Asset Allocation and TAA

The Controversy over Long-Term Equity Return Forecasts

Looking Into The Future - Long Term Expected Return Forecast Equities

In a recent article for the CFA, I discussed the results from four major asset managers that publish their long-term return forecasts. Three of the four institutions share a broadly common view on the expected returns from the major asset classes. On average, they see long-term expected returns (from a US dollar perspective) as shown below. As we move into riskier assets, the expected return increases. Nothing controversial so far...

Future long-term return expectations:

     Asset Class

     Annual Return

     US inflation

     2% - 2.5%

     US cash

     2% - 2.5%

     Commodities

     2% - 3%

     Hedge funds

     4% - 5%

     Global equities

     6% - 8%

     Private equity

     8% - 9%

Source: Aggregated from JP Morgan, Northern Trust, BNY Mellon (see papers below). 

However, the estimate of long-term US equity returns from Research Affiliates is distinctly different from those above. The California-based team estimates 10-year US equity returns at only 1.0% pa – a huge difference!

Whom should investors believe? Key factors that could explain the difference include assumptions about equity valuations, share issuance / buybacks, and long-run earnings growth / profit share of GDP. What returns might investors reasonably expect? Let’s unpack this in more detail.

Reasonable expectations of long-term equity returns

Let's go back to first principles, and employ a classic formula to provide a starting point for our test of reasonableness. We'll assume that company earnings grow in line with expected GDP growth. We then have the following:

LT term equity return = DY + Real (in excess of inflation) Profit Growth + Inflation

LT term equity return = 2.0 + 2.25 + 2.25 = 6.50%

So, other things being equal, a return of 6-7% p.a. for US equities is perfectly reasonable.

But, are other things equal?

Adjusting for stock buybacks and stock issuance

When stocks are bought back, the number of issued shares reduces. EPS growth increases, and long-term returns increase. When new shares are issued, the converse is true. EPS growth is reduced, and LT returns are negatively impacted.

You may have noticed that the “classic” formula above uses real earnings growth, rather than the more obvious option of using real dividends growth. This is largely to avoid the complex question of how to adjust for stock buybacks and stock issuances. 

A recent paper from Research Affiliates, Stock Buybacks - Oasis or Mirage? argues that, in the US, the level of stock issuances is larger than stock buybacks. Therefore this will be a drag on long-term returns.

This subject is also covered in the JP Morgan paper referenced below, albeit with different conclusions.

Profit share of GDP

Can we safely assume that earnings growth will match the growth of the real economy?  Will the profit share of GDP be maintained at current levels? A recent report from McKinsey, "The new global competition for corporate profits" argues that the 30-year expansion in corporate earnings growth is set to come to an abrupt halt. They project that global corporate profits, currently almost 10% of world GDP, could fall to less than 8 percent by 2025.

Can large companies hold their profit share?

Even if the profit share of GDP holds at current levels, can the large corporations which currently constitute the index hold onto their own profit share? Undoubtedly, new technologies and new players will emerge, just as they have done over the last 30 years. But what difference will this make?

Equity market valuation

The “classic” formula used above is based on the assumption that the dividend yield holds steady at 2%. If it is going to change, we need to make a valuation adjustment. This is the key reason that the estimate from Research Affiliates diverges so substantially from the herd. Research Affiliates argue that the US market is overvalued, and assume a 50% reversion to the long-term mean valuation, based on their measures. This accounts for a return reduction of almost 2.5% pa on their figures.

Demographic trends and sales by the baby boomer generation

Apart from assuming a random reversion to the mean, there may be other structural factors which could drive equity market valuations in the coming years. Absolute Return Partners recently authored a paper Structural Trends in Equity Market Valuation which covers this in some detail. The paper looks at the structural factors that drive equity prices, including corporate profits as a share of GDP, dividends and share buy-backs, demographics, interest rates and valuations.

Recommended Reading (PDFs)

The papers below, from JP Morgan, Research Affiliates, Northern Trust and BNY Mellon, summarise the long-term return expectations for major asset classes:

Long Term Capital Market Return Assumptions (JP Morgan) 2016

This detailed 90-page document contains the thinking behind JP Morgan's long-term outlook for all major asset classes and alternative strategy classes. The document is divided into 3 sections: a) thematic trends and issues, b) rationale and methodology, and c) assumptions.

​Forecasting long-term equity returns (Research Affiliates) 2015

This paper sets out the approach taken by Research Affiliates for forecasting long-term equity market returns. Other papers in the series cover other asset classes.

Five-Year Return Forecasts for Major Asset Classes (Northern Trust) 2015

This study has a shorter time frame and therefore depends less upon quantitative considerations. The authors assume a continuation of slow global growth, resulting from high indebtedness, de-leveraging, demographic factors and slowing emerging markets.

​10-Year Capital Market Return Assumptions (BNY Mellon) 2015

This annual report from BNY Mellon covers 50 global asset classes and is designed to inform the decision-making process for strategic asset allocation.

For further reading see our full list of asset allocation white papers

Feedback