How do we report data errors in WealthData to have them fixed. In particular I see:
GLD Data Spike: 4/7/2016
HYG Data Spike: 11/10/2020
IEF Data Spike: 3/16/2016
QQQ Data Spike: 8/1/2016
SPY Missing Date: 11/21/2016
SPY Data Spike: 12/21/2020
I tried performing data cleanup but nothing was fixed/changed.
GLD Data Spike: 4/7/2016
HYG Data Spike: 11/10/2020
IEF Data Spike: 3/16/2016
QQQ Data Spike: 8/1/2016
SPY Missing Date: 11/21/2016
SPY Data Spike: 12/21/2020
I tried performing data cleanup but nothing was fixed/changed.
Rename
Reporting them here like you’ve done is fine, I’ll take a look 👍
I compared these against other providers and here are my findings:
GLD Data Spike: 4/7/2016 - other providers show the same spike
HYG Data Spike: 11/10/2020 - other providers show the same spike
IEF Data Spike: 3/16/2016 - other providers show the same spike
QQQ Data Spike: 8/1/2016 - inconclusive, Tradier and Norgate have this spike but Yahoo data doesn't
SPY Missing Date: 11/21/2016 - got it filled
SPY Data Spike: 12/21/2020 - other providers show the same spike
So most of these are valid data, with one missing date corrected. Thanks!
GLD Data Spike: 4/7/2016 - other providers show the same spike
HYG Data Spike: 11/10/2020 - other providers show the same spike
IEF Data Spike: 3/16/2016 - other providers show the same spike
QQQ Data Spike: 8/1/2016 - inconclusive, Tradier and Norgate have this spike but Yahoo data doesn't
SPY Missing Date: 11/21/2016 - got it filled
SPY Data Spike: 12/21/2020 - other providers show the same spike
So most of these are valid data, with one missing date corrected. Thanks!
Thanks Glitch!
The SPY data spike on 12/21/2020 should be removed and the high corrected to 368.805.
That one was an obvious and terrible outlier. If it were representative of the real market, you'd see a bump in QQQ too. Of course the Time-Of-Sales can't be recovered from nearly 5 years ago, but based on the volume it appears that several block trades went off at a price far outside the market.
That one was an obvious and terrible outlier. If it were representative of the real market, you'd see a bump in QQQ too. Of course the Time-Of-Sales can't be recovered from nearly 5 years ago, but based on the volume it appears that several block trades went off at a price far outside the market.
I'll add that to the backlog!
ETFs probably didn't get the "Wealth-Data" treatment" for historical trading back in 2016, but as a reminder, Wealth-Data isn't trying to match other providers showing the same spikes. Spikes are always real prints (they'd be removed if the trade were busted) but the point of Wealth-Data is to make the data more realistic for backtesting - like for the SPY correction above.
You can repeat the same intraday process for the other spikes in the list - like GLD here, whose Low should be corrected to 118.03.
fwiw,
QQQ's high for 8/1/2016 should be 116.13.
IEF's high for 3/16/2016 should be 109.345
HYG's low for 11/10/2020 should be 85.74
You can repeat the same intraday process for the other spikes in the list - like GLD here, whose Low should be corrected to 118.03.
fwiw,
QQQ's high for 8/1/2016 should be 116.13.
IEF's high for 3/16/2016 should be 109.345
HYG's low for 11/10/2020 should be 85.74
Glitch when will the data for SPY be updated for the missing date and presumably the bad spike? I updated my dataset but the issues for SPY remain.
Cone what reference provider/data are you using to provide corrections for the data spikes? Will WealthData be updated or do we need to make these corrections locally?
I have some other priorities I'm working on, so unsure. Keep in mind WealthData is a free service we offer, we don't maintain a dedicated team to manage the data such a paid service like Norgate would offer. I'll post back here after it's done.
If the spikes are a problem for your backtests, then by all means, right click them in the chart and "Edit Bar Data".
Your Response
Post
Edit Post
Login is required