OHLC with C > H and QC of data

[jdumont0201]

Hi, OHLC stock data is plagued with inconsistent off the charts values with C >H. At specific timestamps, it shows the current trading value as 'c' instead of the close of the bar.

As I write, trading value of MSFT is 165 and it wrongly appears in the historical data of many years ago.

Example: https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1585010909&to=1585039609&token=YOURTOKEN

{ c: [ 165.6 ], h: [ 143 ], l: [ 141 ], o: [ 141.28 ], s: "ok", t: [ 1585036800 ], v: [ 41406 ] }

https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1547809000&to=1547809609&token=bpv606vrh5rabkt31ka0

{ c: [ 164.67 ], h: [ 106.5 ], l: [ 106.5 ], o: [ 106.5 ], s: "ok", t: [ 1547809200 ], v: [ 500 ] }

This is a plot of C over two years. The exact same value 165ish pops all the time. https://imgur.com/a/ofU5sB5

Don't you have basic quality control on your data? Thanks.

[jdumont0201]

OHLC with O < L also happens.

https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1548279000&to=1548281009&token=YOURTOKEN

Was

{ c: [ 106.71 ], h: [ 106.7651 ], l: [ 106.56 ], o: [ 106.48 ], s: "ok", t: [ 1548277200 ], v: [ 12271 ] } now { c: [ 106.71 ], h: [ 106.7651 ], l: [ 106.65 ], o: [ 106.7 ], s: "ok", t: [ 1548277200 ], v: [ 12271 ] }

Two problems:

• O<L
  • I suspect the O value to evolve spontaneously too, as 106.7 is a different value compared to a previous extract of mine.

[occasionallydavid]

FWIW I cancelled my subscription until the rate at which these data issues turn up verifiably dissipates. It seemed almost pointless to file tickets about many instances of missing data I found because I have no doubt you are already aware of them.

It might be worth slapping a [BETA] tag on all the endpoints in the docs then focusing solely on removing that label from candles and ticks, nothing else really matters if these pieces are so spotty.

[bowyern]

I see O or C outside the L/H bounds on about 45% of the 5m candle data I've looked at.

[finnhubio]

Hi guys, we are aware of the issue. Our apology. It will be fixed by the end of the week.

[finnhubio]

This is a plot of C over two years. The exact same value 165ish pops all the time. https://imgur.com/a/ofU5sB5

Don't you have basic quality control on your data? Thanks.

this issue specifically has been fixed.

[jdumont0201]

The same data on two different timespans (from/to) can also show different values. Example with hourly MSFT at ts 1163682000.

https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1163675388&to=1163688188&token=YOURTOKEN

{ c: [ 29.12, 29.19 ], h: [ 29.11, 29.23 ], l: [ 29.03, 29.1 ], o: [ 29.03, 29.12 ], s: "ok", t: [ 1163682000, 1163685600 ], v: [ 29687, 2475343 ] }

is different when narrowed down.

https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1163675388&to=1163684188&token=YOURTOKEN

{ c: [ 29.03 ], h: [ 29.1 ], l: [ 29.03 ], o: [ 29.03 ], s: "ok", t: [ 1163682000 ], v: [ 7001 ] }

[soffio]

The same data on two different timespans (from/to) can also show different values. Example with hourly MSFT at ts 1163682000.

https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1163675388&to=1163688188&token=YOURTOKEN

{ c: [ 29.12, 29.19 ], h: [ 29.11, 29.23 ], l: [ 29.03, 29.1 ], o: [ 29.03, 29.12 ], s: "ok", t: [ 1163682000, 1163685600 ], v: [ 29687, 2475343 ] }

is different when narrowed down.

https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1163675388&to=1163684188&token=YOURTOKEN

{ c: [ 29.03 ], h: [ 29.1 ], l: [ 29.03 ], o: [ 29.03 ], s: "ok", t: [ 1163682000 ], v: [ 7001 ] }

I also have this problem.

[GF-Huang]

@finnhubio What is your data source? Why is there such an error? I am very concerned about this issue.

[finnhubio]

@finnhubio What is your data source? Why is there such an error? I am very concerned about this issue.

There was an error when we built this data from tick. It will be fixed by the end of Sunday. We are already importing the new data in.

[finnhubio]

The same data on two different timespans (from/to) can also show different values. Example with hourly MSFT at ts 1163682000.

https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1163675388&to=1163688188&token=YOURTOKEN

{ c: [ 29.12, 29.19 ], h: [ 29.11, 29.23 ], l: [ 29.03, 29.1 ], o: [ 29.03, 29.12 ], s: "ok", t: [ 1163682000, 1163685600 ], v: [ 29687, 2475343 ] }

is different when narrowed down.

https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1163675388&to=1163684188&token=YOURTOKEN

{ c: [ 29.03 ], h: [ 29.1 ], l: [ 29.03 ], o: [ 29.03 ], s: "ok", t: [ 1163682000 ], v: [ 7001 ] }

Regarding this issue, you need to expand the "to" params. For example in the 2nd request, you are requesting data from 8:00 -> 8:36 AM, so the API will build the 1H candle using 36 minutes of data. You need to expand the "to" param to 8:59AM and you will get the same result.

Try this instead of your 2nd link: https://finnhub.io/api/v1/stock/candle?symbol=MSFT&resolution=60&from=1163675388&to=1163685540&token=

[finnhubio]

As promised, the primary issue with "OHLC with C > H and QC of data" has been fixed. Can you guys try it out and let us know ?

[finnhubio]

As no bug is reported after the fix, we will close this issue for now.