In 1890 Jacob Riis published “How the Other Half Lives,” an indictment of the horrific tenements of New York that lead to a significant housing reform movement. I recently concluded the reading of a book titled, “How the Other Half Banks” by @ an indictment of the U.S. banking system excluding a significant portion of the U.S. population from traditional financial services. Full disclaimer, the author is a friend of my sisters but she was nominated for a Pulitzer Prize so I am not just referencing the book for the aggrandizement of a family friend (frankly, I am not sure I have ever met her). The book is about financial inclusion. In the United States almost 28% of the population is unbanked or underbanked. Unbanked meaning that a person doesn’t participate in the financial system at all, operating solely in cash, and underbanked meaning the person participates in non-traditional financial products such as check cashing services, loan sharks, and pawn brokers. “The answer to the implicit question contained in [Baradaran’s] title, ‘How the Other Half Banks,’ is simple: The ‘other half’ hardly banks at all.” (Nancy Folbre, NYT)
We could dive into significant detail on how we have managed to let this happen in the U.S., but suffice it to say that de-regulation of the banking industry allowed banks to ignore potential banking customers that would not produce profits for the bank. If you put your businessperson’s hat on, you may think that this makes sense; customers that don’t make money typically aren’t foundational to a sustainable business. So, why should a bank see customers any differently? Herein lies the rub. Baradaran argues that bank and country are inseparable. This was made particularly evident in the 2008 financial crisis wherein the U.S. Federal Government committed $16.8 trillion to relief in the wake of mass mortgage defaults and financial instruments tied to mortgages going toxic with gargantuan piles of money doled out to the biggest banks to prop them up. Clearly, the government has an interest in making sure that the banking system doesn’t fail.
In wake of the crisis there has been a significant increase in bank regulation. The question becomes, is the government support and regulation of the nations largest banks designed to serve the best interests of our nations people? With this level of support could not policy makers insist on the creation of products that meet the needs of our nations underbanked and unbanked?
Interestingly, for some time there was an alternative banking system (by today’s standards) that assisted this segment of the population; and it functioned quite well. Postal banking existed in the U.S. from 1871 to 1966. You guessed it, your local post office doubled as a bank. Postal banking very well could be a solution for the underbanked and unbanked. “…the post office inspector general’s office, a small regulatory branch of the post office, issued a White Paper report in January 2014 proposing” postal banking for the present. But “…the postmaster general has not publicly supported the proposal and no congressional committee has seriously considered postal banking.” (Baradaran) Thinking purely from a business perspective, with a public/private joint venture, the existing post office infrastructure could provide a significant opportunity to offer right market fit products for the underbanked and unbanked. I mention this because if the right people get thinking about solving this problem, even if this post is just something that sparks further research, a solution means significant improvements to quality of life for almost 90 million Americans.
I’m going to make a serious transition here. The book got me thinking about payday loans.
Ideas are in the ether and we can pull them down and do something with them or wait until someone else does. I’m going to claim pulling this one down (although I may not be the first) and put it out there into cyberspace for the thinkers and executioners of the world.
Machine learning is a huge part of the direction of technology based businesses at the present. Machine learning refers to a subfield of computer science that studies pattern recognition and computational learning in artificial intelligence. Translation: your computer learns things about you without anyone specifically instructing it to do so. An example would be that you are shopping on Amazon and you shop for a pair of running shoes. Amazon delivers up a list of products that were shopped for by others who shopped for this same pair of running shoes. See below:
This list is customized by what Amazon’s machine learning algorithm knows about you. These suggestions often get you to buy more. Amazon leverages data and machine learning to make your shopping experience more efficient and pleasant.
Think about payday loans in the context of machine learning. Payday loans as they exist today are predatory by almost any sane measure. Interest rates can be up to 1900% per year. What if you created a machine learning algorithm that assessed risk of default on payday loans? You could use a FICO score as a baseline and build on top of this (but I would imagine that poor FICO scores are fairly common among the underbanked and unbanked, this being said, you could add the factors that FICO uses into your algorithm). Factors that could be included in the machine learning algorithm could be:
- job stability
- marital/family status
- level of education
- browsing history information
- keystroke efficiency measurement
You could even create an algorithm that measured the potential client’s reason for the loan to help evaluate risk. This list is by no means exhaustive but hopefully it serves to illustrate the thought process.
For the record, companies are working on the above right now. This is not new thinking.
Perhaps the business model is new: Your client registers online and provides all of the above information during the application process. You as the business owner have contractual relationships with ATM (Automatic Teller Machine) networks for integration of your product. In other words, if you approve a client for a payday loan (perhaps from their mobile phone) they can then go to the closest network ATM to retrieve their loan using a 16 digit code you text them upon approval. Eliminating the need for brick and mortar payday loan locations reduces operational overhead significantly.
Now, for the social good element. Each time a client repays their loan as agreed, any subsequent loan accounts for the previous repayment of a loan. The risk profile improves and as a business you can make a loan to this same client with greater confidence that it will be repaid. The typical payday loan starts at almost 400% annually, but by employing an efficient risk management algorithm a business may be able to cut this rate in half on the 2nd loan, maybe to a quarter on the 5th loan, and so on. This allows a person in a tough situation to dig themselves out rather than perpetually be in debt to ruinous payday lenders.
There you have it in brevity.
What do you think about this business idea? What do you think about postal banking?Leave a comment below.