http://scriabinop23.blogspot.com/2008/01/yield-curve-inversion-recession.html take a look. basically, I went through all the past yield curve inversions in the 45 yrs and the results are pretty clear. 6/7 of the total inversions led to recession. all led to economic slowdown. discuss.
its the benchmark evaluator of the yield spread inversion ... shortest versus long maturity. Perhaps also because the 30 yr bonds haven't been consistently floated ?? Not sure why the 30 yr is not used as often as the 10.
The benchmark I see most often is the 2-10-yr, at least at: http://www.briefing.com/Investor/Public/MarketSnapshot/BondMarketUpdate.htm Wonder what your results would be if you tried the two year data instead of the three month, or any other combinations along the curve: http://federalreserve.gov/releases/h15/data.htm
OK. No spreadsheet for you yet, but the 2 yr vs 10 yr is a bit noiser than my methodology. Starting with 1976 to present data, generally the 2 yr inverted against the 10 yr several (4-10 or so) weeks before the t-bill makes the move. The inversions for every example ended at about the same time (t-bill vs 10 yr, vs 2yr vs ten yr). 1982 had some inversion even though recession was formally over (in the 2 yr data). The t-bill dataset had none of this noise. The 1998 and 1990 data had quite a few more weeks of two year inversion vs the t-bill/10yr data. Again, though, the inversion was light and the data was generally wobbly. [lot of crossing the flat line] An informal glance reveals the 2yr vs the 10yr reveals practically the same thing vs. the tbill-10yr spread, but with more noise and an earlier and longer inversion. The tbill spread reveals clearer data and is a better indicator. If you want i can send you the data in my spreadsheets.