LFP236 – Making Business Loan Decisions in Minutes w/Chirag Shah CEO Nucleus Commercial Finance

It was not that long ago that business lending decisions took days, weeks, or even months. This lengthy period – and then need to perhaps apply to multiple banks was a severe challenge for many organisations, many of whom over time have gone bust whilst going through such processes.

But what if using the best of technology this extended time-period could be reduced to minutes moving way nearer to B2C BNPL decisions? Even a few years back this would sound highly unlikely in anything other than special cases. However Nucleus Commercial Finance have funded £2.6bn of loans and processed over 100,000 loan applications for over a decade with successful results and now make decisions in less than 5 minutes and aim to get that time even lower.

Needless to say this is no easy task and in this episode founder and CEO Chirag Shah leads us through the journey to that. Key elements are naturally the data – where Open Banking and Open Accounting are the technical conduits to getting data, permission to access this data ahead of the need for a loan and naturally the challenge of ensuring that automated decisions are at least as good as if not outperform those of human credit officers.

Topics discussed include:

  • monsoon stopped play
  • monsoon patterns in South and South East Asia
  • Chirag’s career journey to founding Nucleus in 2011
  • the unrecognised root of Fintech in terms of Hedge Funds independence of banks and strong use of technology in the 90s/00s
  • Nucleus origin story
  • key components of the model
  • SME lending even a decade ago in terms of process and timescales
  • 6 million SMEs in the UK and monitoring India as a potential but plenty of scope left in the UK
  • avoiding trading off decision time for decision quality
  • analogy with BNPL’s approach to rapid credit decisions
  • lack of credit rating agencies large stores of data unlike with B2C
  • main challenges are around data rather than decision
  • embedded lending in the SME market
  • variable rates of success by Fintechs
  • the archaic nature of the data available data at Companies House
  • “there are lots of SMEs who can’t produce management accounts”
  • using Open Banking and also Open Accounting
  • the challenge of accounting standards ever-more muddying the water over the actual business situation
  • EBITDA and “adjusted” EBITDAs
  • overcoming this by building models based on the raw banking data
  • deriving the business’ unit economics
  • the roots of these challenges in the 19th invention of “management accounts” breaking with all tradition
  • the challenge of getting access to SMEs accounting data before they need a loan decision
  • businesses are getting more comfortable with granting access to open data but only when they need a loan
  • less comfortable with granting ongoing data
  • how Nucleus overcomes this by sharing their modelling of the business performance as a Dashboard with the SME which helps it all the time not just at “needing a loan” time
  • how Nucleus overcame the challenge that long term loans make it hard to calibrate models – it being say 5yrs until you find whether your model was making “good” decisions
  • evolving to superior automated credit decisions than “human” decisions
  • parallel running the system and credit officers for 18mts
  • segmenting the SME market
  • Nucleus’ new platform Pulse as a decision making tool for the businesses as well as being used by Nucleus for lending decisions
  • partnering with accountants

And much much more 🙂

Share and enjoy!