Choosing a Spot Estimate from Trades and Best Quotes
Summary
The document discusses how to estimate a spot price from tick data when the latest trade and the midpoint of the best bid and ask may tell different stories. A stale trade may lag a changing market, while a quote midpoint can move with the spread, including during volatile periods or in illiquid markets. The question also asks how providers might aggregate ticks into minute-level prices, but the answer does not give a specific aggregation procedure.
The response treats the last trade as a useful default because it records a completed transaction between counterparties. It cautions that quotes can be informative even without a trade, particularly when both sides are refreshed and stand well away from the last execution; in that circumstance it suggests considering the bid–ask midpoint as a better estimate. This is a qualitative judgment rather than a precise rule: it does not define how old a trade may be, how wide a spread must be, or how to handle missing quotes. The approach therefore needs explicit thresholds and market-specific handling before it can be applied consistently.
Key ideas
- A completed trade records an executed price, while displayed quotes show available interest rather than a transaction.
- A last trade may become stale as quotes change.
- A quote midpoint may be informative when both sides are regularly refreshed and diverge materially from the last trade.
- The response gives no thresholds or formal procedure for combining trades and quotes into minute bars.
Tags
Full text
# Calculating spot level using tick data # Calculating spot level using tick data What is a proper (or commonly used/accepted) way of calculating some spot value with tick data? At the moment I can think of two options: - Take the latest available best bid, latest available best ask and find mid. (Trade represents both best bid and ask at the same level) - Take the latest available trade and consider it as a spot price. The problem with 1 for me is that spread can change dramatically over time, especially in the beginning of trading which can affect the mid and show unrealistic prices. The problem with 2 is that bid-ask can move significantly without any trades being performed and so latest trade wouldn't represent the market situation any more. If the considered asset is not very liquid there come lots of problems (i.e., no bid available at all). I can think of a few tweaks, but I'm quite sure I'm trying to re-invent the thing and there must be a common solution for it. Is there? I.e., how do data providers calculate minutely prices based on tick data? ## Answer by Matt Wolf (score 2) https://quant.stackexchange.com/a/8971 Neither is entirely correct. I think you need to find a middle ground here because: a) Bid and Ask do not have to be identical when a trade occurs. Someone may lift the offer and that results in a trade without ever having bid the asset b) The last trade may have occurred some time ago as pointed out and true value will most likely lie somewhere between bid and offer. Having said that, I believe market accepted standard is to take the last trade as true value because that is where two counterparties agreed to conduct business. Just because someone shows buy and someone else sell interest does not mean that anyone is willing to trade at levels away from their shown interest and thus true value is not reflected in current bid/offers alone. I would take the last trade as "true value" but if bid/offers are significantly away from the last traded price and if bid and offers are regularly refreshed, meaning market participants stand ready to trade at such indications then I would mark true value as the bid-offer mid price.
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.