At last count, I remember seeing a list of 90+ “alternate lending” (fintech) startups in India. These, of course, cover various niche white spaces within the $15Bn+ lending landscape in India – P2P lending, cash advances, B2C retail loans, micro-loans, buy now pay later, POS credit, SME financing, purchase financing, bill discounting, and the likes. Then, we have the large and medium sized banks and NBFCs keen on having their own digital-channel-alternate-offerings.
And not quite alternate lending, Yet!
We have several examples from the western world, such as Funding Circle, Klarna, Lending Club, Sofi, that one could learn from. Unfortunately, most conversations in India are about working around the first principles – the traditional processes – rather than disrupting them.
The real extent of “alternate” is the presence of a web-portal or a mobile app (for loan origination). But these alternate lending companies need to consider alternate data, alternate processes, and alternate lending strategies. NOW. The promises of “algorithms and machine learning”, “alternate data for alternate lending”, “instant decisions”, “truly digital”, “low cost”, “faster approvals” need to be fulfilled.
Can it become truly alternate then?
Alternate lending companies need to find the right balance –
- Risks Vs Bets
- Data vs. Technology Vs. Analytics
- Faster Decisions vs. Thoughtless Experiments
There is also a need to focus on what you must, and defocus from what you don’t need to build yourself. The need to not fear that you’ll lose your secret sauce. Remember? “The secret ingredient is… Nothing.”
As an analytics/ data/ insights guy – I think most alternate lending players need to really invest NOW (not later) in extracting the juice out of their alternate data. Whether it’s working with companies like Algo360, CreditVidya, Lenddo, etc. that are innovating on the extraction and use of alternate data in lending, or building in-house capabilities, the ones that turn the tap on alternate data faster, will probably differentiate faster – on cost – on risk- on efficiency.
Whether we are talking about parsing information from SMS, or triangulating locations, or measuring social engagement levels with social media and call behaviour data, or just establishing behavioural patterns with device usage, utility bills, professional credentials, etc. – if you invest now, you benefit earlier than your competition.
Try deep learning using MATLAB