed that coders using AI took 20% longer to complete tasks than unassisted coders, with the checking and rewriting of code prolonging the task. The softness in software is likely to create some buying opportunities especially for names that were lagging already.