I'm thinking about providing a quarterly report/newsletter or saas service which reports on an organisations own market data and various market data quality, comparison and similarities/dissimilarities metrics and benchmarks. I have built some software which automates database building, processing, aggregation and data analysis and would use this to automate the report. Data licensing issues aside would this type of service be of interest to market data providers, exchanges, brokers, funds, investment firms etc and how best to approach them?
Professional Market Data Service
- Thread starter kingquant
- Start date
Baron
ET Founder
I know this goes without saying but the interest you'll get in that will depend on the quality and comprehensiveness of the report and how much you're going to charge for it. Without seeing your final product, your question about the interest you'll get just has a lot of unknown variables at this stage.
Agreed. I would build a prototype report and develop it going forward. Historical reports would be available for evaluation and would include the standard metrics/coefficients used etc. Any charges would be in line with the quality/comprehensiveness of the output. Once a minimum viable report is available then companies can take a view as to their level of interest. Just not sure what type of firms to approach first?
So you would validate the data that a firm has?if they own data from ORATs, you report on the quality of orats vs what? how would you determine the right value for a particular field?
secondly, outside of fundamental financial data and calculated data (like implied volatilities), I think there’s only a few underlying sources and a bunch of interpretations on how to view those underlying sources. Once those methodologies are validated then what’s the follow up newsletter going to say?
Firstly, any report would not attempt to cover all markets and data types. There are many sources of data and data providers. Professional firms that have the budget may no doubt limit their data sources to a select few so any service may only be applicable to data providers, emerging funds etc. Providing the data is available there are techniques to compare data quality, similarity measurements, trends, issues etc for these types of problems. Not saying it is easy but some value might be delivered to those that need to have and retain a degree of confidence that the data they are using/publishing is accurate on an ongoing basis.
Last edited:
Any sane pro has multiple sources of data and performs lots of verifications.
Heck, I've got both fiber and starlink, and operate multiple independent VPSs on different hosts. When there's a trading signal discrepancy between data source A/B/C/D I check why. Some verifications are post-trade, so I have careful wind-back procedures to mitigate bad signals.
And I'm a one man band.
An issue that comes up quite regularly is handling of spinoffs (surviving entity), special dividends, and foreign dividends - everyone handles it a little differently. It's not that they're wrong as such - they're handled differently. The fun happens when a source varies their methodology arbitrarily...
Heck, I've got both fiber and starlink, and operate multiple independent VPSs on different hosts. When there's a trading signal discrepancy between data source A/B/C/D I check why. Some verifications are post-trade, so I have careful wind-back procedures to mitigate bad signals.
And I'm a one man band.
An issue that comes up quite regularly is handling of spinoffs (surviving entity), special dividends, and foreign dividends - everyone handles it a little differently. It's not that they're wrong as such - they're handled differently. The fun happens when a source varies their methodology arbitrarily...