Autores
Barry Gill Paul Marsh

Given his unique access to one the most impressive databases of asset class returns, we interviewed Paul Marsh about his views on diversification and asset correlation trends over time.

Every year for the last 25 years finance academics Paul Marsh, Mike Staunton and Elroy Dimson have cleared their diaries, loaded up reams of asset class data, and started analyzing and writing. The result has become an annual market returns bible that industry practitioners far and wide use to inform their thinking and asset allocation calls. Originating in partnership with Credit Suisse, the final report is now known as the UBS Global Investment Returns Yearbook.

We decided to ask Paul Marsh what diversification lessons he has gleaned from such rigorous and disciplined annual research over the years.

Barry Gill

Can you provide a brief history of financial diversification?

Professor Marsh

The intuition that diversification reduces risk goes back centuries. The expression “don’t put all your eggs in one basket” can be traced back to a quote from Don Quixote in 1605 but may have been commonplace before then.

The scientific measurement of diversification has a shorter history. Just over seventy years ago, Harry Markowitz (1952) published “Portfolio Selection,” which laid the foundations of modern portfolio theory and won him a Nobel Prize. He showed that a portfolio’s risk is not defined by the average riskiness of its individual assets, but by the extent to which the returns on those assets are correlated or move together.

Stocks or assets whose returns tend to be negatively correlated with each other are the most valuable diversifiers, but they tend to be the exception. Those with zero or low correlations are also good diversifiers. However, as long as assets returns are less than perfectly correlated (i.e., a correlation below one), then diversification can reduce risk.

Markowitz demonstrated that diversification reduces risk so that investors can earn the same return with lower risk, or a higher return for the same risk. He argued that “diversification is the only free lunch in finance.” Investors were urged to diversify across stocks, countries and asset classes.

The benefits of global diversification were demonstrated by a French academic, Bruno Solnik who published an influential article 50 years ago entitled, “Why not diversify internationally rather than domestically”. At that time, cross-border investment was the exception rather than the rule.

This had not always been the case. At the start of the 20th century, there was extensive cross-border investment. Over the 20th century, there was a U-shaped pattern of globalization, with international investment commonplace in both 1900 and 2000. During the period in between, from the First World War through to the 1970s, many barriers and costs inhibited cross-border investment.

Since the 1970s, these barriers and costs have been progressively swept aside. Following the 1971 breakdown of the Bretton Woods system of fixed exchange rates, most major currencies have floated freely, removing the risk of sudden large devaluations. While investors still face exchange rate risk, currency, interest rate and equity market risk can now all be hedged cheaply. Barriers to international capital movement have been substantially dismantled. Information is rapidly and widely available and in ever greater volume. Accounting, tax, governance, trading and issuance systems are being harmonized. Low-fee passive vehicles including ETFs abound, facilitating cheap global diversification.


In terms of equities, what is an optimal number of stocks to hold in order to achieve effective levels of diversification?

Almost every finance textbook contains a chart plotting risk (measured by the standard deviation of the portfolio) against the number of stocks held, and showing how rapidly diversification across individual stocks reduces risk. Conventional wisdom is that a small number of stocks – say 10 to 20 – is sufficient to provide market-mimicking returns. In the 2022 Yearbook, we showed that while these charts are correct, they are misleading. That is because they focus on the portfolio’s standard deviation, rather than its residual risk or tracking error. Many more stocks are needed to create a well-diversified portfolio that tracks or mimics the market. For the US market, even with 100 stocks, we showed that the tracking error is still 3.3% per annum.

This does not detract from the fact that diversification across stocks is a highly effective way of reducing risk. In most countries, the risk of a well-diversified equity portfolio – say of an index fund for that country – is around half that of a typical individual stock.


What do you make of Henrik Bessembinder’s research which finds that only a small minority of stocks drive market returns over time – and that the majority of stocks perform worse than cash? What, if anything, does this tell us about active versus passive investing?

Bessembinder’s research is interesting and persuasive. In an initial paper published in 2018, he shows that the majority of US stocks (57.4%) have had lifetime buy-and hold returns below that on Treasury bills. Since 1926, the best-performing 4% of companies explain the net gain for the entire US stock market. In a subsequent paper published in 2021, he examined some 64,000 stocks from 42 countries and showed that the same pattern held for non-US stocks.

This is caused by the strong positive skewness in individual stock returns. The nice thing about investing in a stock is that you can’t lose more than 100%. However, you can achieve returns of well over 100%, or even 1000% or more. This positive skew in returns means that the average return on a portfolio of stocks is well above the median return. The positive premium over bills that we observe for overall stock markets is driven by very large returns for relatively few stocks.

