Africa’s financial services sector is entering one of the most significant transitions in its history. Artificial intelligence, digital lending, behavioural analytics and open banking are rapidly changing how financial institutions assess borrowers, price risk and extend credit.
Yet amid all this advancement, one lesson stood out during the recent East African Banking School Conference held at Diani, Kenya: the future of lending will not be determined by technology alone, but by the ability to combine data, human judgement and responsible finance.
For many years, lending decisions largely depended on collateral, financial statements and the experience of credit officers. Today, those traditional indicators are increasingly being complemented by behavioural data, mobile money transactions, digital footprints and machine learning models. Financial institutions can now analyse thousands of data points within seconds to estimate the probability of default.
However, conference discussions repeatedly emphasised an important caution: algorithms should support lending decisions, not replace professional judgement. Every credit model should be explainable. If a bank or microfinance institution cannot explain why a customer was declined or approved, the institution risks embedding bias, weakening governance and exposing itself to regulatory and reputational challenges.
The quality of lending will therefore depend on the quality of data. Inaccurate, incomplete or outdated data inevitably produce poor credit decisions. Financial institutions must invest in data governance, ensuring that information is reliable, complete, timely and secure. Equally important is protecting customer data. In an era of increasing cyber threats and stringent data protection requirements, trust remains one of the banking industry’s most valuable assets.
Another important shift is the changing profile of Africa’s borrowers. With the continent’s median age below 20 years, Generation Z represents the next frontier of financial inclusion.
Yet many young borrowers own few traditional assets. They rent rather than own homes, rely on ride-hailing services instead of purchasing vehicles, earn income from informal or digital platforms, and conduct much of their financial lives through mobile phones.
This raises a fundamental question: Are financial institutions still lending using yesterday’s collateral for tomorrow’s customers?
Perhaps the industry should begin viewing behavioural consistency, cash-flow patterns, digital transaction histories and demonstrated financial discipline as forms of collateral alongside traditional security. Character and repayment behaviour may become valuable predictors of future credit performance.
Fortunately, financial institutions already possess significant amounts of customer data. Mobile money transactions, bank statements, bill payments, savings patterns, school fee payments and business cash flows all provide useful insights into customers’ willingness and ability to repay. The challenge is no longer collecting data, it is converting that data into better lending decisions.
Equally important is recognising that not all growth is good growth. Conference participants observed that smaller ticket loans often perform better than large exposures because they allow lenders to build borrower relationships gradually while limiting downside risk.
Progressive lending enables institutions to reward responsible repayment behaviour with larger facilities over time instead of taking excessive risks at the outset.
Technology is also reshaping partnerships across the financial services ecosystem. Increasingly, banks, fintech firms, credit reference bureaus, mobile network operators and data analytics firms are collaborating to improve customer acquisition, underwriting and collections. Products such as Fuliza in Kenya and Songesha in Tanzania illustrate how partnerships can expand financial access while creating sustainable business models.
Yet the conference also challenged lenders to rethink responsibility in digital finance. Digital loans have increased financial inclusion and generated attractive returns. However, they have also contributed to over-indebtedness among some borrowers, particularly where multiple lenders compete aggressively for the same customers. Responsible lending requires balancing commercial objectives with customer wellbeing.
One particularly striking observation was that many non-performing loans are created long before customers default. Weak credit appraisal, inadequate due diligence and poor underwriting decisions eventually translate into costly recoveries and write-offs. In many respects, collections merely reveal mistakes made during loan appraisal.
This reinforces the need for institutions to strengthen credit assessment rather than relying solely on aggressive debt recovery strategies.
The conference further highlighted climate change as an emerging source of credit risk. Following the economic disruption caused by the COVID-19 pandemic, climate-related events, including floods, prolonged droughts and other extreme weather conditions, are increasingly threatening household incomes, agricultural production and business continuity.
Financial institutions should therefore begin integrating climate considerations into credit appraisal, portfolio monitoring and stress testing.
Perhaps the greatest lesson from the conference was that successful lending will continue to depend on people. Artificial intelligence can analyse patterns. Algorithms can rank risks. Data can improve predictions. But trust, ethical judgement, customer understanding and professional scepticism remain fundamentally human capabilities.
For Kenya’s banks and microfinance institutions, the competitive advantage of the future will not simply lie in adopting more technology. It will lie in developing institutions that combine high-quality data, skilled people, responsible governance and customer-centred innovation.
Those institutions will not only grow healthier loan portfolios but will also become more attractive to investors seeking well-governed financial institutions capable of delivering sustainable returns.
The future of lending, therefore, belongs neither to algorithms nor to intuition alone. It belongs to institutions that successfully integrate technology with human judgement, innovation with responsibility, and growth with trust.
Professor David Mathuva is an Associate Professor, Accounting & Financial Markets at Strathmore Business School
