In an earlier post , I explained that stock trading costs consist of the visible commissions and the less visible costs due to stock spreads . In the case of commissions, it is obvious who gets the money – the brokerage. In the case of stock spreads, it is less obvious where the money goes. If traders lose money due to spreads, who gets this money? You could imagine a stock trading system where potential buyers and sellers each submit a price and a number of shares, and a computer tries to match them up. This sort of system might work well for a very liquid stock that trades millions of shares each day, but it wouldn’t work as well for thinly-traded stocks. Suppose that you are looking to sell 100 shares of little known XYZ stock, and for three days running there haven’t been any reasonable bids to buy the stock. You would be quite unhappy. To make the system run more smoothly, each stock has one or more market makers whose job is to create a market in that stock. Market ma...

In an earlier post I discussed how using a limit order to beat the spread doesn’t really reduce spread-related trading costs. Does this mean that investors should always use market orders when buying or selling stocks? Definitely not. Let’s continue with the example of SPNS that had a quote of bid $1.45x200 and ask $1.60x2500, and we wanted to buy 6000 shares. In the earlier post we put in a limit order at $1.52. But there is a significant risk that this order won’t be filled. Maybe we really want to own this stock. (Please note that I don’t own SPNS, and I don’t know anything about it other than its quote worked well for my example.) We could just put in a market order and see what happens. One possible result is that we get 2500 shares at $1.60, and the rest at $1.61. Great. We get all the shares we want at a reasonable price. But what if we get the first 2500 shares at $1.60 and then 500 at $1.70, 1000 at $1.95, and 2000 at $2.40? This works out to $11,600 when we expected to p...