There is much evidence that individual/private investors typically hold concentrated portfolios containing relatively few stocks. Bessimbinder’s results imply that the average individual with a concentrated portfolio is thus likely to receive less than the return on the overall market.

What does this tell us about active versus passive investing? Both camps claim it supports their case. Active managers argue that their skills are needed to find the relatively few stocks that really make a difference. Passive managers argue that you need a highly diversified portfolio that tracks the market in order to harvest the market risk premium. Both are correct, but the choice between active and passive hinges on whether the active managers have the requisite skills to find those big winners and can outperform on an after fees basis.


Is global diversification always a good idea?

Solnik showed that you can reduce risk by diversifying across countries. He urged investors around the world, particularly US investors, to diversify globally. However, although before the event, global diversification is always a good idea, it does not always have a good outcome.

In the 2022 Yearbook, we compared domestic versus global investment for investors in 32 countries from 1974 – the date of Solnik’s paper – onwards. For each country, we examined the increase in the Sharpe ratio (reward to risk ratio) from investing in the world index, compared with investing domestically in the investor’s home market.

Global investment led to higher Sharpe ratios than domestic investment in the vast majority of countries. However, there were a few exceptions, and one of these was the world’s largest and most important market, the US. US investors would have been better off staying at home. With hindsight, following Solnik’s advice proved a costly mistake.

For a US investor, domestic investment beat global investment for two reasons. First, US equities performed exceptionally well. Over this period, US stocks beat non-US stocks by around 2% per year. My co-authors and I have documented this continuing outperformance of US equities and describe it as a case of “American exceptionalism.”

Second, global diversification failed to lower volatility for US investors. The US equity market was among the world’s least volatile as its size, scope and breadth ensured that it was highly diversified. Over this period, the average volatility of non-US countries in the world index was almost double that of the US market. US investors had less to gain from risk reduction than their foreign counterparts.

This is a cautionary tale. It is a reminder that investing is subject to considerable uncertainty. Good investment decisions, based on sensible criteria, can sometimes have disappointing outcomes.

Prospectively, and without the benefit of hindsight, the case for global diversification remains compelling. Our advice to investors from all countries, including the US, is that they should diversify globally. This is very likely to reduce risk and increase the Sharpe ratio, but it is important to recognize that this is not guaranteed.


How have correlations and hence the benefits of global diversification changed over time?

In the 2022 Yearbook, we looked at how the average correlation between the returns on all pairs of countries had changed over the period since 1970. We found that the average correlation between developed markets more than doubled from 0.37 in the early 1970s to 0.75 in the most recent period examined. The corresponding increase for emerging markets was from a very low base of 0.05 to 0.49.

These increases have coincided with and been driven by the removal of barriers, and by the increased globalization of economies and markets. Ironically, with correlations now at a higher level unmatched in the past, this has reduced the potential gains from diversification.

Despite this, some sources still cite quite high potential gains, based perhaps on old data or unrealistic assumptions. Our estimates suggest that investors around the world can now expect a more modest, but still useful level of risk reduction from global diversification.


Much is made of the 60:40 equity bond portfolio. What lessons can we draw from your datasets and decades of research into asset prices in terms of stock-bond diversification?

There is nothing magic about the 60:40 portfolio. The appropriate mix will differ across investors, depending on their investment objectives and risk preferences. However, holding both stocks and bond does provide substantial risk reduction. We show in the Yearbook that the downside risk of a blended stock/bond portfolio is substantially lower than that of an all-equity, or even an all-bond portfolio.

There are two reasons for this. First, bonds are less volatile than equities. Second, bonds are imperfectly correlated with stocks. For example, measured over the entire 20th century, the correlation between US stocks and bonds was just 0.19. This positive, but low correlation, provided good scope for diversification between stocks and bonds.

Negative correlations are even better. Over the period from 2000-21, the US stock-bond correlation was −0.29, which was highly unusual by historical standards. The US was not alone, and stock-bond correlations were negative in most major world markets over this period. This negative correlation meant that stocks and bonds served as a hedge for each other, enabling investors to increase stock allocations while still satisfying a portfolio risk budget.

All good things come to an end. By the start of 2022, we had entered a different environment. Inflation had risen. Monetary policy had shifted from ultra-loose to an interest rate hiking cycle. Real interest rates were now rising sharply, rather than falling. As we said at the time in the Yearbook, “we therefore do not recommend placing reliance on a continuation of negative stock-bond correlations.” Indeed, in 2022, stocks and bonds plunged together, resulting in a drawdown for a 60:40 equity-bond portfolio of more than 30%.

