I enjoy looking at markets in a much longer term perspective. I feel this is particularly important as we have passed an inflection point. The last 40-odd years have been characterised by falling bond yields and rising valuations. The future is likely to look rather different.
When I was at the hedge funds, I never bothered to read the Credit Suisse Global Investment Returns Yearbook, now renamed the UBS yearbook, compiled by respected academics Dimson, Marsh and Staunton and in its 25th year.
My reasons were I didn’t have time and it was too historical - I was stupid. I now read the published extracts with interest, and this article is based on those. UBS have kindly supplied me with a review copy, and I shall report more fully on that in a future Substack.
Before getting into it, a word from this week’s sponsor:
The upcoming London Value Investor Conference on 15th May features renowned Value Investors from the UK and around the world – including Joel Greenblatt, Peter Davies, Gary Channon, Sarah Ketterer, Ian Lance, Alex Roepers, Simon Gergel, Cole Smead, Brian McCormick, Clive Gillmore, Vera German, Juan Torres, Django Davidson, Jacob Mitchell and others. If you’re interested to attend next month, you can visit the conference website by clicking the logo above and use the code “Behindthebalancesheet24” for a discount of £360 until April 17th. If you’ve not heard of the LVIC before, head to the homepage to watch the overview video of the 2023 event – which featured David Einhorn and Howard Marks: https://www.valueinvest.com/london/
In my article on the yearbook last year, I said that a benefit of looking far into the past can show how much things change. For example, the US wasn’t the largest market in the world in 1900 - that title belonged to the UK. Today, the US market is home to 60% of the world’s capitalisation, and the UK is under 5%.
Stockmarket Changes in Last 125 Years
Source: Elroy, Marsh and Staunton, 2024 UBS Global Investment Returns Yearbook
This year, they also included industry weightings for both the US and UK. For simplicity, I have only shown the US data.
Industry Weightings in the US – Change in the Last 125 Years
Source: Elroy, Marsh and Staunton, 2024 UBS Global Investment Returns Yearbook
By 1900, railroads, the go-go technology stocks of the 1850s, had become the dominant sector in the US market with 63% of the total market capitalisation. I don’t know how many stocks there were then, but I would imagine by this point there had been a significant degree of consolidation.
How many times have you heard commentators question the concentration of today’s market in the big tech stocks ? They could conceivably become a larger proportion of the total market!
An aside from Penn Station, New York, as I make my way to D.C. Many of the stations en route are called Penn Station - the old Pennsylvania Railroad built many of these stations back in the early 20th century. Each railroad used a different station in major cities so the station usually took the name of the company.
Oddly, the proportions of other industries haven’t changed as much as you might expect. Banks and Financial services are only slightly larger than they were; utilities are a little smaller; telecoms is rather smaller than telegraph; industrials is slightly larger than iron, coal and steel and other industrial; while there are a few new sectors like health, media and insurance.
Drawdowns
I particularly enjoyed the section on risks and returns. First let’s look at the good news. There have been four golden periods for global stock markets which are shown in the table below:
Best Periods for Returns in US Stock Markets
Source: Elroy, Marsh and Staunton, DMS Database 2024, 2024 UBS Global Investment Returns Yearbook
Unsurprisingly, one of these was the tech boom in the 1990s, which was much better in the US than globally. Global markets were held back by the meltdown of the Japanese bubble which fell from epic valuations. The 1920s was a better period for global markets and again the US delivered even better returns in the period. The other two periods highlighted are the recovery from extremely high rates in the early 1980s and the period of reconstruction post WW II.
I am not clear why the period from 2009 - 2023 does not make it into the table – the nominal gain was massive and inflation surely hasn’t been that bad.
The 1920s continues to stand out as being an era of excess and it feels like the 2020s have started on a similar trajectory.
The authors also show actually start with the worst periods which are unlikely to surprise you: two world wars, the Global Financial Crisis, the fallout from the dot.com boom and the roaring twenties plus the 1970s oil shocks. The magnitude of the falls are much smaller than the booms. And even the 1929 crash looks less scary in real terms.
Worst Periods for Returns in US Stock Markets
Source: Elroy, Marsh and Staunton, DMS Database 2024, 2024 UBS Global Investment Returns Yearbook
The authors look at the distribution of returns in good and bad times with the positive histogram shown in the first chart below:
Real Returns Annually from US Equities
Source: Elroy, Marsh and Staunton, DMS Database 2024, 2024 UBS Global Investment Returns Yearbook
I think it’s more helpful to know the magnitudes beyond the 10 percentile buckets and also to see the averages for each. I prefer my version of this analysis in the chart below but this data set is shorter, at under 100 years.
