Choosing a Theoretical Price from Limit Order Book Quotes and Depth
Summary
The document discusses why an order book does not determine a unique theoretical stock price. Possible reference values include the last traded price, the unweighted midpoint between the best bid and ask, and a midpoint weighted by displayed order quantities. Which measure is appropriate depends on how much information the analyst believes visible order depth contains.
For a market maker, a single theoretical value may be insufficient: the trader can hold separate estimates for buying and selling and use them to set quotes, which need not sit at the top of the book. The answers also distinguish quoted prices from estimates: adding volume at a bid does not necessarily improve the price available to a buyer. No order book values or empirical tests are provided, so the discussion gives interpretive choices rather than a universally validated pricing formula.
Key ideas
- The last trade and the best bid-ask midpoint are distinct candidate reference prices.
- A depth-weighted midpoint assumes displayed order quantities contain useful information about value.
- Market makers may use different theoretical prices for their buy and sell quotes.
- More displayed volume at one price does not by itself improve the execution price available to a trader.
Tags
Full text
# Order books theoretical price # Order books theoretical price Suppose we have the following bid/ask spread for a particular stock: What is the theoretical value of the stock? Now suppose we have: What is the theoretical value of the stock? Now suppose a major event in the market occurs and the particular volume at a price point changes as following: What is the new theoretical price? ## Answer by JCPBBT (score 2) https://quant.stackexchange.com/a/42372 It is totally up to the type of use or the type of interpretation that you want to give to the theoretical price and also it depends on what are your belief about the information content of the queue of orders that are currently resting and are visible in the limit order book. Depending on whether or not you believe that the information displayed in the limit order book adds to your final result for the theoretical price, you could take: - the last price at which the security was traded; - the mid-price calculated as an unweighted average of the best bid and best ask prices; - the mid-price calculated as a weighted average of the best bid and the best ask prices (depending on what is your belief about information content of the depth of the LOB on each side you could decide to include only level 1 information or use the entire depth as in Cao, Hansch, and Wang (2004) 'The Informational Content of an Open Limit Order Book' ). However, in one of your comments I read that your interested in how a market maker would price that security, and in that case the answer may not be a single number since they are requested to post quotes on both sides of the LOB and therefore they will have a theoretical price at which they believe it is reasonable to sell the security and a theoretical price at which they believe it is reasonable to buy the security and based on that they will post their quotes in the LOB (take also in consideration that they are not obliged to post at the top of the LOB so their theoretical prices may differ from the best bid and ask). ## Answer by David Duarte (score 0) https://quant.stackexchange.com/a/41894 Well, the price is not changing throughout your examples. The bid/ask price will be the highest buy order and lowest sell order. If by theoretical price you mean a mid, you would just average the two and maybe say that price is valid up until the min amount between the bid and ask. Doing some kind of weighted average doesn't really make sense because if you are buying for example , just because the volume went up on the bid, doesn't mean the price has improved for you.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.