How Do Household Characteristics Determine the Levels of Saving Deposits in the Euro Area?

The sum of real and financial assets less debt from loans.

Dummies: single (MS_S), married (MS_M), in consensual union (MS_CU). Basis for comparison: widowed and divorced.

Dummies: doing regular work for pay (LS_RW), retiree (LS_R). Basis for comparison: the remaining labour statuses.

Dummies: lower secondary (E_LS), upper secondary (E_US), tertiary (E_T). Basis for comparison: primary or below.

Dummies referring to the household’s expectations regarding its future incomes (in next 12 months): they will increase more than prices (I_M), they will increase less than prices (I_L). Basis for comparison: they will remain at the same as prices.

Dummies: purchase of home (S_P), other major purchases (S_OP), invest in financial assets (S_FA), provision for unexpected events (S_UE), paying off debts (S_PD), old-age provision (S_OA), travels/holidays (S_TH), education or support of children and grandchildren (S_ES), bequest (S_B), taking advantage of state subsidies (S_SS).

Dummies: not willing to take any risk (R_N), take average financial risks expecting to earn average returns (R_A). Basis for comparison: take at least above average financial risks to earn at least above average returns.

The Akaike and Schwartz-Bayes information criterions are applied to validate the selection of independent variables for models and compare the goodness-of-fit of the models to empirical data (Gorecki 2010, pp. 40–41). For verification of collinearity of the independent variables, the variance inflation vector is used (Kufel 2013, pp. 64–65). The homoscedasticity of variance is examined by White test (Kufel 2013, pp. 60–61). For all the models, this problem is identified, thus, the structural parameters of the models are estimated according to the Heteroscedasticity Consistent Covariance Matrix (HCCM) (Kufel 2013, pp. 136–137). The significance of the structural parameters of the models is verified by Student’s t-test while the residual normality by Doornik-Hansen test (Kufel 2013, pp. 117–118). It should be explained that due to the use of household-level data the coefficients of determination (R2) are relatively low (Gorecki 2010, p. 40, 43). The primary criterion for the selection of independent variables for the models is their merit, referring to the objective of the study. The formal and statistical criteria regarding two basic assumptions are considered (Kufel 2013, pp. 54–55), i.e. that the independent variables are uncorrelated or poorly correlated in the model and the independent variables are relatively highly correlated with the dependent variable.

They relate to, e.g., options, futures, index certificates, precious metals, oil and gas leases, future proceeds from a lawsuit or real estate that is being settled, and royalties.

For example, a subsidy to building society savings.