2003). Both tables reflect average population mortality, have a maximum age of 100
years, and are subdivided for males and females.
As proxy for the risky asset, we use the return of a broad-based stock market index.
Stock market returns are assumed to be lognormal and i.i.d. (see, e.g., Hull, 2005).
For the United States, we use data from 1926 to 2006 from Morningstar (see
Morningstar, 2007). After deducting typical transaction costs of an index-investment
fund of 0.7% per annum, the mean of Rt is given by 1.1151, and the standard
deviation is 0.1996. For Germany, we use 1955-2006 data provided by Professor R.
Stehle, Ph.D., Chair of Banking and Stock Exchanges, Humboldt-Universitat zu
Berlin (Germany), which give mean 1.1264 and standard deviation 0.2792 for Rt
after assuming identical transaction costs.
For the risk-free asset return, the short-term money market is used as a proxy. Given
the same sample periods, the Rf is set to 1.0361 per annum for the United States (see
Morningstar, 2007) and to 1.0472 for Germany (see IMF International Financial
Statistics Online database, http://ifs.apdi.net/imf), again after assuming typical
transaction costs of an index-based investment of 0.18%.
For inflation, we use the same sample period, resulting in a value of 0.031 for the
United States (see Morningstar, 2007) and 0.0279 for Germany (see Federal
Statistical Office, 2007).
The labor income process is calibrated to match empirically observed life-cycle
income profiles. Ideally, we would like to have estimates that reflect the labor
income process as given by data used also in the regression analysis later. But our
data are cross-sectional and thus not well suited for this. Panel data that include both
detailed longitudinal information on labor income and asset allocation for longer
sample periods are, especially for Germany, not available. Thus, we decided to take
income profiles from the literature that uses panel data.
11