Can the economy hold on until AI pays off?

Reading time: 5 minutes

A year dominated by two themes

This year, the economy and financial markets have been dominated by two themes: the Gulf conflict, and AI. The Gulf conflict has transmitted a negative supply-side shock to the global economic and financial market system via a rise in oil prices, while AI has delivered a positive demand-side shock through capex expenditure and consumer spending driven, in part, by a boost to wealth. The Gulf conflict has been a negative for economic growth, but has largely been offset by the positive growth impulse from AI. However, both shocks are a positive for inflation.

The Gulf conflict is the swing variable 

Regular readers of the Brief will be aware of our three Gulf War scenarios: the Benign, the Malign and the Malignant (Economic impact of the Gulf war). Following the Memorandum of Understanding between the US and Iran in early June, inflation and the economy largely tracked our Benign scenario as oil prices fell below US$80/bbl for Brent Crude one-month futures. After a spike in the oil price in the second half of July, oil prices were still below US$80/bbl by early August. But this was to change.

Since early August, oil prices have climbed steadily, to be over US$100/bbl. With the conflict escalating and with no clear path for resolution, we are now at risk of shifting from the Benign scenario to the Malignant scenario, where the oil price remains above US$100/bbl well into 2027 and beyond. As a result of the persistently high oil prices, we have seen inflation reverse direction across the major global economies. In countries like the US and Australia, core inflation rates are accelerating, with annual rates well above central bank targets.

The change in direction of inflation has caused central banks to change direction of monetary policy. The US Federal Reserve (Fed) and European Central Bank (ECB) have been forced to end their easing cycles and raise interest rates. The Reserve Bank of Australia (RBA) and the Bank of Japan (BoJ) have had to continue with their tightening cycles, while the Bank of Canada (BoC) and the Bank of England (BoE) have been forced to abandon an anticipated easing of policy rates. More tellingly, all the aforementioned central banks are expected to tighten policy over the coming 12 months.

Where are rates headed?

But the more worrying factor is where rates are headed. For example, the RBA Cash Rate is headed for 5% by May next year according to market pricing, up from 4.35% currently. The US fed funds rate is headed for 4.8% by July 2027, up from 3.875%. The ECB policy rate is to reach 3.40% by July 2027, up from 2.5%, with the market expecting comparable increases for most other major central banks.

These levels of central bank policy rates are consistent with our Malignant scenario, not our Benign scenario. Importantly, the rise in interest rates since the start of the war have placed significant upward pressure on the cost of capital. For example, the inflation adjusted real 10-year US Treasury yield, often thought of as the risk-free rate of investment, has lifted by 0.9 percentage points (ppts) since early March. To place this increase into context, if such an increase in the cost of debt were to be absorbed by the US stock market without offsetting increases in earnings expectations or risk appetite, the likely outcome would be around a 15% fall in equity values.

AI the saviour?

To date, AI has supported global stock markets and economies. The performance of the US market over the year has been nothing short of astonishing. Year-to-date, the US market (as measured by the S&P500), has risen by 12.6% with earnings per share (as measured by IBES) up by around double the rise in the equity price, offsetting the drag on valuations caused by the rise in interest rates and the cost of capital. Importantly, the earnings outlook remains robust, with the outlook for the coming twelve months for a further 20% earnings uplift. Can AI continue to deliver? That is to be determined, but a question we can ask is, “how much does AI need to deliver?” A rough back-of-the-envelope calculation would suggest that the impact of a 1ppt increase in the real risk-free interest rate could be offset by a (permanent) 1ppt increase in real earnings growth.

Can AI deliver such a lift? The promise of AI is to permanently lift the economy’s growth rate in productivity, which in turn would lift the economy’s real rate of growth, and hence, the real growth rate in earnings. We don’t know by how much AI technology can lift productivity and growth, but we can point to historical precedents. History shows that there is historical precedent of technology delivering sustained increases in productivity growth of around 1ppts.

That precedent occurred over the 20th century and is known as the “one big wave” in the academic literature associated with Robert Gordon. Gordon argues that technological innovations (mainly over the first half of the 20th century), such as electricity, indoor plumbing and sanitation, running water, the internal combustion engine, railways and automobiles, telecommunications, modern chemicals and household appliances, led to a lift in US productivity growth of about 1ppt point over the 50 years from 1920 to 1970. This is the hurdle that AI must clear if the earnings growth promised is to materialise. It is the hurdle the global economy must clear if AI is to continue to offset the ongoing fragmentation of the global economic order.