So what happened this summer?

The S&P 500 hit an all-time high before losing -8.2% from mid-July to early August. But stockmarkets recovered unnervingly quickly going into September. With the dust settled, it is worth considering if there are any lessons to be learned.

10 September 2024

During the 2022 Hungarian Grand Prix, Max Verstappen suffered an outright spin on lap 41, conceding the lead, but went on to recover and win the race almost as if nothing had happened. The parallels with equity markets over the summer are not dissimilar. In the first few days of August, the S&P 500 fell -5.8% while the technology-heavy Nasdaq-100 Index fell -7.7%. By the seventh of the month, the S&P 500 Index was -8.2% off its mid-July record high. And then, just as in the Hungarian Grand Prix, the main US equity market went on to recover by adding back +8.8% to 30 August. For investors, it is not sufficient to just shrug and continue without pausing for at least a moment, since markets may (or may not) be trying to communicate something more profound about their prospects. As traders return to their desks after the summer break it is worth reflecting for a moment on this episode before the ‘now’ of the return to work takes over.

There were arguably three main protagonists behind August’s sudden volatility. In turn: technical over-positioning in specific areas of the market, fears over the sustainability of the artificial intelligence (AI) revolution, and concerns around the US economy. Each represents a valid enough reason in theory for a market reassessment both at the time and in the future, so the real question is, whether they were justified or not. Starting with positioning in the market, there is a strong argument to say that the sheer speed of the sell-off pointed to a technical ‘flash crash’-type event rather than an incremental pricing in of a more profound change in the market environment. Specifically, Japan was always going to be a risk from this perspective given that its low interest rates are seized on by hedge funds as a way to finance higher-yielding ‘carry’ trades. Any change in Japan’s potential interest rate outlook, ie higher rates, was bound to hit the carry trade and boost the Japanese yen, with exporter-heavy Japanese equity markets selling off correspondingly. But the argument that potentially higher rates in Japan justified a full global equity sell-off both then and in the future seems weak when one considers that Japanese equities represent just 5% of the MSCI AC World Index and that the Bank of Japan since moved to reassure investors that interest rate normalisation would from now on take care not to upset capital markets. Furthermore, traders may be more wary about sudden reversals in the carry trade going forward and take steps to risk-manage their positions more effectively, leading to less extreme reactions than those seen in early August.

Japanese equities highly sensitive to relative yen strength:

Chart 1: USD/JPY versus MSCI Japan Index (from 3 Sep 2004 to 2 Sep 2024)

Turning to the AI revolution, it is fair to point out that valuations in technology had become somewhat stretched by the end of July, with the tech-heavy Nasdaq-100 Index trading at nearly 29x forward earnings, implying an earnings yield of just 3.4% compared with the risk-free 10-year US Treasury yield of 4.0% at the time. However, stretched valuations alone are not a catalyst for a sell-off (they can get more stretched) and certainly do not mean that that the AI revolution is somehow invalidated. It is true that technology shares – particularly the so-called Magnificent Seven – were hit hard during the early August sell-off but this is primarily because they had been so profitable for so many investors and, combined with their liquidity, made them ideal candidates to sell first in the event of market upset. But look at the fundamentals, and the revolution looks intact. Nvidia supplies the chips that power AI applications and its latest results are revealing, with second quarter revenue coming in at a whopping USD 30 billion, more than doubling from the same period a year ago, as did net income of USD 16.6 billion. These numbers beat Wall Street’s already sky-high expectations and, along with the ambitious investment plans of the AI ‘hyperscalers’ that own and operate data centres, suggest that the supply-side of the AI revolution is at least in very good shape.

Ambitious AI investment plans should drive tech stock revenues:

Chart 2: Capex from the main AI hyperscalers (from 31 Dec 2004 to 31 Dec 2024)

Onto the US economy, and a soggy jobs report in early August showing just 114,000 non-farm payroll additions appeared to raise fears of recession and found itself included in many commentators’ rationales for the market volatility. Likewise, evidence was growing that US consumers were starting to become more careful. But these were scarcely evidence of a pronounced and certain economic slowdown. The National Bureau of Economic Research (NBER) looks at a combination of payroll employment, industrial production and real income before declaring recession conditions. These were all consistently slowing in the months preceding the technical recessions of 1990 (early 90s recession), 2001 (dot.com bubble) and 2008 (global financial crisis). But the three months up to July 2024 by contrast has been far more upbeat than these historic slowdowns. Payrolls have been growing, industrial production stalled in July but had been expanding in May and June while real disposable incomes grew in all of May, June and July. If there is a slowdown in the works it is likely to be a soft one since the overall picture of US economic health appears sound for now.

It would be too breezy to pass off early August’s sell-off as mere ‘summer volatility’, given that US valuations – particularly in technology – are not cheap, the Japanese carry trade is perhaps less obvious than it was, and US economic data is becoming a little more mixed around the edges after recent strength. Combine the above with lower levels of liquidity and juniors manning trading desks as their bosses were away, and the conditions for a temporary air pocket in market progress were optimal. Reassuringly, fundamentals do seem sound for now. And if they were to deteriorate, for example export-crushing appreciation in the Japanese yen, AI bottlenecks around electricity supply or a far more pronounced US consumer slowdown, then there still remains the prospect of interest rate cuts in the US starting mid-September. Federal Reserve chair Jay Powell came as close to a promise as central bankers can get when he declared in late August that “the time has come for policy to adjust”. In the meantime, investors might reflect on that Verstappen spin – unnerving at the time, but carrying on regardless eventually paid off.