3 ms·
> But at that point you understand pensions when fully funded actually cost less than 401k's, assuming similar investments. Because 401k's end up with some dead
by WkndTriathlete 8y ago
> But at that point you understand pensions when fully funded actually cost less than 401k's, assuming similar investments. Because 401k's end up with some dead people having large surplus's and pensions don't.
a) 401(k) and 503(b) plans are expensed and accounted for in the present. You may have the best actuaries in the world but if your current pension-expecting workers beat the odds you're in a world of hurt, even without over-promising and under-funding pension benefits. (No less than Warren Buffett himself famously warned of this problem back in the 1970s, and if there's one thing Buffett knows it's risk assessment and insurance.)
b) 401(k)s are only more expensive than pensions assuming everyone retires at the same age and gets the same benefit on retirement between the two investments. However, as is common, pension retirement age can be as low as age 50. Paying out those pension benefits for those extra 10-15 years (vs. retiring at age 60-65 for 401(k)) probably adds up to paying more for pensions than 401(k)s, especially given that some employees can apparently goose their pension eggs by working overtime in the last few years of employment.
- Retric 8y agoA) The problem is not that this is hard, the problem is their are incentives to get it wrong. If you accept saving to much is a net benefit as it reduces future spending, then you can operate very safely. B) This just comes down to discounting the pension properly or limiting retirement age to 65. In the real world doing pensions correctly has not worked out. My point is simply they have an innate mathematical advantage not that it's directly applicable.