Network Effects: Why Financial Software Markets Keep Tipping Toward a Handful of Winners

When you log into a community bank’s online account software, swipe a card at a checkout counter, or open a Bloomberg Terminal, there is a high likelihood that you are interacting with one of the very few companies that dominate that specific area of financial software development. This is the predictable result of network effects: the same economic theory behind why most of a search engine, social network, or operating system’s revenues are produced by one of the major players. In the case of financial software, network effects combine with high switching costs and fixed costs to yield markets where you see the existence of a very limited number of entrenched incumbents rather than normal textbook competition.

Network effects explain how as an increasing number of users use a product, their experience with and value of that product increase; examples of network effect-type products include: telephones, eBay, Facebook, etc., as well as financial services platforms, as financial services are primarily about connecting users to users like traders to traders, banks settling payments with banks, and merchants accepting credit cards everywhere. Financial software development serves as the infrastructure layer through which these fixed costs and data-driven advantages are embedded into platform design.

The Bloomberg Terminal serves as an excellent case study of the network effect. Bloomberg controls approximately one-third of the global financial data market with over 355,000 users around the world and leases its terminals at an average rate of about $27,660 per year per user. While many competing products exist that provide similar raw data, such as LSEG’s Eikon, the majority of traders stick with Bloomberg rather than switching to other platforms because of features like Instant Bloomberg, the terminal’s built-in chat network, which allows traders to access other users already connected to the platform. Data can be replicated; networks cannot.

Core banking software provides an example of a somewhat related but different network effect. The Federal Reserve Bank of Kansas City analysis showed that three vendors—FIS, Fiserv, and Jack Henry—serve roughly 70% of U.S. banks. In this case, the concentration is more closely related to the switching costs rather than to direct network effects between the end-user and the financial services vendor. Core banking software contracts are negotiated for a minimum of 7 years and restrict a bank’s ability to transfer its accounting records, compliance reporting, and payment processing to another vendor because the financial institution is essentially tied to the vendor's infrastructure. Removing that vendor’s systems while still under contract would therefore be expensive and operationally risky, which in turn allows established vendors to increase prices without significant concern of customer attrition. 

Compounding both of these factors are the significant fixed costs associated with developing any new products in this industry. It can take many years of developing security-hardening, regulatory certification, and uptime guarantees before a financial software vendor can start to generate revenue from that product; while the amount of security, regulatory, and uptime costs does not significantly increase based on the number of clients/customers using the product, the established vendors can spread those costs over a much larger base of clients/customers, thereby allowing the established vendors to sell their products at much lower prices while maintaining significantly higher margins than any new vendors. Data inherently embodies a more subtle expression of this equivalent reasoning in terms of transaction volume. Payment network and fraud risk scoring providers continue to improve their ability to detect possible fraud and determine pricing for related risks as transaction volumes grow; therefore, the level of improvement is at a significantly greater rate for larger payment-processing firms due to their larger volumes of transactions, which, in turn, attracts the higher volume of transactions needed to continue their growth.

The impact of this explosion in transaction volume on the competition and innovation rate is showing itself through pricing and speed of innovation. For example, the Bloomberg terminal charges the same fee per user for access to its services as it did many years ago, while many companies now provide much lower-cost alternatives to Bloomberg data. So, while the price of Bloomberg data has remained unchanged, the barriers faced by community banks in launching new digital banking services due to the dependence on their vendors for any changes to be made continue to impede the ability of these banks to innovate and compete.

In recent years, many government agencies have also begun to view this situation as a more significant financial stability issue than simply a competition issue. An outage or price increase by any of the relatively few firms that operate as "the backbone" of the U.S. banking system would have repercussions beyond the firm’s direct customers. The Federal Reserve, CFPB, and other federal regulators have been scrutinising core-provider contracts and payment network provider practices much more closely than they have done for many years.