With your access to the data, could you possibly provide us with as estimate for the distribution function of SL residents' wealth? With the number of registered accounts, I'm assuming it will be roughly Pareto distributed. A maximum-likelihood estimate of the Pareto index would be fantastic. And, hey, while we're at it, confidence intervals on the value would be cool too....

This information would help give us more context for the total wealth, sources, and sinks figures posted by Robin.
Thanks a bunch!