Downloading all data from a GOVSRCH EIkon panel

sumit

Hi,

Is there a way I can download (via python) all the data shown in GOVSRCH? I tried the excel approach, but there's a limitation of 4000 records.

Regards

Sumit

jason.ramchandani01

@sumit So I think you need to separate things into 2 bits - the first bit is to identify a universe of bonds satisfying some criteria, the second bit is to get some information (such as prices etc) about those instruments. The Search API will help with the first part. The article and its accompanying notebooks contain examples entirely relevant to your bond use case - consider this from the following filter notebook

Example - Search for Brazil global sovereign bonds in USD with a duration greater than 1

rdp.search(
    view=rdp.SearchViews.FixedIncomeInstruments,
    filter = "RCSIssuerCountryLeaf eq 'Brazil' and \
              RCSCurrencyLeaf eq 'US Dollar' and \
              RCSTRBC2012Leaf eq 'Government & Government Finance' and \
              OptionAdjustedDuration gt 1",
    select ="RIC, ISIN, MaturityDate, CouponRate, IndustrySector, FaceOutstandingUSD, OptionAdjustedDuration"
)

or this from the same notebook:

Example - Floating Rate Notes

rdp.search(
    view = rdp.SearchViews.GovCorpInstruments,
    filter = "CouponRate gt 2 and \
              RatingsScope(RatingType eq 'MIS' and CurrentRatingRank lt 10) and \
              RCSCouponTypeName eq 'Floating Coupon' and \
              IsActive eq true and \
              RCSTRBC2012Name eq 'Healthcare' and \
              RCSCurrencyLeaf xeq 'Euro'",
    select = "CouponCurrency, IssuerName, IssueDate, RIC, CouponRate, ISIN, Ticker, RCSTRBC2012Leaf, MaturityDate, AssetState",   
)

Between these two examples and others in the notebook you should be able to progress sufficiently. Note: maximum results for this API are 10,000 so you may need to build things up by country or sector or type or maturity date buckets in order to get all the instruments you need. A simple check for this is if the length of the return is 10,000, most likely it has been truncated so you need to segment your filter further. I hope this can help.