For a deeper look into our Eikon Data API, look into:

Overview |  Quickstart |  Documentation |  Downloads |  Tutorials |  Articles

question

Upvotes
Accepted
5 0 0 2

Discrepancies/Errors between Python API and Refinitiv Eikon data

Hello,


I have been recently retrieving brokerage recommendations from Reuters using the Python Eikon API.

I am interested in the price target and the grade recommendation assigned by a brokerage to a stock.

Because I need this information to be accurate, I have been handchecking data returned and have found many

errors and discrepancies between what the Python API returns and what is shown by Eikon in the Reuters terminal.

For example, I use the following code to pull recommendations for ticker IRT:

import json

symbol_list = ['IRT']
fields = [
              'TR.TPEstvalue.brokername',
              'TR.TPEstValue', 'TR.TPEstValue.date',
              'TR.BrkRecLabel', 'TR.TPEstvalue.analystname',
              'TR.RecLabelEst',
              'TR.RecEstValue', 'TR.BrkRecEstConfirmDate'
         ]
data = ek.get_data(instruments=symbol_list, fields=fields, raw_output=True)
print(json.dumps(data, indent=2, sort_keys=True))


The results for brokerage RBC are the following:

screenshot-48-min.png


Python API says RBC is at a Neutral rating for ticker IRT.

However, the following is returned for when looking at recommendations on the Reuters Terminal:

screenshot-49-min.png


As you can notice, the true grade is an Outperform/Buy rating, not Neutral.

I have posted before on this issue (see "Possible BrkRecLabel and RecEstValue Data Errors"), and this is not an issue of fields retrieved.

I have also been in contact with the Eikon Help Desk, who confirmed data is correct on their end (see case 11302953).

This is a Python API data issue that needs to be corrected. Let me know if more examples would be helpful. These discrepancies appear

in many brokerages and stocks so I have many available.

eikon-data-apirefinitiv-dataplatform-eikondataerrorpython api
icon clock
10 |1500

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

Upvote
Accepted
12.1k 59 39 61

Hi @bshapiro

I've reached out to content specialist who provided an explanation as to what is happening. Specifically, it appears to be a common issue if you mix specific data sets - target price and broker recommendations - in one request. Each data set is sorted differently in the backend.

A proposed solution is to perform 2 separate queries for the different datasets and combine them.

For example:

import eikon as ek
import pandas as pd
from functools import reduce
...

symbol_list = ['IRT']
fields = ['TR.BrkRecLabel.brokername', 'TR.BrkRecLabel', 
          'TR.BrkRecLabel.date'
         ]
data1, err = ek.get_data(instruments=symbol_list, fields=fields)
data1


symbol_list = ['IRT']
fields = ['TR.TPEstValue.brokername', 'TR.TPEstValue.analystname', 
          'TR.TPEstValue', 'TR.TPEstValue.date'
         ]
data2, err = ek.get_data(instruments=symbol_list, fields=fields)
data2


Combine them:

result = reduce(lambda x,y: pd.merge(x,y, on='Broker Name', how='outer'), [data1, data2])
result

ahs.png


ahs.png (51.4 KiB)
icon clock
10 |1500

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

Upvotes
5 0 0 2


I can confirm Refinitiv Eikon data is correct since I have the most recent RBC publishing that declares the recommendation:


screenshot-50-min.png


icon clock
10 |1500

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

Hi @bshapiro,

I've reached out to the content specialist team to get involved. Sorry for the long delay - I will keep a close watch on this and ensure someone will get back to you.

Upvotes
481 2 1 2

Hi @bshapiro ,

Thank you for your feedback.

This a content data from fundamental data service that Python API can't fix.
I raised this issue to relevant team.

icon clock
10 |1500

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

@pf

Hello, it has been about a week since I asked this question. Are there any updates to this issue? And can you put me in contact with the relevant team working on this?

@pf

Hello, can I have a response on the status of this issue?