We have a Time and Sales tab that follows the largest trades in a name each day and you can check historically. This can be downloaded too. We track the price paid, our smoothed theoretical value, and the current profit of the trade. https://gyazo.com/e6a5295eb83081e0b3f821c949ba2d51
I disagree with this post. I was an electronic market maker. We lived for market orders. Assuming you will always trade, a limit order will be superior to a market order. The only reason to use a market order is because of the assumption “you will always trade.” The cost of ensuring an execution is losing control of price.
My personal statistics is obviously limited to the stocks and ETFs that I traded, and as such anecdotal; but to be a bit more specific, I traded many stocks, ETFs and CEFs that either have a bid/ask spread of $0.01, or an "effective" spread of $0.01 or less, i.e. the spread of fills when placing marketable orders or orders within the quoted spread is usually less than $0.01. Price improvements under the zero commission plans at IB and Schwab are on average roughly 50% of half the quoted spread (midpoint to bid or midpoint to ask) in my experience; so for a $0.01 quoted spread, the "slippage cost" per share per trade would be ca. $0.0025. "Smart routed" non-marketable limit orders at IB for my type or orders come with commissions of ca. $0.0035 to $0.0065 depending on execution venue; let's say roughly $0.005 per share. In my experience the average is about the same if I directly route to various different venues. Also, market orders with commission plans fill at similar effective spreads as market orders under zero commission plans, in my experience. Lastly, non-marketable limit orders for low volatility and low volume stocks ETFs or CEFs take a looooong time to fill in my experience, even when I see the volume per consolidated tape steadily increasing, which tells me that I'm at a structural disadvantage; most retail trades seem to get filled off-exchange, and on exchange usually only when the market moves against the order (with exceptions of course). Putting all this together, it is clear that the zero commission market orders carry less all-in trading cost for my type of trading. The commission of non-marketable limit orders alone would be higher than the slippage of zero commission market orders, even without considering the fill quality i.e. slippage of the limit orders. Likewise, assuming that non-marketable limit orders on exchanges (or smart routed limit orders at IB filled either on or off exchange) are typically filled when the market moves against the order (or else matching orders are internalized or front-run by market makers by $0.0001), the slippage with respect to the new bid/ask alone would be higher than the slippage of market orders. I would be glad to hear specifics of alternate statistics, or evidence like studies or papers.
My opinion or my entitlement to it don't matter; I'm trying to learn by discovering factual evidence via discourse in this forum. In my experience market orders for stocks and ETFs too are typically executed between the bid and ask. I also believe that most stock as well as option retail orders are filled between bid and ask, regardless of order type. But that doesn't answer the question. More interesting would be studies on "effective spreads" by order type (market vs limit), which are harder to come by. Effective spreads are often measured with respect to the bid/ask or average price after some delay after the trade, to eliminate the market impact of the trade itself on the quotes or on the average traded price. I found some studies for options on the internet, but not specific to this question, and I also think the market structure changed over the years and also depending on index options vs equity options etc.