Dimson et al have estimated that from 1900 to 2011 the following (geometric) ERPs are realized per country:
1. Denmark: 2%
2. Australia: 5.9%
3. Netherlands: 3.5%
4. U.S.: 4.4%
5. World: 3.8%
Source: Dimson, Marsh and Staunton 2011.
Academic estimates range from 0 to 7% ERP (!) And in practice valuation advisors since 2008 also advise 8% as top of the range based on ex-ante model (implied ERP’s)
Earlier this year the Economist made the following ERP calculation based on the Campbell/Shiller’s Dividend Yield Model:
Dividend Yield + Real Dividend Growth rate + / – Any change in the Price / Dividend Ratio – Risk-free interest rate
Historically (1900-2011) this resulted in:
1. World dividend yield (1900-2011): 4.1%
2. Real dividend growth: 0.8%
3. Rerating of the market: + 0.4%
4. Real Equity Return: 5.4%
5. Risk-free rate long term: 1.9%
6. ERP = 3.5%
Interestingly, the dividend growth has been historically lower than the GDP growth:
This is because companies issued net 2% per year of new shares, hence dividends per share grew merely by 0.8%.
What does this say about the dividend growth assumed in the terminal value of the dividend discount model? Do we correct enough for dilution of dividends? I am sure investors in European banks have felt this effect on there returns recently.
What does this say about future ERPs?
1. In the early 20th century there was no widely available business information, deep market investment portfolios of institutions (pension funds, insurers) or retail investors and diversification, causing low valuations (with high ERP’s as a result).
This implies future ERP’s should be per definition lower.
2. Other arguments are: a net outflow from equities making the P/D market rerating a significant downward pressure on future ERP’s. See McKinsey’s analysis: Growth and Stability in a new investor’s landscape.
3. Of course this also implies that at some point in the future (after the market rerating has fully occurred due to the funding outflow) you will have higher future ERP’s again like in 1900.
4. World deleveraging has only just begun (just moved from private to public) so if disposable income decreases (due to gov. cutting back) so do savings, investments and dividends. According to another McKinsey analyses deleveraging can take up to 7-50 years. See McKinsey’s analysis debt and deleveraging: uneven progress to growth.
5. Many of us look for ERP’s at US data which has been the most successful economy of the last 100 yrs. Whereas an investment in 1900 in Russian or German equities would have had catastrophic results.
Ok I think I can make my case for lower future ERP’s contrary to what the implied ERP models indicate.
An Ex-Ante ERP calculation according to the dividend yield method results in:
• Dividend yield now: 2.7%
• Dividend growth rate: 0.2%
• Expected market valuation adjustment (P / D ratio) is downward because current P / D ratio is higher than historical: (37x vs. 24x)
• 2.7% + 0.2% – 1.4% – 2% = a negative ERP?
Numbers are from mid March 2012, copied from the economist.com. P/D ratios have come somewhat down since.
I do not think I can make a case for people receiving less than risk free for taking risk.
This stands in stark contrast to the ex-post and ex-ante ERP’s proposed by advisors (5% -8%) to correct for the current low risk free rate.
All credits due to the economist.com for this analyses. I just reproduced it for discussion purposes.
Your thoughts please…
- Rethinking the equity risk premium (CFA Research publication 2012)
- Economist.com: How much should people get paid for investing in the stock market?
Updated link: Explanation of the ERP by Buttonwood at Economist.com