In reply to Nevis-the-cat and Trangia:
Thanks for the constructive debate. Just to be clear, I'm an investment banker (Boo! Hiss!) so the logic and mechanics of discounting, risk free rates, risk premia, all that stuff are second nature to me. It's the freehold reversion bit I am struggling with - here's why:
In my example, in 100 years' time, F (the freeholder) gets its hands on a building which *today* is worth £10m. However, all other things being equal, in 100 years' time, the building will not be worth £10m, but (say property prices rise by, ooh, 2.9% in real terms (Nationwide BS data from 1975), the building will be worth in round numbers £174m in today's money.
Now, in reality the building will be a crumbling wreck by then, but in theory it could still be pulled down and rebuilt. Assuming the same sale value of the finished block of £174m, and on a developer's rule of thumb 1/3 land cost, 1/3 building cost, 1/3 profit, we already own the land so it would cost us 1/3 of £174m (say £54m for round numbers) leaving us with a "worth" of £120m in today's money in 100 years' time...and *not* the £10m (i.e. today's actual value) that the usual formulae would use. Feel free to pick holes in these numbers but the logic is surely correct?
Now, I understand that when buying the freehold, the valuation is constrained by various legislative requirement (don't ask me why!), principally the requirement to use today's value of the building as the value at the end of the lease. However, as I noted this is not a sale of a freehold, but a voluntary and commercial transaction by the entity which already owns the freehold...so I can't see why we wouldn't use the "correct" (?!) value for the asset in 100 years' time?
Now, all of that is fine, but may well produce a result which is unattractive to the buyer, in which case we're down to good old-fashioned bargaining...but that's for another day.
Further answers on a postcard welcome...