Definitely maybe: Should investors be cautious about history repeating itself?

As we say hello to a new quarter century, it’s a good time reflect on longer-term asset class performance. US equities have had a famously strong run. 

When making investment decisions, however, it’s important to fight behavioural biases such as recency bias, which is where investors are overly influenced by recent events and performance. We believe that analysing performance over a longer history can be helpful in combatting recency bias.

Outside of investing, particularly within culture, people seem to suffer from the opposite of recency bias. Nostalgia is the belief that things were inherently better in the past than they are today. It’s regularly seen when people discuss politics, sport and music. I may be letting nostalgia influence me here, as the last one is something I actually agree with! So let’s try and combat recency bias and analyse US equities from a time when I think music was better: when Oasis released their 1994 debut album Definitely Maybe. 

Over the last five years, investors have had significant challenges to navigate, including the first global pandemic in a hundred years, war in Europe and the biggest inflationary spike in almost 30 years. Over that period, US equities – as measured by the S&P500 Index – have returned an impressive 15% p.a. But how sustainable is that? One thing that might help answer that question is to look at what’s been driving equity returns.

Total equity returns can be broken down into dividends and capital gains, the latter driven by changes in price-to-earnings (P/E) and earnings, which is driven by changes in revenue and profit margins.

Margin expansion changes are often linked to the economic cycle, while multiple expansion occurs when investors become more positive about the future prospects for equities. 

The last five years have seen particularly strong revenue growth and 5.2% p.a. of multiple expansion. Analysing a breakdown of equity return drivers over a longer period might potentially be able to help inform decisions going forward. As shown in the chart below, returns from dividends and revenue growth have been relatively stable drivers over the last thirty years, combining for 7.2% a year. It’s within margin and multiple expansion that we have seen much more volatility. 

Assessing the last five years in the context of the last thirty years, 5.2% p.a. of multiple expansion – while impressive – Is far below the 13.2% p.a. we saw from 1995 to 1999. But the multiple expansion we’ve seen, little by little, is edging up US equity valuations. 

Were we to see the same multiple expansion over the next five years, that would take valuations above the peak of the tech bubble in 2000. Although we might not see the mean reversion we saw from the extremes of 1999, but we shouldn’t be overly influenced by recent performance and expect that the last five years’ performance will be repeated.

We derive longer-term expected returns through our strategic model, though one way a simple baseline return expectation could be established is to take the average dividend and revenue growth with no assumed margin or multiple expansion. That might potentially suggest returns of around 7-8%, which in comparison to US government bond yields of around 4% suggests a risk premium of around 3-4%. We believe this is more consistent with a risk premium one might expect from equities over the long term.

Whichever way you look at it, US equities have had a very strong run, driven by revenue growth and multiple expansion. This has left valuations elevated but nowhere near supersonic levels of the dot-com boom, following which we saw a sharp pullback.

So is it worth being cautious about recent returns from multiple expansion repeating over the next five years? Definitely. 

Could that mean we’ll see mean reversion and a pullback in equity valuations? Maybe.