We’re in a time of greater ever choice for our everyday current accounts. New entrants from traditional brands including the Post Office and a growing band of challenger banks, including Metro Bank and OakNorth, are opening up competition in the market that consumers are beginning to take advantage of.
However, the greatest competition for our traditional banks may be lying in wait elsewhere – online.
Companies have a huge number of challenges to consider when entering the personal banking space. Not only do they have to be able to raise the capital required to be able to lend, invest and meet regulatory requirements, but they also need to be able to set competitive interest rates for loans and savings (and have the technology and know-how to do so in a sustainable and profitable way).
Finally, and crucially, they need to able to attract not just customers, but the right sort of customers – people who will repay their debts but also be profitable for the bank.
Banks decide who to lend to and what interest rates to set by using complex mathematical and statistical models. The more accurate the models, the fewer defaults on loans, the more precisely interest rates can be matched to the risk of defaults, and the more attractive banks can make themselves to credit worthy customers.
Those who have less accurate models risk falling behind, and losing out to those more able to fit their offer to customers’ needs.
But there might be a bigger issue on the horizon for our banks than simply competition from more recent market entrants like Tesco and the Post Office. Traditional banks also no longer have a monopoly on modelling expertise, or access to serious funding.
Tech companies including Amazon, Alibaba and Paypal are all reportedly exploring SME lending. If these companies are looking seriously at lending as a new market opportunity, you can be sure another major player is. Above all others, banks need only to look to their search engine to find possibly the biggest threat to the status quo: Google.
The last reported balance sheet for Google suggests cash and equivalent assets of over $18.3 billion, and investments that could be turned into $46 billion – considerably more than the shareholders’ equity of several established UK and US banks.
That’s a lot of money to start a banking business with. What's more, Google’s search engine is based on extraordinary modelling expertise and data processing capacity.
In a sector where predictive power is king, Google already has a huge amount of data about us they could translate into accurate – and attractive – interest rates for consumers.
A survey from uSwitch last month found that a quarter of millennials say they'd rather bank with Google than with their current bank. It seems the writing is already on the wall, and banks need to act now.
Fortunately, there are ever-better models being developed that banks can use to stay ahead of the game, and ensure they can remain competitive in the face of these new challengers.
The 14th biennial Credit Risk and Credit Control Conference, held at the University of Edinburgh Business School this week (26-28 August), is set to welcome some of the world’s greatest experts on credit risk to unveil the newest and most exciting thinking in the field.
One of the developments being unveiled is the intensity model, developed by Professor Jonathan Crook and Dr Mindy Leow at the University of Edinburgh Business School.
It could provide the key to the ‘next generation’ of credit risk modelling, protecting against institutional instability, optimising portfolio performance and allowing more accurate provision calculations, as well as paving the way for more accurate personalised consumer interest rates.
The model is already attracting attention from the international banking community and may have an impact on consumer interest rates as well as financial institutions’ stability and profitability in the future.
The intensity model is a leap forward compared to current risk modelling used by banks. It not only provides a way to assess the probability of a customer defaulting in any given month rather than just the chance of default any time in a 12 month period, but it could also be used to assess when a customer is likely to miss a payment even if they do not go into default.
Beyond that, the model also predicts the chance of various degrees of ‘cure’ in each month in the life of a loan. As a result of this, banks using the model may be in stronger a position to tailor interest rates to consumers based on far more accurate probabilities of late payments, repayments or defaults, giving them a competitive edge.
The model is likely to become highly relevant very soon. It will help financial institutions meet the requirements of a new accounting standard being introduced in January 2018, IFRS9, which asks banks to calculate the present value of credit losses expected in the future. It provides a huge opportunity for lenders to identify the capital at risk from defaults and funds lost due to further missed payments.
And it could also help banks more effectively stress-test their systems by predicting delinquency and default levels in the event of another credit crisis, to a more accurate degree than ever before.
This really is the next generation of risk modelling. As new players enter the market, whether they are new challenger banks or tech companies like Amazon and Google, with cutting-edge predictive technology at their disposal, employing these methods will enable banks to retain a competitive edge as well.
Those institutions most able to entice the best customers with competitive rates could come off best in this new market environment.
Sourced from Professor Jonathan Crook, director of the Credit Research Centre, University of Edinburgh Business School