consumption wealth and business cycles in Germany

Consumption wealth and business cycles in Germany
BRITTA HAMBURG MATHIAS HOFFMANN JOACHIM KELLER
CESIFO WORKING PAPER NO. 1443

This paper studies the long-run relationship between consumption, asset wealth and income in Germany, based on data from 1980 to 2003. While earlier studies — mostly for the Anglo- Saxon economies — have generally documented that departures of these three variables from their common trend signal changes in asset prices, we find that for Germany they predict changes in income. Asset price changes are found to have virtually no effect on consumption — both in the short as well as in the long-run. We offer an explanation of this finding that emphasizes differences between the bank-based German financial system and the rather market-based Anglo-American system: stock ownership by private households is much less widespread in Germany than in the Anglo-Saxon economies and the share of publicly traded equity in household wealth is much smaller in Germany than in the U.S., the UK or Australia.



The fact that pension systems are so fundamentally different in the two countries is also likely to help explain why the share of equity in household portfolios is so much lower in Germany than in the United States. In Germany, both the public as well as most employer-sponsored retirement schemes are financed on a ”pay as you go” (PAYG) basis. Conversely, private mutual funds and pension funds play a much more important role in the Anglo-Saxon economies. In particular the U.S. saw considerable growth in the number of tax-deferrable defined contribution plans such as the 401 (k) throughout the entire 1980s and 90s (Poterba, Venti and Wise (1994, 1998)). As a consequence of the minor role that equity holdings play in Germany, real estate wealth dominates the net wealth position of the German private sector. About two thirds of net asset wealth come under real estate wealth. In the US, by contrast, real housing wealth accounts for only about a third of total net wealth.

It would therefore seem conceivable that temporary fluctuations in cay capture temporary fluctuations in housing wealth. But while residential real estate prices in the US are characterized by long and pronounced swings over our sample period (1980-2003), prices in Germany have remained relatively flat.3 Here too, differences in the financial system, in particular in the profiles of mortgage finance system, may play a major role: e.g., mortgage equity extraction is not used in Germany (as opposed to the US) and according to the BIS study by Tsatsaronis and Zhu (2004), banks’ lending behaviour is more conservative and requires home buyers to provide relatively high levels of collateral for their mortgages. Against this backdrop, it is not surprising that Germany has not seen any pronounced cycles in the residential real estate market and that, therefore, fluctuations of the cay relation can hardly be attributed to real estate wealth.


These facts clearly support the notion that fluctuations in income must be a relatively much more important lever — in fact the only remaining — in explaining fluctuations in the German consumption-wealth ratio. But it would seem that our results suggest that stock market prices do not have a transitory component in German data, even though there are compelling theoretical reasons to believe that stock markets in general should have transitory components that reflect time-varying risk premia. We make the following remarks: first, we do find that the cay residual is statistically significant in predicting excess returns on the German stock market. But in terms of economic significance, the predictive power of cay for asset prices and in particular for equity returns in Germany is negligible relative to what is found in the U.S., UK or Australian data sets. Secondly, this finding is in no way tantamount to saying that stock markets in Germany could not have sizeable transitory components. What we say is that the consumption-wealth ratio of the average German household does not help to identify these transitory components. We report some evidence below that suggests that the U.S. consumption-wealth ratio has considerable forecasting power for German stock returns.


Our finding that household disposable income — largely derived from labour — constitutes the transitory component of the consumption-wealth ratio in German data may seem at odds with the general notion that German labour markets are very rigid and that wages may display only very sluggish adjustment. But note that it is the sum of all household labour incomes that enters the intertemporal budget constraint from which the cay-relationship is derived. This sum is equally determined by temporary fluctuations in the unemployment rate and in hours worked, so that sizeable temporary fluctuations in labour income do not a priori contradict the view that the German labour market is rigid. Indeed, we report below that cay has considerable predictive power for fluctuations in the unemployment rate and other business cycle variables.


In one important respect, our results differ markedly from those reported in the papers by Lettau and Ludvigson for the U.S. and Fernandez-Corugedo et al. and Voss et al. for the UK and Australia: temporary stock market fluctuations have almost no impact on the budget constraint of the average German household, because stocks account for only a minor share of German household net worth. Conversely, the consumption wealth ratio has considerable predictive power for income at business cycle frequencies: in Germany, cay predicts business cycles, not stock market cycles or the prices of other assets.

This paper has studied the link between consumption and wealth in Germany during the period 1980-2003. Very much as earlier studies for other countries, we can identify an empirical approximation of the consumption-wealth ratio as a cointegrating relationship between consumption, asset wealth and income — the cay residual. In keeping with most versions of the permanent income hypothesis, we find that consumption mainly reacts to permanent innovations in asset wealth and income. But whereas earlier studies for the U.S., Australia and the UK have documented that this cointegrating relationship predicts changes in asset prices, in particular risk premia in the stock market, we find that cay mainly predicts income changes in German data. Our explanation for this phenomenon is that stock market wealth accounts for a much smaller share of household net worth in Germany than in the Anglo Saxon economies so that temporary fluctuations in stock markets have only very limited impact on German private household net worth. We have interpreted this observation in the light of well-documented structural differences in the financial and pension systems of continental Europe and the Anglo-Saxon economies.