US Equity Real Returns 1928-2023
Source: Behind the Balance Sheet from Aswath Damodoran NYU Stern Data
Even more interesting is the time taken to recover. The four great bear markets of the last 100 years are shown in the next chart, together with the time taken to recover in real terms. There is an important distinction here – in 1973/4, stocks fell by 48% and only took 26 months to recover the nominal high. In real terms the drawdown was 56% and the recovery took over 10 years.
Time Taken to Recover from a Bear Market
Source: Elroy, Marsh and Staunton, DMS Database 2024, 2024 UBS Global Investment Returns Yearbook
Similarly, for the 1929 crash, the period to recover in nominal terms was I think over 25 years – deflation reduced this to 15 years as shown in the chart. Later recoveries have been faster – this reflects the quick and significant action taken by the authorities post the Global Financial Crisis (and in the Covid-19 pandemic).
The use of real returns is significant, especially when looking at data on a longer timescale – as of the end of 2023, the authors highlight that US stocks were still 10% below the November 2021 high in real terms.
Do Bonds Have More Fun?
Clearly the answer is no! But the yearbook also provides some comparative data on which to assess the relative attractions of equities and bonds. And in my latest podcast, Jonathan Ruffer speculates that we may be entering an age of bonds. Equities have done so well that we tend to forget that bonds were the asset class of choice for many years.
Worst Periods for Bonds
Source: Elroy, Marsh and Staunton, DMS Database 2024, 2024 UBS Global Investment Returns Yearbook
What stands out is how effective bonds were in deflationary times. Note also how attractive returns were from the interest rate high in 1982. Clearly, the last 40 years have been exceptional in interest rates declined for most of the period. I keep saying returns may not be as good without that tailwind. That must be true of bonds, although it’s conceivable that it may not be true of equities.
Real Returns Annually from US Treasuries
Source: Elroy, Marsh and Staunton, DMS Database 2024, 2024 UBS Global Investment Returns Yearbook
It’s striking how much of an outlier 2022 was. And although equities have been better than bonds, bonds have still been more fun than cash.
Corporate Bonds
The last extract from the book covers corporate bonds. I was stunned to read that “Corporate bonds [have] an outstanding value of some USD 44 trillion, almost half that of the value of global equities”. I spoke with the CIO of a massive bond shop and he questioned that number. The US represents 20% of the total global bond market. Here is their breakdown by type of bond in the US, sourced from an industry body:
US Debt Market Breakdown
Source: Elroy, Marsh and Staunton, 2024 UBS Global Investment Returns Yearbook
The authors calculate that there are $10.6tn of corporate bonds in the US and some $45tn globally. The total is dominated by financial issuers which comprise 50% of the total (29% in the US). They also highlight industrials, consumer discretionary, utilities, energy, healthcare and technology.
I am no bond expert but I wasn’t surprised that US financials were over- represented. What surprised me though was that tech features as one of the largest sectors; and the value of debt outstanding is a staggering number. If it’s right, then corporations will have to find at least $1tn per year to fund extra interest payments in coming years.
Country Analysis
The Yearbook also includes some analysis of individual markets including Switzerland, Japan and the US. The authors show annualised real returns and risk premia for the periods 1900-2023, 1974-2023 and 2004-2023, as illustrated in the chart.
US Returns and Risk Premia
Source: Elroy, Marsh and Staunton, 2024 UBS Global Investment Returns Yearbook
The UIS equity market has delivered 6.5% pa real since 1900, the best performance of all the countries they cover. The world ex the US delivered 4.3% pa. That differential is astonishing. The US has delivered a near 50% better return than every other market in the world over 120+ years.
Part of it must be the impact of two world wars which destroyed massive wealth, but left the US stock market relatively unscathed. Part of it must be the US system which has favoured wealth creation.
But it seems less likely that this will be repeatable. I always wonder about the commentary around just buy an S&P 500 ETF and forget it – it has been very right for a very long time, but will it always be correct? I wonder.
Conclusions
And that’s the great benefit of a book like this – it forces you to ask that type of long term question. I am grateful to UBS for publishing the extracts and to Elroy Dimson for allowing me to quote from it and to show you some of these charts.