Then from there, I click through the 65% #, assuming they have a good study on 65% of people stop after a year. Nah, they don't. It's super complex but tl;dr: specific cohort, and somehow the # getting on it in year 2 is higher than the # of people who quit in year 1.

I have a weak to medium prior, after 10m evaluating, that the entire thing might be built on more sand than it admits.

Lot of little slants that create an absolute tone - ex. multiple payouts over the "lifetime" of a life insurance policy. (sure, it's technically possible)

Also there's no citation for the idea this mortality slippage happened because of GLP-1, and it's been out for...what...a year? Maybe two?

That's an awful lot of people who were about to die, saved in the nick of time by...losing weight? Again, possible, I'm sure it even happened in some cases.

Enough to skew mortality slippage from 5.3% to 15.3%?

I thought they were 98% accurate?

Wait...is the slippage graph net life increase slippage? Or any slippage?

Because it's very strange this explosion happened in exactly the year of a global pandemic that had sky-high mortality rates for older people.