I've recently got my 13 year old son interested in learning about stock trading and options. He recently came to me with this question that I did not have an answer for. I am hoping someone in this forum can help.

He noticed that NVDA stock in Google finance showed a spike (up and immediately down) in after hours trading:

Google Finance after hours NVDA stock chart

but we could not find the same in other places for example marketwatch:

Marketwatch NVDA after hours stock chart

Another example: Walmart (WMT) stock today, Google finance shows a major dip after hours but Marketwatch shows no such thing.

Does anyone know why? Is this just a bug in Google stock charting software or something else?



Answer: something else (it's complicated!)

After-hours and pre-hours trading is not subject to any price barriers or halts, and is mostly unregulated accordingly. Trading errors/spikes are a significant trap for the unprepared, especially when there are corporate actions involved such a dividends, spinoffs, splits etc.

Different sources will include/exclude certain trade conditions and/or exchanges/trading venues. For regular market-hours trading, the rules are well-defined by the Consolidated Tape Association and the UTP Plan.

For after-hours and pre-hours trading, there are no rules regarding trades. Some sources will report all trades. Some sources will only report a subset of trades.

In the case of NVDA on trade date Jul 22 2021, there were no corporate actions involved (there was a 4:1 stock split a couple of days earlier, but this is unrelated).

Your attachment shows that NVDA closed at $195.94 but spike in after-hours trading above $205 according to one source and not to another.

Looking at high-cost instutional-level data feeds shows that NVDA had a single trade for $206.815 at 17:09:38 for 240 shares, performed through a dark pool.

The 15 most prior traders were from 17:08:45 through 17:09:10 at prices ranging from 195.65 thorugh 195.85 with volumes ranging from 1 share through to 128 shares from dark pools and NYSE Arca.

The 15 most subssequent traders were from 17:10:00 through 17:10:57 at prices ranging from 195.75 thorugh 195.85 with volumes ranging from 1 share through to 100 shares from dark pools and NYSE Arca.

The trade conditions present on the "spike" trade of 206.815 from a Dark Pool were: Form T (i.e. afterhours, which all the other trades had too), and "Average Price" which is meaningless on dark-pool-reported trades.

Did the trade occur? Yes. Was it cancelled? No. Is it generally indicative of the price of NVDA? No.

Draw your own conclusions from this example of after-hours trading. For what it's worth, published data from exchanges/consolidate tape do NOT incorporate any pre- or after-hourse trading into daily pricing (OHLC) data.


This is a frequently asked question. More often than not, sharp spikes are due to bad data. For example:

Bad Data 1

Bad Data 2

Bad Data 3

and so on ...

  • We did look for other posted questions but did not find the ones you'd mentioned. Regarding the answer, the data would be the same for all charts? Why would one chart show a spike due to bad data when the others don't? Are the others doing something to neglect the spike and if so how do they know a reasonable legitimate spike from others? If it is an issue with Google charts, somehow their data is corrupted, is there not an incentive for them to fix it? This almost always seems to happen afterhours and I've not seen an instance yet when there is a discrepancy normal hours.
    – CharmQuark
    Jul 26 at 17:15
  • Lastly, I did not understand the dark pool hypothesis above, your thoughts on that?
    – CharmQuark
    Jul 26 at 17:16
  • I don't know if web sites get their data from the same source so I can't tell you if the data should be the same for all charts. If the web site is free, what's their incentive to be perfect? Bad trades do occur during real time - aka a during discrepancy normal hours. As for the dark pool, lot of was presented buy no explanation was provided. I have no opinion on dark pools since I do not know how they operate. Jul 26 at 17:29
  • I agree that at the outset it seems like Google is an outlier with regards to the data and they must be wrong as no one else seems to reflect the same trends. We want to consider other options before concluding on that hypothesis. One way to eliminate the Dark Pool and other hypothesis is to look at paid, professional data providers to see if the data reflects what there in Google or not. I'll attempt it with the paid providers we have access to.
    – CharmQuark
    Aug 8 at 18:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.