We use cookies on all our websites to gather anonymous data to improve your experience of our websites and serve relevant ads that may be of interest to you. Please refer to the cookies policy to find out more.

By continuing, scrolling the page or clicking a link, you agree to the use of cookies.

The Evolution and Growth of Alternative Data 
Two business people sitting at a computer


Last week, together with guest speaker Jonathan Holman, we hosted a webinar: The Evolution and Growth of Alternative Data. It was a fantastic, informative and interactive session with over 170 participants. During the webinar a number of questions were asked and there simply was enough time to get through them all. Jonathan has provided his response to any unanswered questions and we're sharing them with you below.


My results were based on UK companies and UK google searches. However, the other academic research I highlighted shows that the concept has been demonstrated to work all over the western world at least. One needs critical mass of the population using a search engine to make it representative. Google might not be sufficiently popular enough in every market to make it relevant.

From my experience, I wouldn’t say that’s a significant enough effect to damage the general trends shown in search term history.

M4 money supply figures, balance sheets and GDP figures would all suggest that far low levels of lending went on in those decades and so slower processes, with fewer companies present in any ono market, made the speed at which they were able to execute, viable, at the time.

I can’t give a definitive verdict on this, however, in general, the UK’s acceptance of cloud and ML is more receptive than what I understand to be the case in other markets in Europe and the US

Search engines will be relevant source of alternative data, as long as they’re used by a critical mass of the population. Mobile payment platforms, given their success, and the lack of penetration of mainstream banking to all the population, in developing countries, would likely be a great source of alternative data for credit risk assessment.

The wisdom of crowds is normally the best technique to manage this risk. The more people whom judge something, the more accurate the outcome is likely to be. Where people are invested with their own savings, such as in equity markets, this also makes them emotional and volatile. However the general trend is true. The more data, the more benchmarks, the more relative comparisons and the more judgements which have taken place, the more accurate it is possible to make subjective assessments.

It was true in the 1, 2 and 5 year from PD models I built in ML, yes.

Regulation would need to change to open up those platforms, as it has with open banking and psd2.

Missed financial filings, change of stock, management or name / SIC code, number of employees  a proxy for size), industry / activity, elongated reporting of accounts (following on from late filings)

Africa has great mobile payments businesses, Indonesia does too. In India, googlepay is massive now. There are loads of inclusive fintechs which create new data sources in pretty much all developing markets. Even mBills in Slovenia!