Demographic Implications for Capital Markets Euromoney Conference 2005 Rome, September 6 Stefan Schneider Chief International Economist
Structure Demographic trends and their drivers Population dynamics and economic activity Implications for capital markets The rate of return, effects on portfolios
Trend1: Population growth slowing Population Change % 1960-2005 2005-2050 160 140 120 100 80 60 40 20 0 Global phenomenon: global annual population growth to fall from currently 1.2% to 0.4% p.a. in 2050. By 2050 the population will shrink 0.4% p.a. in Europe, grow 0.2% and 0.4% in Asia and North America, respectively. Only Africa will still grow by 1.2% p.a (half of the 2000 rate). Germany France UK EU15 Japan USA Less developed countries Source: Eurostat, UN -20
Trend 2: Global ageing Median age 2050 Years 55 50 Median age (dividing the population in two equal halves) will rise above 45 years in many developed countries. 2005 45 40 35 In Germany one out of three persons will be above 60 years of age. 30 25 Largest relative increase will take place in less developed countries. Germany France UK Japan USA Less developed countries 20 Source: UN
Driver 1: Falling fertility rates Total fertility rate Children per woman 2.6 2.4 Global fertility rate dropped from 5 (early fifties) to 2.6 (2000). Reproduction rate 2.2 2.0 1.8 1.6 1.4 Reasons: - Education, - demographic/economic paradox, - mismatch problems (male/female education). Germany France UK Japan USA Less developed countries Source: Eurostat, UN 1.2 1.0 In its medium variant, the UN expects the global fertility rate to fall to 2.1 (by 2050).
Driver 2: Increasing life expectancy Life expectancy: A linear trend Life expectancy in years Life expectancy at birth: female 2005 2050 Years 95 90 85 80 75 70 65 60 Source: Max Planck Institute for Demographic Research Germany France UK Japan USA Less developed countries 55
Falling birth rates and increasing life expectancy leads to unfavourable age structure EU15: Ageing causes doubling of the dependency ratio EU-15 working age population to peak in 2012. 0.55 0.50 0.45 Population aged 15-64 270 260 250 The number of 15 to 64-yearolds will fall by 40m, or 15%, until 2050. 0.40 0.35 0.30 240 230 The ratio of >65 years to 15-64 years will rise from 1:4 to 1:2 by 2040. 0.25 220 0.20 0.15 Ratio of the population aged 65 years or over to the population aged 15-64 1972 1979 1986 1993 2000 2007 2014 2021 2028 2035 2042 2049 210 200
A number of factors determine the impact of demographics on the rate of return Life-cycle saving & portfolio decisions (supply) Population ageing & pension reform (structural) Capital markets, productivity & growth (demand) Supply side: - Saving behaviour of households - Risk aversion of households Demand side: - Labour force participation (capital/labour ratio) - Productivity (natural rate) - Demand of the public sector Structural factors: - Set-up of the pension system - Degree of international capital mobility - Political and economic stability in the emerging markets None of these factors is exogenous!
Disclaimer: Old data used to analyse a new phenomenon No historic example of an ageing and shrinking population. Economic theory gives only partial, sometimes contradictory answers (demand for capital: relative prices versus substitution effect). Models are calibrated with historic data, but agents will adjust their behaviour to demographic changes (Lucas critique). This applies especially to governments (retirement age). Usage of cross-sectional data for longitudinal analysis confuses age-, time- and cohort effects.
Supply side: The life-cycle hypothesis (1) Income, consumption, savings during the life cycle 225 200 175 150 125 100 75 50 25 0-25 -50-75 Income (right) Monetary units Consumption (right) Financial assets (left) Savings (right) Monetary units 20 30 40 50 60 80 150 125 100 75 50 25 0-25 -50 Life-cycle hypothesis postulates consumption smoothing over life-time. It suggests that savings will rise until 2015-2020. Microeconomic studies find at best lower savings rate after retirement, but no dissaving. Explanation: pension wealth (PAYG) not correctly measured, precautionary saving because of higher life expectancy.
