Global equity markets were particularly volatile in November, driven by factors such as the US Government shutdown, concerns over a potential AI-related bubble and the direction of future US interest rates. The S&P 500 and NASDAQ reached intra-month lows of -4.4% and -6.9%, before recovering to end the month +0.1% and -1.5%, respectively. Returns were muted outside of the US, with the FTSE 100 remaining flat and the Euro Stoxx 50 rising +0.1%.

Similar to the prior month, domestic markets underperformed their international peers. The ASX All Ordinaries Accumulation index fell -2.5%, as markets digested stronger than forecasted economic data. This was largely focused upon two releases, being 1) a stronger than expected jobs report and 2) hotter than anticipated October CPI, showing broad-based inflation across goods and services.

As a result, markets now assume a greater chance of a rate hike rather than a rate cut over the next 12 months.

In bond markets, the US 10-year bond yield declined -6 basis points (bp) to 4.01%, whilst its Australian counterpart rose 22bp to 4.52%. The Australian dollar remained flat, closing at US$0.655.

In the world of AI, a lot can change in six months. Back in July, there was little talk of AI investment in public bond markets. The sector's 'hyperscalers' (Amazon, Google, Meta, Microsoft and Oracle) had been funding their capital expenditure (capex) through strong free cash flow and by tapping private equity, private credit and other sources. This dynamic is changing - fast.

Capex is set to ramp up significantly over the next 24 months. With free cash flow constrained by shareholder returns and share buybacks, hyperscalers are leaning harder on public bond markets.

The last three months alone have seen a glut of debt issuance: $30 billion (bn) of bonds issued from Meta, $25bn from Alphabet (owner of Google), $20bn from Amazon, and $18bn from Oracle.

Major investment needed

The rapid growth of AI-optimised data centres highlights how much investment is needed for businesses to keep pace with the AI race.

Building an AI-specific data centre can cost up to $50bn - some three times more than a conventional facility, depending on the chips involved. Morgan Stanley estimates total data centre funding will hit $3 trillion (trn) by 2028, while JP Morgan and McKinsey put the figure at $5-7trn by 2030.

Oxford Economics suggests the current pace of AI investment since 2023 matches the digital boom in the 1990s, when the internet took flight.

With such major outlays required, capex guidance and forecasts are rising fast. The five hyperscalers are expected to grow capex by 40% in 2026 to $500bn, and by a further 17% to £600bn in 2027.

Powering the future

Increased energy demand is a major implication of AI and this too will require further investment and funding over time.

The International Energy Agency estimates that total data centre electricity consumption will double by 2030. Data centres are essential for AI due to the massive computational power and high-speed networking required to train and run complex AI models. Electricity consumption by AI-optimised servers is expected to increase fivefold by 2030.

BloombergNEF estimates that by 2035 global data centre power needs will hit 1.6TWh, which will take data centre share of global power demand from the current 1.3% to closer to 4.4%. Put another way, by 2030 it is estimated that if data centres were a country, they would be the fourth largest consumer of energy after China, the US and India.

This surge in energy demand adds another layer to the investment challenge - and will likely keep funding needs elevated for years to come.

Bond markets taking notice

Bond markets are sitting up. US investment grade credit spreads have widened by 12% since the end of September - driven in part by a surge in issuance, as markets question how the AI boom will be funded.

According to Morgan Stanley, just under half of the $3trn required by 2028 could come from cash generation, with a quarter from private credit and 10% from other sources, such as private equity and sovereign wealth.

That leaves the remaining 15% roughly $450bn - to be raised in bond markets, with $200bn-250bn potentially coming from investment-grade credit markets. JP Morgan estimates that 14% of the US investment grade debt market is already tied to AI.

In our view, AI-related corporate bond supply will likely continue to grow, both in absolute terms and as a share of the wider bond market. While AI issuance is unlikely to overwhelm the bond market, we do expect further periods of significant issuance from AI-related companies as they invest in their capabilities - potentially causing bouts of indigestion in public credit markets.

What does it mean for investors?

We see the technology sector as one to play tactically.

For strategies focused on shorter-dated bonds, the surge in supply creates opportunities to lock in high-quality names at more attractive levels as markets reprice.

For all-maturity or longer-dated funds, a nimbler approach is needed. This helps control exposure to the wider AI sector and leaves room to add when the next wave of major bond issuances arrives - which we expect in the not-too-distant future.

Final thoughts...

The AI revolution is well underway. The funding to power this new era will be significant. Bond markets are already taking notice as hyperscalers come to market. We expect this trend to continue at pace. Data centres alone will need an estimated $5-7tn by 2030.

For investors, this wall of supply offers opportunities to pick up attractively valued securities ahead of any repricing. Staying nimble and selective will be key as the next chapter of AI-driven growth unfolds.