The in-demand stocks will rise quickly with the limited supply strategy and lesser demanded stocks will drop to be penny stocks. It'll probably be 80/20 where 80% are worth pennies.
Once you have offloaded your entire inventory now you just allow others to trade between themselves and you are done. The market will correct your approximations in short order.
If you are looking for a pricing engine to handle the limited supply strategy, send me an email. Given that this is all funny money I'll just let you use our REST API for the pricing.
And in real stock exchanges, companies issue shares of the company, so they have real value. InTrade links its share value to the likelihood of something happening - but the goals of websites/startups vary greatly, so it's difficult to set a fixed standard that way.
You might also be able to set up a market in which shares only have value because of scarcity, but I think it would be difficult to convince people to buy them in the first place, which is why starting with funny money might actually be a good move in trying to establish this sort of market.
And should sites be allowed to apply for the exchange? Do you think you could allow the companies that own these sites to sell actual shares to accredited investors using a Funders Club model?
If it were bitcoin based tho....
Online trading markets are excellent at exposing 'insider information', but the prices have to be informed by (a) availability of stock and (b) the ability for traders to negotiate prices between themselves.
In this instance, the valuation is based on:
"...a simple formula to give each domain a daily price."
So, what does this achieve? It rewards the user(s) who can best approximate the formula, gather the requisite information and act on it. In other words, the best performers are those that can best emulate the workings of the site.
Ok, it is called "LinkDAQ" for a reason - but there is an opportunity here to build something linked to real value.
It would be better to seed initial stock values from such a formula, but then leave users to trade between them. Stocks "cash out" during [real] funding rounds, dividend payouts or difference in actual stock valuations. If you can refine the market sufficiently, you should have a reasonably informative prediction engine.
I've been a long time user of the iPredict prediction market in NZ (ipredict.co.nz) - but that's small biscuits next to InTrade, for example.