Supply side: The life-cycle hypothesis (2) Economic estimates of the impact of population structure on the household savings rate Coefficients student-t Dependent variable: Household saving rate Control variables: Budget deficit 1-0.632-7.5 Real interest rate 2 0.321 3.4 GDP per capita 8.58 10-6 4.4 Inflation 3 0.193 1.9 Share of population 25-59 4 0.778 2.8 Share of population 60-99 4-4.267-7.0 Share of public health expenditures 5-0.327-3.9 Interaction pension replacement rates & prime-age population -0.015-4.1 Interaction pension replacement rates & old-age population 0.053 5.1 Life expectancy at birth -0.400-0.9 Time trend R 2 w ithin (country-fixed effects) 0.52 Observations Period Countries 1) Cyclically adjusted government net lending as a % of potential GDP 2) Real short-term interest rate 3) Based on CPI indices 4) In total population 5) Public expenditure on health as a % of total expenditure on health Source: OECD calculations 254 1970-2003 30 German savings rate to fall after 2010 Long-term savings divided by disposable income, pension remains at 70% 2000 10 20 30 40 50 Source: Börsch-Supan, Ludwig, Winter, 2003 After 2010-15 the savings rate will fall in OECD countries, causing interest rates (ceteris paribus) to rise. 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00
Demand side: Dependency ratio & potential growth EU 15: Labour force & growth potential 1.2 % yoy, 5Y mov. avg. % yoy 3.1 Lower GDP growth means lower growth of investment spending. 1.0 0.8 0.6 0.4 0.2 0.0 Population aged 15 to 64 years (left) Assumptions: 1) constant labour force participation rate 2) constant retirement age 3) constant grow th contribution of capital and total factor productivity 2.9 2.7 2.5 2.3 2.1 As capital becomes more abundant compared with scarce labour, the return on capital will fall. The effect will be even stronger in the event of pension reforms. -0.2-0.4-0.6 1. & 2. oil price shock Potential GDP (right) 1.9 1.7 1.5 Demand of an ageing society probably more geared towards less capital intensive services. -0.8 1975 1982 1989 1996 2003 2010 2017 2024 2031 2038 2045 Source: Eurostat, EU Commission 1.3 Demand side points to lower yields bottom line?
Bottom line: Interest rates likely to fall Overlapping generation model: Rate of return Simulations show that total return could drop 120 bp, assuming no pension reform and no international capital mobility. Assuming pension reform and no capital mobility (lower curve), the decline would be even bigger. Assuming investment in With pension reform But with capital mobility (top curve) the reduction can be limited to less than 100 bp. Source: Aging and global capital markets, A. Börsch-Supan
Asset demand and portfolio composition German households portfolios of different age groups 700 600 500 400 300 200 100 EUR `000 <25 25-35 35-45 45-55 55-65 65-70 70-80 80< Source: Einkommens- und Verbrauchsstichprobe Life insurance Other investments Equity funds Bonds Equities Other assets with banks Savings accounts Building socities Studies find little demographic effects on asset returns. Some evidence that price/ dividend ratio in equity markets is higher given a large share of prime savers (40 60 years) Little evidence that households actually disinvest as they age (if time effects are taken into account). Not surprisingly, life insurance and savings with building societies show largest age variation. No structural shifts evident between safe and risky assets.
What might change? Lower returns and lower replacement rates in the public pension system will increase need to save further. But postponing the legal retirement age will reduce the need to save. Savers could then resort to (safe) wage income for longer, allowing for more risky assets in their portfolios. This could reduce the equity risk premium (but time effect of the 2000/01 burst of the equity market bubble). Need for duration matching of life insurance companies will increase demand for ultra-long bonds.
Thank you! stefan-b.schneider@db.com