Man! - you guys, cooking up an actuarial shit storm! Great work! I had no idea this was so contentious. I’ll have to write a post.
ERN’s method is fine. Assuming the pension increases with retirement then give all the pension payments 2% per year increases and then discount by the nominal rate (not inflation adjusted). Or a neat actuarial trick is to take the nominal pension values and discount by the nominal rate less the inflation of 2% (the real rate). And don’t forget to delay payment by 8 years by further discounting by 8 years.
For the discount rate I think you want something that reflects the security of the promise. So risk-free Treasury rate is too low, and equity return is too high. It is common in the pension industry to use a long AA bond. Which yields about 4% at the moment (nominal). That reflects the fact that the promise is pretty strong and the pension is actually worth something to you. So your net rate might be around 2%. Crazy low huh! That’s what low real rates do to pension valuations!
Don’t use a perpetuity, that is not realistic and will give you too high a number. If I valued my clients’ pensions with a perpetuity their costs would skyrocket and I would get fired!
BTW the following comments are all true.