on hand ratio constant for measuring the risky asset share, cash on hand also has to
decrease. Consequently, from the perspective of this lowered amount of cash on hand
the overall discounted value of future labor income is relatively larger leading to an
increase in the risky asset share. Note that in reality, i.e., in a situation where more
factors than age will change at the same time, one would expect the individual to
jump at retirement from one line to another, reflecting a drop in the labor income-to-
cash on hand ratio.
Figure 2: Investment into the risky asset as a function in age for German (EVS)
data, Gender = 0 (male), γ = 2, δ = 0.97, Education = Middle
--- put Figure 2 here ---
The effect of education on the risky asset share in our model is solely driven by
differences in expected future labor income. Higher education is associated with
higher growth rates of labor income during the life cycle, resulting in higher
expected labor income (for a given current income), leading to a higher risky asset
share.
Gender enters the benchmark model at several points. Women have a higher life
expectancy and (for the German calibration) lower expected growth rates for future
labor income than men (assuming for the moment no differential in relative risk
aversion). Ignoring the rather small effect of higher survival probability of women,
their risky asset share should be smaller than the share of men. Assuming further that
women are more risk averse (higher γ), we further expect a smaller risky asset share.
The effect of capital market conditions is highlighted by comparing U.S. and German
capital market data in Table 1. U.S. and German investors can expect a similar equity
risk premium (nominal around 8%; in real terms around 5%), but at a considerably
smaller standard deviation. Consequently, U.S. individuals should invest a higher
share into risky assets. For example, an American CRRA-investor with zero labor
15