Python code for 3K ISIN

@Gurpreet Client has follow up

I believe there seems to be a misunderstanding. We aim to obtain 10 years of 19 characteristics from 3,000 companies (some of which we know support historical series and some do not):

• Supports historical series:
  • "TR.DividendYield"
  • "TR.F.ComStockBuybackNet"
  • "TR.NumberofSharesOutstandingActual"
  • "TR.NetDebtToEBITDA"
  • "TR.FreeCashFlow"
  • "TR.SharesOutstanding"
  • "TR.F.PriceToCFPerShr"
  • "TR.PCReturnOnInvestedCapPct"
  • "TR.WACC"
  • "TR.CompanyMarketCap"
  • "AVAIL(TR.F.IntrCovRatioTTM(AlignType=PeriodEndDate),TR.F.IntrCovRatio(AlignType=PeriodEndDate, Period=FY0))"
  • "AVAIL(TR.DPSMean(AlignType=PeriodEndDate, Methodology=InterimSum, Period=NTM),TR.DPSMean(AlignType=PeriodEndDate, Methodology=WeightedAnnual, Period=NTM))"
  • "TR.F.DPSComGrossIssue5YrCAGR(AlignType=PeriodEndDate, InstrumentType=Primary)"
  • "AVAIL(TR.F.EVToLeveredFOCFTTM(AlignType=PeriodEndDate),TR.F.EVToLeveredFOCF(AlignType=PeriodEndDate, Period=FY0))"
  • "AVAIL(TR.F.FOCFYldPctTTM(AlignType=PeriodEndDate),TR.F.FOCFYldPct(AlignType=PeriodEndDate, Period=FY0))"
  • "TR.FreeFloatPct"
  • "TR.Volume"
• Does not support historical series:
  • "TR.HQCountryCode"
  • "TR.GICSSector"

We have two methods to mass obtain this data:

Using Excel:

Using Python:

Client

• A way to download the data using Python with auto-imputing of the data as done in Excel.

Hi @Jirapongse ,

Thank you for clarifying the use of parameters in Excel and Python. The formula we are using in Excel, as your colleague mentioned earlier, is the following:

=@RDP  .Data($B$7:$B$40;D$1:V$1;"Sdate=#1 EDate=#2 Frq=CM Curn:USD RH:Calcdate Null=BLANK CH:FD,IN";C2;$B$4;$B$5)

The code we use in Python is as follows:

import refinitiv.data as rdfrom tqdm import tqdmfor isin in tqdm(isins):    try:        response = rd.get_history(            universe=[isin.strip()],            fields=[                "TR.DividendYield",                "TR.F.ComStockBuybackNet",                "TR.NumberofSharesOutstandingActual",                "TR.NetDebtToEBITDA",                "TR.FreeCashFlow",                "TR.SharesOutstanding",                "TR.F.PriceToCFPerShr",                "TR.PCReturnOnInvestedCapPct",                "TR.WACC",                "TR.CompanyMarketCap",                "AVAIL(TR.F.IntrCovRatioTTM(AlignType=PeriodEndDate),TR.F.IntrCovRatio(AlignType=PeriodEndDate, Period=FY0))",                "AVAIL(TR.DPSMean( AlignType=PeriodEndDate, Methodology=InterimSum, Period=NTM),TR.DPSMean(AlignType=PeriodEndDate, Methodology=WeightedAnnual, Period=NTM))",                "TR.F.DPSComGrossIssue5YrCAGR(AlignType=PeriodEndDate, InstrumentType=Primary)",                "AVAIL(TR.F.EVToLeveredFOCFTTM(AlignType=PeriodEndDate),TR.F.EVToLeveredFOCF(AlignType=PeriodEndDate, Period=FY0))",                "AVAIL(TR.F.FOCFYldPctTTM(AlignType=PeriodEndDate),TR.F.FOCFYldPct(AlignType=PeriodEndDate, Period=FY0))",                "TR.FreeFloatPct",                "TR.Volume",            ],            start="2015-05-29",            end="2025-06-03",            interval="monthly"        )        response.to_csv(f"historic_data/historic_refinitv_data_{isin}.csv")    except Exception as ex:        print(ex)        with open("historic_data/errores.csv", "a") as file:            file.write(f"{isin}\n")

For example, in the case of ISIN CH0010570767 (please refer to the image below, Python on the left side and Excel on the right side), we obtain the following result and, setting aside the discrepancy in the data (which we are resolving separately), the behavior of both extraction methods differs:

• Indexes (dates) differ as you can see.
• Excel auto-imputes missing values (for “NumberofSharesOutstandingActual” it uses forward-fill method), but Python does not, it leaves the blanks.

So we have to handle the missing data ourselves or provide an unknown (if it exists) parameter to the function `get_history`, we rather use this last option.

Our goal is to obtain every company's data (3k companies by 19 features) for a 10 year period with the missing-values already filled using Python.

Best regards,
Pedro Ventura

Hi @Jirapongse ,
This is exactly what we needed — thank you very much.

Would you mind pointing me to the documentation where all the fields and their properties are listed? For instance, I wasn’t aware that "TR.DividendYield" has a "Calcdate" property. Is this documented somewhere?

Thanks again,
Pedro Ventura