I stress that this is depending on when you joined the scheme, but the idea is that a member (or dependants) will always get ten years' benefits out of their membership. So, retire and die at the end of year 1, the death grant is nine years' worth. End of year 2, eight years' worth, etc. After ten years it's down to zero. Changing those rules is a Parliament job.
Finding the money to pay the benefits is the employer's risk, not the members. The Fund has to invest to meet the benefits promised. A purely passive strategy would get the bum's rush on the grounds that we need to outperform the market, not simply follow it. Bear in mind the LGPS is not one scheme, it's about one hundred, broadly one for each county.
I should say that the atmosphere when we listen to investment professionals' predictions is distinctly cool. We don't even place much faith in the reported funding level, which comes from the actuaries; within the last year it's gone from 78% funded to 92%, apparently, but none of us would have been remotely surprised if it had gone the other way, and quite a few think it actually did. It'll take so long for the effects of a particular funding level to come out (assuming it never changed again) that we would have died of old age before we could see whether the predictions were true. The approach that has evolved is a significant proportion of active management to get the growth, passive as insurance, and try to identify systematic poor performers as early as possible and get rid of them. You will appreciate I'm probably missing a lot of detail!