Asset Liability Management (ALM) models for housing corporations work on the basis of stochastic scenario models for important risk and return drivers such as interest and inflation rates, construction costs and sales prices of houses. Given the situation of the housing corporation (current real estate portfolio, asset management strategy, loan portfolio, etc.), these variables directly influence the risk and return profile of these organizations at both short and long horizons. The scenario characteristics are usually based on long term historical time series of relevant indices. For sales prices of houses, different indices exist with different specification levels. House price indices are available at both a national and a local level. And different house price indices can be used for different house types (from apartments to detached). In this paper we analyze the impact of the choice of the specification level of house price indices on the risk - return profile of housing corporations. And we draw some conclusions on the optimal specification level.