Basic Economic theory regarding Trade, is that Trade profits is made through specializations. If everyone is self-sufficient, they reach one kind of equilibrium... but if there are differences in production capacity ratio between different products, then trade will arise (if people have perfect knowledge, and so on... a lot of theoretical assumptions.)
If you can produce 10 swords or 5 Mounts in a day's work, and your trade partner can produce
10 swords or 10 Mounts, you can either be satisfied with a self-sustained equilibrium of 3 Mounts and 4 Swords, or produce only Swords and sell swords for Mounts with your trade partner. Together you can then consume 10 Swords and 10 Mounts, so you if you get a really good trade deal will have 5 Swords and 5 Mounts instead. Those extra resources (1 Sword & 2 Mounts) is the profit made from trade, that is to be divided between the parties involved.
To make the theory even more complex, different people can have different preferences. That is, Mounts may be worth more to you than to them, and if you get more mounts and they get more Swords, you will still win even more as your needs are more fully met. (You may not like using Swords, but like using Mounts to flee with, and hence would be willing to trade nearly all your swords for Mounts; while the opposition wants lots of Swords and is more than willing to trade extra Swords for Mounts. You will consume even less Swords than before Trade, but you will be far more happier with the extra Mounts.)
In reality however (or simulation of such) things do not work this cleanly. The "invisible hand" is disturbed by monopolies and strong economic forces with superior information, why having a Global Auction house is a no-no for a functioning economy, as people can manipulate it too much for it to function properly.
On the other hand, there is also the point of transaction costs. If the Transaction costs are greater than the Trade profits, then it is not profitable to trade. That is, if you use your 5 gaming hours each to produce your Swords and Mounts, and you and your trade partner would each then spend ½ an hour per item produced, and the collective profit is 3 Swords, then if you both spend (3x½=) 1½ hour or more combined looking for customers, agreeing on a trade place, traveling to that place, making the transaction, and then travel to an area where you can resume your regular business... then you are at a LOSS.
Which is why Local Player Vendors must be made, as Play Time is the most valuable thing we have, and some do not wish to waste it.
And Vendors to sell things need to be according to need, so we need the Gnome Vendors for permanent stores with specific Crafters, Market Vendors to be hired selling things in Cities for Crafters to sell in bulk at need; and have NPC Merchants to sell minor amounts of items for any Gatherer wanting to get some quick Gold as well as starting Crafters to sell some low-grade stuff to other Newbs. For which there will be a market, when the transaction costs in time have been reduced.
And, there will still be a market for Player Merchants to play in, as the transaction costs is their cut of the business... apart from the face-to-face negotiation possibilities, making skilled Merchants having greater possibilities of profit.
This is just off the top of my head, and I can say quite a lot more about this subject, using many other theories, or discussing other aspects of these even, but I doubt few would care to read such Walls of Texts anyway. Not to mention that there is a reason why I rarely torment the population on these fora with solid theoretical backgrounds, as I doubt few non-academic or non-economists would care about the difference between an opinion and a proven theoretical viewpoint in this matter.