Abstract
The financial accounts of the household sector within the system of national accounts report the aggregate asset holdings and liabilities of all households within a country. In principle, when household wealth surveys are explicitly designed to be representative of all households, aggregating these microdata should correspond to the macro-aggregates. In practice, however, differences are large. We first discuss conceptual and generic differences between those two sources of data. Thereafter, we investigate missing top tail observation from wealth surveys as a source of discrepancy. By fitting a Pareto distribution to the upper tail, we provide an estimate of how much of the gap between the micro- and macrodata is caused by the underestimation of the top tail of the wealth distribution. Conceptual and generic differences, as well as missing top tail observations, explain part of the gap between financial accounts and survey aggregates.