2) Ingrid Ellen showed that REO properties produce crime. The interesting part--they induce violent crime, rather than property crime. The data set she put together, which used block faces, instead of census blocks, was awesome.
3) Len Lin may have solved why real estate appears to have better Sharpe Ratios than stocks. If they were really better, one could arbitrage between real estate and stocks. But because real estate is illiquid--it takes a long time to sell it--one cannot arbitrage it. When one adjusts formally for illiquidity, the Sharpe Ratio of real estate is the same as stocks.
4) Stephanie Yates Rauterkus presented some promising work that suggested that "walkability" near CBDs enhances value, but elsewhere might not.
It was a really good few days. I learned a lot of other stuff too.