I will divide domains in the following categories:
1. Good names.
May be worth thousands of USD if you have the right one, e.g. the same name as an existing company that wants to set up a new site.
2. Good sites with high traffic and pagerank.
May also be of value to a (startup) company, since it may drive customers to their site. They can use the site as is or the site has subscribers (existing customer base) to different services that a buyer may profit from.
3. Sites with positive cash flow.
The bigger the better. The value may range from some thousand USD til millions.
Example:
Business.com was sold for about USD 10 million. Not difficult to figure out why.
- Great name.
- Great site.
- Great traffic.
- Positive cash flow.
Advice: Do not assume that there is a free lunch. You may be disappointed. Hard work on your site is my best reccomendation. If you have the right name and sell it for a good price, you can compare it to winning in a lottery. The expected outcome of such a lottery is in my view negative, but never the less some people win. Loosers pay, winners stay.
Expected outcome (E):
p(i) = probability of success.
s(i) = site i sold for USD?
C = Total cost of n sites.
Then:
E=p(1)s(1)+(p2)s(2)+ ... + p(n)s(n) - C
It is up to you to figure out how easy it is for you to get E > 0 after some time (years).
In my view there is an optimum amount of sites that you should work on. Some will say, concentrate on one site. Other people say, I have 15 - 20 sites. On two of the sites I earn enough to feed me and my family. The rest is a zero sum game.
If you think you find great names, you may buy them, but the number of possible names are constantly increasing. .EU was recently introduced. See my point on names? I have about 20 sites, and that is good enough for me. My children (I have 4) may get and administer some of them in the future. Hopefully, when that time comes the site has traffic and page rank (if that is a fruitful concept in some years).