Recent thoughts on Treasury Market Liquidity

I have a recent working paper that considers how to look at Treasury on-off spreads.  The paper uses a decomposition approach.  First regress on-the-run and off-the-run yields on a set of macro factors, and then study the difference in the coefficients.  I argue that it is not the spread itself which captures the liquidity of Treasury markets, but rather it is the difference in the coefficients of macro factors on bond yields.  The paper will be made available on SSRN.