If buying something does not affect the supply, or if the ability to buy something is not limited by an available supply, then the price discovery mechanism will not work properly. For every buyer there has to be a willing seller. If nobody's buying at the price you want to sell at, then a sale should not happen. The seller should have to lower his or her price to meet demand if he or she wants to complete the sale.
That was always one thing that bugged me about original Elite. Say I was carrying 35 T of computers into an agricultural system where the price was 101.5 CR. When I got there, there were simply no computers available for purchase they were selling so hot. But as the seller, I had no way of increasing the price. Who knows, maybe I could have gotten 150 CR for those computers.
So, if we're dealing with open markets, prices in principle should be constantly in flux. Commodities trading in real life is very volatile, and is very sensitive to market pressures as well as monetary fluctuations. Today, when central banks engage in quantitative easing, investors will hedge on weak currency by buying commodities.
That means prices are affected not only by supply and demand of the goods traded, but also the supply and demand of money itself. An additional dimension that affects pricing is access to credit. Can you borrow money to buy stuff? The more access to credit there is (i.e. the lower the lending standards), the more money is available for consumers, traders, etc. to buy stuff.
In fact, how much people borrow has a tremendous impact on overall consumption. It's the primary mechanism by which new money is created. This is a chart I produced at work that shows the percent increase of credit outstanding in the U.S. (that is, all debts public and private) versus percent increases and decreases of the Gross Domestic Product
based on U.S. government data that you can see here in Excel. What I did was plot data back through 1945, calculated percent increases and decreases, and then started looking for relationships.
What it shows is that whenever there is a credit contraction or even a credit slowdown, consumption (i.e. GDP) slows down or even contracts. This too has an impact on prices. Less consumption means lower demand, and therefore lower prices. When credit is really moving fast, as in the U.S. housing bubble, prices then tend to go to the moon. In the case of economic bubbles, those tend to be driven by monetary and financial policies, in particular, credit policy.
Inflation also tends to have an impact on credit allocation, inasmuch as high inflation usually leads to higher interest rates. This is done to take more money out of the system and lower prices. It also creates a disincentive to borrow money at higher rates. See here, inflation tends to lead to higher interest rates, and eventually, higher interest rates tends to drive down inflation:
So, for the purposes of the game, I'd suggest that the following elements lead to price discovery on a system-by-system basis: 1) supply and demand of goods (let sellers set a price and see if there are buyers at those prices); 2) credit availability (which the more there is the higher prices can go, and leads to higher consumption); 3) economic growth (i.e. how much consumption is taking place which demonstrates demand); 4) inflation; 5) interest rates; and 6) balance of trade between systems.
I think done in a simplistic way, the economic information determining prices can be pretty realistic, and whatever formula is used can be based on real data from economic history. Price discovery therefore can become very realistic, predictable in some capacity based on certain data points, and traders can time their sales to make more money. Or they could be caught in an unforeseen recession that makes their six-month old information obsolete when they arrive and forced to book a big loss on their trade just so they could afford fuel and missiles.