What can game theory teach us open data and finance?
Hi,
What are the fundamental elements of success for humans as a species?
Obviously, not an easy question to answer. However, I’d like to think I’ve arrived at one- cooperation and coordination. Interestingly, I revisited von Neumann’s seminal paper ‘On the theory of game of strategy’ this week; the paper that made game theory an invaluable part of our understanding of choices and consequences in economics, politics, biology, and computer science, you name it.
The idea that the world largely operates in a zero-sum/non-zero-sum mindset is fascinating. That with a zero-sum mindset, banks were seen as antagonists, or fintechs were seen as eating banks’ lunches. In a non-zero-sum mindset, an open data economy would thrive, everybody wins and nobody loses - coordination and cooperation.
Recently, a slew of banks jumped on the Open Network for Digital Commerce (ONDC) bandwagon. For one, the massive amounts of data that will be generated in the network, ideally, should make ONDC a non-zero-sum game where everybody benefits from the network. The same data, banks believe, will aid them in offering and cross-selling more products, fostering efficient credit decisions and innovative lending products. Makes sense then to buy a stake in ONDC, right?
‘The technology is a commodity, the power is in the network”
Except that I think most banks are still struggling to correctly harness the massive amounts of data they’ve amassed in the last couple of decades. For one, most banks don’t have an API-driven strategy and two, building a sustainable data utilization strategy in the age of AI is easier said than done.
APIs have transformed banking, but not banks
A 2020 Mckinsey survey on APIs in banking revealed that only roughly three-quarters of banking APIs are used for internal purposes, and banks plan to double the number of internal APIs by 2025.
The same survey also showed that reducing IT complexity through internal APIs was the top objective, followed by enabling agility and partners. Sadly, ‘innovation’ only made fifth place.
Source: McKinsey
When we talk about digital transformation, we talk about filling the customers’ need for a full-stack solution. Before the API explosion, the only way to deliver this full-stack was with proprietary service offerings and formal relationships between firms that determined who was or wasn’t in the service stack.
In an API world, competition and differentiation are about speed, agility, and personalization. And that’s at the core of digital transformation - it has everything to do with banks’ ability to deliver products and services and little to do with consumers’ channel behaviours or preferences.
A study from the European Corporate Governance Institute says - “Developing the technological architecture necessary to connect to and, importantly, compete in an Open Finance ecosystem can be extremely costly. While a small handful of large banks and other financial institutions may be in a position to absorb and amortize these costs, they will often be prohibitive for many smaller banks and fintech disruptors—thus creating a potentially significant barrier to entry.”
And this is where the problem is, only the largest banks have the resources to implement and execute API integration, testing and compliance, and the legal and contractual reviews necessary. And even with third-party APIs, they should be able to continuously evaluate whether a vendor’s API can help them support a ‘superior customer experience’
Getting the API strategy right would then require banks to a) clearly define the differentiated customer experience and products they want to offer and b) continue to focus on APIs that enable them to connect ecosystems to deliver these differentiated offerings.
In India, API Banking took off only recently with Yes Bank and RBL Bank taking the lead in 2017. Most significant players like ICICI Bank, Axis Bank, Kotak Bank, and Smaller Players like SBM, AU SFB, and Equitas SFB have now established comprehensive API banking platforms. The common thread between these banks was a focus on the monetisation of APIs with multiple business partnerships.
It boils down to this - are banks willing to move their legacy business models to compete in an API-based world?
Building a sustainable data utilization strategy in the age of AI
Consider Tesla - they haven’t been great at delivering fully autonomous cars and Tesla’s hardware (for cars manufactured after 2016) doesn’t come close to the ‘level 5 autonomous driving as defined by the society of automobile engineers. But it's the only company anybody can think of when they hear of self-driving cars. Why? They harnessed the power of data using AI much before anyone else did - they recognized that in order to get to level 5 autonomous driving, they need to train their models from the get-go. With over 500,000 cars on the road, Tesla has ensured each of these cars has the technology to passively capture data continuously; the data necessary to train self-driving algorithms.
Tesla’s doing it in automotive, Amazon in tech and Netflix in show business - forward-thinking companies usually have a strategy to build sustainable competitive advantage by leveraging data and AI.
How can financial services do this?
Getting rid of internal silos - Lending to MSMEs is the prime example. I wrote about how private banks are getting it right a couple of weeks back. Essentially, leveraging transactional data to proactively identify clients that may need extra liquidity and making it instantly available is the key in using data better
Enriching data - Much like Tesla, banks can continually mine customer behaviour. Especially in the ONDC network with a layer of Open Credit Enablement Network (OCEN), every interaction will be digital. From capturing typing rhythm to the angle at which they take a selfie to their gait, data can be captured and fed into algorithms to do all kinds of interesting stuff.
Getting more data - Proprietary data coming straight to your platform is one thing, but the account aggregator framework allows for the movement of data across parties. Banks should leverage data from sources outside of their own platform to really, holistically understand consumer behaviour. The linking of proprietary data with anonymized, aggregate-level data from other sources for competitive benchmarking and strategy development could put banks on a sustainable growth path.
Rejigging your business model - A sustainable strategy involves going beyond the low-hanging fruits and making difficult choices. Generating proprietary data that can be harnessed to train machine learning algorithms will not be an investment with instant returns. A broader time horizon is necessary to change your growth strategy, which might, on the face of it, appear like a bad bet. Think about mortgages - imagine how different mortgage servicing would be if banks looked at it not just as a fee-generating business, but as an opportunity to gather data on behaviours and aspirations of a household over long periods of time!
An oft misquoted statement, owed to Charles Darwin (he didn’t say it), and one that’s used in innumerable business presentations sums up the non-zero-sum game that banks need to play - “It is not the strongest of the species that survives, not the most intelligent that survives. It is the one that is most adaptable to change”
And that’s the idea of an open data economy, of the likes of ONDC. It's a grand idea made to suit the smallest of players but also advantageous to every player involved. A win-win. In order for banks to win too, they need to adapt to technology that’s evolving at full tilt and cleverly mobilize and channel data.
See you next week!
Cheers,
Rajat