When the AI music stops, what's the playbook?

For investors who like to dance when the music is playing, the current AI trade has been a once-in-a-generation Woodstock. It’s been a great party, but no-one likes a hangover.

Thus far, we have been happy to lean into the trade.

Our thesis has been relatively simple. So long as the AI capex boom is cashflow funded and not debt funded, the market’s earnings generation appears sustainable. Indeed, the earnings growth has been extraordinary. Well reported elsewhere, the current earnings upgrade cycle is remarkable outside of periods immediately following a recession. Share markets reflect this.

Everyone loves a free lunch

Call us old school, but diversification remains one of the sharpest risk management tools available. Whenever returns become concentrated in a single stock, sector or thematic, we blunt the benefits of diversification.

With that in mind, here are three areas we are looking to diversify within an increasingly concentrated global market:

Small Caps

Small caps are perhaps the surprise package of the last 12 months. Global smalls have delivered higher returns than the AI-dominated global index, but what we really like is the way the returns have been delivered.

When we examine the spread of sectors contributing to small-cap index returns, the lack of concentration in Technology is striking. Inherently this presents a more diversified return profile that is less reliant on a single investment theme.

Furthermore, valuations remain more attractive, with multiples closer to historic averages rather than trading at an AI-premium.

Our thesis on adding smalls to the portfolio earlier this year was based on the view that economic growth would continue to broaden across the US and Europe. Some of that strength is undoubtedly linked to AI-related capital spending and the spillover into local economies, but it also reflects a broader improvement in economic activity.

We continue to believe that the users of the AI, in many cases the real economy industrials, have more to gain from the application of the technology than the builders of the AI infrastructure. The hyperscalers, in contrast, have seen their business models become more capital-intensive seemingly overnight. Small cap exposures align well with this view.

Drugs & baked beans

Neither drugs nor baked beans sound particularly exciting when the market is focused on chips, silicon and chatbots.

One defining feature of the current market has been the growing dispersion between the AI haves and have-nots. Capital has flowed aggressively towards AI beneficiaries, leaving many industries behind despite delivering solid earnings growth and maintaining more stable, defensive revenue streams.

Healthcare and Consumer Staples are two notable examples. These sectors also happen to be important ingredients within the Quality factor, which has likewise underperformed in recent years.

When looking up the playbook for ‘investing after AI’ (unfort it doesn't yet exist) it may make sense to start with how some of these out of favour and unpopular industries have traded since chipmaker Broadcom disappointed the market last week, triggering the latest bout of market volatility.

Infrastructure

Beyond AI, the other elephant in the investment room has been the steady return of inflation and interest rates alongside this. Infrastructure offers several attractive characteristics in both environments.

• AI earnings tailwind – Datacentres need power. Regardless of whether hyperscalers ultimately earn attractive returns on their AI investments, the US remains structurally short both power generation and the gas infrastructure to support it. The tailwinds to earnings, particularly for regulated asset exposures, are material.

• Inflation protection – Many infrastructure assets benefit from regulated asset bases or CPI-linked revenue streams, allowing inflation to be passed through via linked pricing mechanisms. Should the world move further into a higher-for-longer interest rate environment, revenues may be partially protected. In contrast, the multiples of many expensive technology shares may find this environment far less rosy, as investors experienced in 2022.

Scars and lessons learned

Many of us still carry scars from the last technology bubble in 2000, when valuations collapsed and there were no earnings to support share prices. This time around feels different, and we have drawn comfort from the dazzling earnings growth being delivered by the current AI boom. 

It’s been an amazing time for the cap-weighted Indexes.

We all enjoy the occasional free lunch, but diversification remains at the top of the menu. While the AI music continues to play, we have elected to spread our bets more broadly. Increased allocations to Small Caps, Real Assets and Quality businesses (Healthcare/Staples) continue to screen attractively and offer a compelling source of diversification in today’s market.