Return expectations: 2021 edition

The bad news is it’s six months late.

The good news is it’s only six months to wait for the next one.

Just like last year’s edition, this post aims to answer the question of how much UK-based investors can expect from their investments over the next 10 years.

Expected returns

As with last year’s post, I’ve collected 10-year equity return forecasts from as many sources as I could find, and adjusted them to be relevant for sterling-based investors.

For those interested in seeing the underlying data, I’ve included links to all data sources in the ‘Data’ section. For details of all calculations, refer to the ‘Methodology’ section.

Without further ado, the expected 10 year returns from each of the twelve sources are below. Returns are in sterling, are nominal (i.e. haven’t had inflation deducted), are as at 31^st December 2020, and represent the returns for a global market-cap weighted portfolio.

GMO are the clear outliers, predicting -2.1% annualised returns. But that’s not really a surprise given they’re famously bearish every single year. 10 out of 12 sources are predicting between 4% and 6% nominal returns over the next 10 years, so that’s at least a reasonable clustering.

The median result is a 10-year expected return of 5.1%, meaning UK investors can expect around 5% annualised nominal returns from a global equity portfolio over the next 10 years, on average.

It’s always useful to think of expected returns in ranges (you never get “average” returns, especially in the equity markets), and helpfully four of the twelve sources included ranges around their return forecasts. I’ve included them in the chart below, taking the median of the forecast returns at each quartile. You can see the 5.1% median return for 2021 is the median from the previous chart.

Overall, return prospects don’t seem to have changed much since last year for returns.

The median expected return stays exactly the same at 5.1%, with only slight upward revisions to the most extreme scenarios (5^th and 95^th percentiles), and a 1% upward revision to the 1^st quartile expected return from 2.5% to 3.4%.

So the picture is broadly the same – expected returns somewhere around 5% nominal, with returns likely to fall somewhere between 3.5% and 7.5%.

In terms of what that means for more balanced investors, we can infer (with a much higher level of confidence than with equity returns) the returns from government bonds based on their current yields.

With yields on global government bond funds standing at roughly 0.5%, that paints the following picture for varying levels of equity/bond allocations:

Given the low expected returns for both equities and bonds (equities due to high valuations and bonds due to low yields), returns are lower across the board.

Balanced investors with a traditional 60%/40% equity/bond split can expect roughly 3.5% annualised returns over the next 10 years, likely falling somewhere between 2.5% and 4.5%.

Again, these are just estimates, but comparing this to much longer-term historic returns of around 6%-7% on a real basis (see later), the main takeaway is that investors should be tempering their expectations for what they can expect from the equity markets over the next decade.

Digging in to regional exposure, all sources expected higher returns from Emerging Markets and Europe over the US (although this obviously comes with higher risk). High relative valuations in the US are likely to drag returns down for global investors.

Fees

Remember, all these expected returns assume no fees.

I won’t bang on about it again here, because I’m sure everyone’s heard more than enough of it by now. I’ll just say that with returns so low, it’s more important than ever to keep fees low.

As we saw in this post, fees are the best predictor of future returns, and luckily they’re also one of the very few things we have control over. Given returns are likely to be lower going forward, keeping fees low ensures you keep a larger slice of your returns.

How reliable are 10-year return forecasts?

It’s true that the vast majority investment forecasts aren’t worth listening to.

One day I’ll get round to writing up the hundreds of pages of notes I’ve collected on exactly how useless most forecasts are. And the return forecasts in this post should be treated with an equally healthy dose of scepticism.

So why am I writing about them?

Although most forecasts spouted by CNBC pundits, macroeconomists, and other “often-wrong-but-never-in-doubt” gurus have little evidence to support them, there’s a surprisingly large amount of evidence behind the relationship between starting valuations and subsequent 10-year returns. So it’s at least worth considering.

Although equity returns are completely unpredictable over the short term, in the longer term (around 10 years), returns become increasingly predictable as they become a function of starting valuations. Expensive markets with high valuations tend to have lower returns over the next 10 years, and cheaper markets with lower valuations tend to have higher returns.

The details of the relationship between starting valuations and returns will be the subject of a later (multi-part) post, and although the correlation is about as strong as we can hope for in investing, the relationship is still far from perfect. Not only is there considerable variation around expected returns, but attempting to use country/regional valuations as a means of generating alpha has been repeated proven to be futile.

What these return expectations are useful for is giving a good ballpark idea of whether returns are likely to be higher or lower than we’ve seen in the past.

They can serve as a useful starting point for setting expectations around the range of returns your investments are likely to generate. This in turn is useful for setting expectations around how much you’ll need to save, how long you’ll need to work for, and how much future spending can be facilitated by your portfolio.

Long-term return expectations

These returns are relevant for investor looking over the next decade. But for those lucky enough to have multi-decade horizons, the impact of starting valuations is diluted, so longer-term historical average returns are likely to matter more for returns.

Over the very long-term, using one data source from 1870 and one from 1900, real returns for equities are in the region of 5% – 7%.

Investors hoping to achieve these sorts of returns over the next decade are likely to be sorely disappointed. But those with longer, multi-decade, time horizons may well see some reversion to longer-term averages, which are much more positive than for the next decade.

Methodology

• Returns from all sources were converted to a global equity return – weighting US, non-US, and EM returns by MSCI ACWI weights where a general ‘global equity’ return wasn’t provided.
• All real returns were adjusted to nominal returns by adding back any assumed inflation.
• Non-sterling returns were converted to GBP where only USD returns were given, using an average conversion rate backed out from sources where comparable USD and GBP returns were given.
• Each estimated quartile return was calculated by taking the median quartile returns of all sources (NB: not all sources provided data by quartile).

Data

Data for long-term real returns: