Greenhorn it is! I'll still mention the real names occasionally just to make the emails more searchable.
Interesting thoughts.
Brighthouse and Greenhorn do reference "Adjusted Earnings" which at least equalizes the quarterly change in liabilities. But I take your point on EPS being a bad measure. And it's quite lumpy anyway. To me it seems like book value is the better measure since that's what Greenhorn emphasizes. I also did my back of the envelope DCF model. It uses Revenue and Net Income margin to calculate Net Income payback over time. It also looks absurdly cheap on that basis. It looks to me like the entire enterprise value will be earned back in just under 3 years if present Revenue and Net income levels are maintained. This is also seen in the P/E being ~3X. I get that earnings may not be the best measure, but would you disregard a P/E of 3 completely?
For annuities being a bad business, how do you reconcile that against the idea of insurance being a good business for Berkshire? Presumably insurance generates float for Buffett to use whereas that's not the case here. But if we take float out of the equation, does that otherwise make insurance and annuities both equally bad businesses? Or there is a reason to prefer insurance? Or am I missing something when you say annuities are a bad business? (You probably know more than I do in this regard).
Ultimately to me the book value discount coupled with potential for accelerating buybacks is the key positive dynamic. And then I just have a lot of trust I place in Greenhorn to do deep dive fundamental analysis a million times better than I could. Generally, my edge (assuming I have one) is critical thinking, and even though this is not "first principles" it seems much more effective for me apply my critical lens to Greenhorn rather than directly to Brighthouse, if that makes sense.