An interesting concept.

Personally I think the biggest factor is what the majority of people think will happen to the value in the near future. I'm sure you are familiar with the self-fulfilling concept of market speculation, where expecting a price decrease will convince you and others to sell/not buy, which in turn reduces demand and contributes to price reduction.

I think that this factor probably realistically outweighs all the other ones. Simply the error of this factor I think is probably larger than the total effect of other factors combined. Particularly since factors such as current value and number of transactions could be lumped in with this. Find a way to break down the information that influences what people expect to happen to the value and you will be golden.

No sarcasm intended, http://www.whatdoestheinternetthink.net/ could potentially be useful if used strategically.