lizzieinvancouver / decsens

0 stars 0 forks source link

try sims with declining variance? #7

Closed lizzieinvancouver closed 4 years ago

lizzieinvancouver commented 4 years ago

Would need to do enough to see when it leads to declining sensitivities on the scale that we see.

lizzieinvancouver commented 4 years ago

@AileneKane Switch to 20 year windows.

lizzieinvancouver commented 4 years ago

To do after seeing PEP725 results.

lizzieinvancouver commented 4 years ago

@lizzieinvancouver Maybe? Based on the PEP725 data, varmat for 1950-1970, then 1970-1990, then 1990-2010:

> mean.betpen
         cc  matslope matslopelog  meanmat    varmat  meangdd matslopeconfint11
1 1950-1970 -6.053967  -0.4157587 7.910215 1.2324478 72.47753         -8.045821
2 1970-1990 -6.843966  -0.4895361 8.120306 0.8399633 72.15105        -10.293221
3 1990-2010 -2.115403  -0.1773680 9.037770 0.8897746 60.03246         -4.456350

So it does go down! And a lot, 25%, but the sensitivity does not change over that time....

lizzieinvancouver commented 4 years ago

I think issue #9 makes clear what would happen here.