Here are my two cents,
If the domain has no value other than it's traffic (For example, CheapFlights.com has a lot of value other than it's traffic. It's a lot more brandable than any available domain names), then it comes down to risk VS reward.
A US government bond for 10 years all but gurantees you 5% interest per year. The risk is none-existent, and if the US government falls I'm guessing so will domain name value. It will pay you no less and no more than 5% every year for 10 years.
Now let's take a domain name which made $50 in the past year. If you buy it for $1,000 and the revenue remains the same for the next ten years, you will make the same amount of money as you would with the US goverment bond.
However, the revenue from the domain is not a sure thing. It can go up, and it can go down. PPC prices do indeed go up, more people connect to the Internet and it stands to reason that if nothing unexpected happens the domain will make more money every year.
However, "if nothing unexpected happens" is the key factor here. Ten years is a long time, perhaps some new .com regulation will exist in five years that will alter the the value of domains. Perhaps someone will hack into your godaddy account and steal all your domains (It is more likely to happen then someone stealing a US government bond from your bank, which is ofcourse government insured).
Now it comes down to how you asses the risks that revenue drop and the chances that it will rise. If everyone had a degree in economics, it's likely that this alone will determine the value of domains and assets in general. But people usually don't think this way, they go with intuition and sometime actually have personal feelings for a domain. (I registered MyBobo.com because I like to say "Bobo" at random occasions).
Personally, I noramally wouldn't pay more than two years revenue for a Sedo parked domain unless it had value other than it's current traffic. The main reason is that it's very dificult to get data about the traffic sources a sedo domain is receiving. HitBot abusing aside, if traffic is coming from expired links for example it is likely to fade off. If it's coming from bookmarks it is also likely to fade off. Only if it's coming from just random type-in then it's likely to stick and even rise.
Because it's very dificult to get this information, and the resources one needs to spend in researching it may themselves make the purchase not economic (You'd spend so much time researching domains that when you do find one that is worth purchasing it wasn't worth the total amount of time you spend finding it).
These were some long cents, hope someone reads this through... :cheeky: