The term “fintech” refers to several different business models that are nested together by their relationship with financial services.
Four core business models have gotten most of the traction with investors and customers. Some are more advisable than others.
Monetize transaction data
Here, the idea is to monetize transaction data.
Often, it is consumer data, like credit card processing data or insurance data.
While these data sets can be valuable, it is not a good place for startups to play in most cases.
Getting usable transaction data requires processing a large volume of transactions, and large payment processors, insurance firms, and the like are more than aware of the value of their data at this point.
The multi-sided marketplace
The multi-sided marketplace idea is to “create a stock exchange for ___,” and so automate an intermediary function on a digital platform.
I go in-depth on this in another article. While it is not a bad model if it works, getting this kind of exchange platform off the ground is a massive undertaking.
Particularly for derivatives or forex traders, who tend to be less gullible than consumers.
Move fast, break stuff, and deal with regulators later
Airbnb, Uber, and Lyft have built large companies doing this.
Unfortunately, it doesn’t work so well in finance, which tends to be regulated at a national rather than a local level.
Moreover, the rationale behind regulations in finance is often much more robust than in, say, cab-hailing. Financial regulators are less likely to be intimidated by press releases claiming that their silly rules obstruct innovation.
Regulators are already biased against smaller financial firms because they’re expensive to surveil and don’t offer much in the way of career prospects, so a business model built on rule-breaking doesn’t work.
Invent a better pricing or risk model than big institutions
Data scientists often assume that, because large banks’ IT infrastructure is often stuck in the 1990s, they can use their superior technical skills to enter the lending, insurance, or credit card space and make better bets than the big banks.
This is rather stupid. While banks may not have the best digital infrastructure, the industry has invested billions in developing algorithmic risk management systems.
Those systems are either quite robust from the volume of transactions executed, or discarded because they didn’t live up to expectations (usually, the latter).
The biggest flaw is that risks in things like lending and insurance tend to occur from segmentation. The theory behind these startups is they can segment and make better loans without the risk pooling of a diversified financial institution.
I haven’t looked at the empirical data in-depth, but it seems like a poor idea.