A neural network model that processes input data consisting of financial ratios is developed to predict the financial health of thrift institutions. The network's ability to discriminate between healthy and failed institutions is compared to a traditional statistical model. The differences and similiarities in the two modelling approaches are discussed. The neural network, which uses the same financial data, requires fewer assumptions, achieves a higher degree of prediction accuracy, and is more robust.

The majority of studies examining the impact of group homes on neighborhood property values have found that group homes do not adversely effect property values. In our study of seven group homes neighborhoods in DuPage County, Illinois, we find that properties which are proximate to group homes experience a decline in value following the announcement of a group home's pending establishment. In our analysis, observations across time and space are incorporated into a format that is similar to an event study. Our model is the first in this literature to accommodate different price levels and appreciation rates across neighborhoods. (JEL R20)

No abstract