This was a timely reminder that diversification should be thought of as a long-run strategy. In the short run, especially during crises, it can let investors down. However, while negative stock-bond correlations may be a feature of the past, looking ahead, we expect, based on history, that the stock-bond correlation will remain low. If so, it clearly continues to make sense for investors to diversify across stocks and bonds.


You just mentioned that diversifications fails when investors need it most, i.e., in times of crisis. What can we learn from history here?

There is extensive evidence from multiple researchers that correlations between stocks and between countries tend to rise quite sharply during times of crisis – everything tends to fall at the same time – thereby making diversification less effective. The most recent examples are the Global Financial Crisis, COVID-19 and the bear market in 2022.

The cost of elevated correlations depends on their duration. There is evidence that the higher correlations arising from crises are quite short-lived. The extent to which domestic and international diversification can fail investors in a crisis is thus limited to quite short intervals, and then they matter only if they coincide with the timing of realizations where the investor is effectively a forced seller. For long-term investors, the enhanced correlations are of less consequence.


To hedge, or not to hedge? Do you have a view on currency hedging?

Hedging exchange risk appears attractive as it reduces one element of the risk of cross-border investment. However, it is not that simple and the investor’s time horizon also matters. In the short-run, exchange rates can vary greatly, and the changes are impactful. While hedging exchange rate risk reduces short-term volatility, our research for the Yearbook has found that its benefits have shrunk in more recent periods. For equity investments, the risk reduction from hedging equities is less than half that obtainable from global diversification.

For long-run investors, hedging is less necessary and may even be counterproductive. Our research shows that although currencies have been volatile, over the long run, parity changes were largely responding to relative inflation rates. This means that long-run investors are already protected to some extent from currency risk.

Conceptually, hedging involves taking a short position in foreign interest rates and a long position in the investor’s domestic interest rate. While helping to hedge short-term currency risk, this introduces a new form of risk and source of volatility, namely taking a position in real interest rates at home versus abroad. Our research has found that, over longer horizons, hedging can on average lead to an increase in the overall volatility of real returns and thus prove counterproductive.


Is there anything interesting you have realized about diversification that we have not touched on here?

There are just two issues that I’d like to mention. First, there is the question of increasing stock market concentration. The outstanding historical performance of US stocks has led to the US dominating world markets, with US equities now accounting for more than 60% of the world’s investable free-float capitalization. At the same time, many stock markets around the world have themselves become more concentrated. The largest three US stocks now make up 16% of the market’s value, while the top ten account for 29%. This is the highest level of concentration since 1966. This increased concentration provides new challenges for investors seeking to diversify domestically and globally.

Second, diversification should not be seen as a goal in its own right, to be pursued at all costs. The ultimate aim is good risk-adjusted performance, and viewed from that perspective, it is possible to be over-diversified.

Overdiversification occurs when an investor or fund manager has information or insights that are not being fully exploited because the portfolio is too diversified. You cannot beat the market by holding it. To beat the market, you need the skill to generate alpha – sometimes known as excess return or abnormal return or return above the benchmark. Assuming you have such skill, you need to take large-enough positions to exploit it. Otherwise, diversification becomes “diworsification”.

The Red Thread: The Diversification Edition

The craft of correlations

S-11/24 NAMT-1891

The Red Thread The Diversification Edition

The art of harnessing asset correlations

External to UBS

Portfolio of Gaia Vince

Professor Paul Marsh

Within London Business School he has been Chair of the Finance area, Deputy Principal, Faculty Dean and both an ex officio and elected Governor of the School. He has advised on several public enquiries; was previously Chairman of Aberforth Smaller Companies Trust, and a non-executive director of M&G Group and Majedie Investments. He has acted as a consultant to a wide range of financial institutions and companies. Dr Marsh has published articles in Journal of Business, Journal of Finance, Journal of Financial Economics, Journal of Portfolio Management, Harvard Business Review, and many other journals. With Elroy Dimson, he co-designed the FTSE 100-Share Index and the Deutsche Numis Indices for the UK stock market. Together with Elroy Dimson and Mike Staunton, he wrote the influential investment book, “Triumph of the Optimists”. They have also produced the Global Investment Returns Yearbook annually since 2000, which is now published by UBS.

Related insights

Contact us

Make an inquiry

Fill in an inquiry form and leave your details – we’ll be back in touch.

Introducing our leadership team

Meet the members of the team responsible for UBS Asset Management’s strategic direction.

Find our offices

We’re closer than you think, find out here.