Since widespread internet adoption in early 2000s, various consumer businesses have been in constant race to adopt it into their core operations. Financial services sector lead this race - with first online banking account access in 1994 until making an online account opening a standard practice today. However, one of the most important consumer banking products, mortgage loan, was always seen as too complex or risky to be sold and processed fully online. Will we ever see it being fully automated?
Taking a step back, one might consider how looking for a real estate looks now. It is fairly digitized with various tools in place:
- search engines (e.g. Google Search)
- specialized search engines (e.g. Streeteasy.com)
- unstructured "bulletin board" type databases (e.g. Craigslist.com)
- credit score or mortgage offer aggregators
All of the above tools might be used to fully online research real estate market, find a wanted property and roughly understand one's financial ability to purchase it. The moment when one needs to "step out" of online world is when it comes to securing a mortgage. While almost all banks offer some kind of online pre-qualification for the loan, closing a mortgage requires delivering hard copies of documents (e.g. bank account statements, tax filings) and signing a mortgage agreement.
What would need to happen to make it fully online? Firstly, banks would need to utilize to a larger extent the data they already own (e.g. bank statement data), thus limiting the need to deliver hard copies of documents. Secondly, banks need to follow the path of many mortgage pricing fin techs and master their risk assessment models, leveraging newest AI and machine learning developments. Lastly, banks should consider non-client data (e.g. local real estate market condition) for broader risk assessment thus minimizing need for paper documents.
No comments:
Post a Comment