What is Algorithmic Underwriting?

algorithm underwritingAlgorithmic Underwriting

Most people have come across credit scoring as an established way of assessing credit applications. Algorithmic underwriting is almost a next generation system, however it does things very differently and if the borrower is unaware then it can often catch them out. Where credit scoring looks at the fixed factors, algorithmic systems concentrate on the live variables.

To understand how it works and what it does then let me cover a few points about credit scoring.

A typical credit scoring system will cover areas such as;

  • Home ownership, tenant or living with parents
  • Type of job, manual, professional and how long you have been it that role
  • Other credit options, debit card, credit cards etc
  • Age, postcode, general demographics
  • Industry sector, SIC code, length of time trading etc

You will have seen all these questions asked on a typical application form.

When it comes to algorithmic underwriting then it is very much the variables that get assessed with this being done via an automated advisory system. The algorithm takes into account areas such as;

  • Bank statements, balances and transactions
  • VAT records
  • Accounts, filed and current year
  • Companies House info

The big thing to understand is where these records are pulled from.

What To Know

Lenders using algorithm based systems will typically ask for a series of information from you. Normally this is uploaded via an on-line platform with the system asking for;

  • Company registered number
  • VAT number and VAT returns
  • HMRC log on details
  • Bank statements downloaded or via a read only access to your on-line system
  • Read only access to cloud accounting or downloads of current financials

It all looks fairly reasonable, and to a large degree it is. Where borrowers risk falling over is when the information available to the lender is not up to date or inaccurate.

Let me expand.

Common Pitfalls

Bearing in mind where the systems pull their information from then borrowers often fall over because of the following;

  • No recognised bookkeeping system. Spreadsheets or paper records will work against you
  • VAT filed late or returns filed late. Much more difficult to get around late payment
  • Systems read bank statements and will tie up payments to credit records, they will also look for warning payments such as fixed sum HMRC or PAYE payments
  • Incorrect details at Companies House. If the director details are wrong or the SIC code is incorrect it could work against you, as could the annual return or annual accounts being filed late. Do check Companies House as I have seen the incorrect sex of directors and incorrect dates of birth. With SIC codes, it doesn’t matter what you say you do, the SIC code takes preference
  • Annual accounts where the level of debtors and creditors is inaccurate. The algorithms will pick up the level of debtors and liabilities to work out cash generation

These systems are very clever and very accurate in the responses they generate. Where they can mislead is where your company is borrowing and that borrowing will make a material difference to the business. I appreciate that all borrowing should make a difference, however where there will be a material major shift in profits then the automated systems won’t see it. The systems are looking at now rather than looking forward.

If you are borrowing and it will majorly improve profits then you may be well advised to look at a lender that is not using these systems.

Either way, checking your details at Companies House and making sure your bookkeeping is current will get you a long way.

For any questions about this post or for any details on the best route to borrow for your business then please get in touch.

By Dave Farmer

let's talk business lending

[contact-form to=’hello@limeconsultancy.net’ subject=’Algorithm Underwriting’][contact-field label=’Name’ type=’name’ required=’1’/][contact-field label=’Email’ type=’email’ required=’1’/][contact-field label=’Phone’ type=’text’/][contact-field label=’Comment’ type=’textarea’ required=’1’/][/contact-form]

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.