Why Structural Credit Risk Models Still Matter for African Lenders

The Problem with Traditional Credit Assessment

Most financial institutions across Africa rely on traditional credit scoring methods to evaluate whether a borrower will repay a loan. These methods which are often based on accounting ratios, financial statements, and historical default patterns have served lenders reasonably well for decades. However, they have two key limitations being that they only look backward and sideways resulting in a company measuring what it has done and how it compares to peers, not what it can do under stress.

The lack of forward-looking analysis was identified years ago by the International Accounting Standard Board which led to the introduction of IFRS 9, which is the accounting standard that requires banks, financial institutions and companies to calculate Expected Credit Losses on a forward-looking basis.

This becomes an issue in developing and emerging countries due to these countries having significantly higher level of volatility than developed countries. Financial institutions in developing and emerging countries face borrowers with irregular financial reporting, limited historical data, rapidly changing business conditions, and exposure to macroeconomic shocks that haven't been seen before.

Traditional scoring models struggle in these environments because they assume the future will look like the recent past—an assumption that often fails when exchange rates spike, commodity prices plummet, or regional instability hits.

What developing and emerging country lenders actually need is a way to assess credit risk on a forward-looking basis, one that penetrates the structure of a company's finances. They need to understand the relationship between a company's assets, its debts, the impact of adverse volatility and the probability it won't be able to meet those obligations.

What Are Structural Credit Risk Models?

Structural models approach credit risk from first principles. Instead of asking "What ratios predict default?", they ask a more fundamental question: "At what point do a company's assets fall below its liabilities?"

The foundation of modern structural modelling is the Merton Model, developed by economist Robert Merton in 1974. The insight is elegant: think of a company's equity as a call option on its assets. The shareholders own the company (the assets), but they have borrowed money (the liabilities) with the net of the two being the value of the equity. If the assets grow, shareholders benefit from an increase in equity value. However, if the assets fall below the debt obligation, the equity falls to zero meaning that the shareholders' option expires worthless, and the company defaults. The default doesn't happen at some arbitrary accounting threshold—it happens when asset value hits a mathematical trigger point based on the firm's debt structure and volatility.

In practice, Structural Credit Risk Models provide the analyst with functionality to infer what the company's total assets are and how volatile they're likely to be. These factors aren't always visible on a balance sheet, so the model derives them from equity market information (stock price and volatility) and accounting data (total liabilities). From there, it calculates the probability of default as a mathematical function of the distance between current assets and the default threshold. The wider that distance, the lower the default probability.

The power of this approach is that it forces you to think dynamically. Asset volatility matters—a stable utility company and a volatile tech startup can have the same equity value but very different default probabilities. Debt maturity matters too. A company with short-term debt obligations faces more immediate risk than one with debt that matures years away. Leverage matters—the ratio of debt to assets is fundamental to default risk.

Why Basic Structural Models Weren't Enough

The original Merton framework made simplifying assumptions that work reasonably well in mature, liquid markets but break down in more complex real-world scenarios. It assumed, for example, that default happens instantly when assets hit the threshold, with no recovery value. It also didn't account for the fact that borrowers might have multiple layers of debt with different seniority levels, or that some debt obligations are contingent on other events.

Over the past two decades, researchers and practitioners have extended these models to handle greater complexity. Hybrid approaches combine structural modelling (which captures asset dynamics and leverage) with accounting and financial ratio analysis (which captures earnings quality and operational health). Some modern variants introduce fractional recovery—the idea that when default occurs, creditors often recover something, not zero. Others allow the default threshold to vary rather than being fixed, reflecting the reality that management might choose to restructure debt before assets are completely exhausted.

These extensions matter because they bring the model closer to how credit actually works in practice. A company doesn't necessarily wait until assets mathematically hit zero; it might restructure or renegotiate earlier. Lenders don't automatically lose everything in default; they recover proceeds from asset sales. And the dynamics of default aren't purely mechanical, they're influenced by management decisions, market conditions, and legal frameworks.

Why This Matters for African Financial Institutions

Developing and emerging country lenders face a particular challenge that makes structural models especially valuable. Many borrowers operate in developing, emerging or frontier markets where:

• Data is less complete. Not all companies have lengthy credit histories or highly transparent financial reporting. Structural models don't rely solely on historical default frequencies—they derive default probability from the market value of equity and the company's leverage. For listed companies, this is more forward-looking and less dependent on sparse historical data.
• Business conditions change rapidly. Interest rates, exchange rates, commodity prices, and political stability can shift significantly within a loan's tenure. A structural model captures the impact of these changes directly through asset volatility. If a borrower's business becomes more volatile—say, a trading company in a region with currency instability—the model reflects that increased risk immediately, rather than waiting for actual defaults to appear in historical datasets.
• Market-based signals are available. For borrowers whose equity trades on regional exchanges, the stock price embeds a real-time assessment of asset value and volatility. Lenders can use this information to update their risk view continuously, rather than waiting for quarterly financial statements.
• Leverage dynamics are critical. In markets where debt maturity is often short and refinancing risk is real, understanding how a company's probability of default changes as its leverage ratio changes is invaluable. Structural models make this relationship explicit and quantifiable.

For smaller borrowers without traded equity, lenders can still apply structural concepts by using accounting data to estimate unobservable variables—a practice that brings the benefits of structural thinking even when market prices aren't available.

Modern Implementations: Blending Methods

The most effective credit risk systems today don't rely on structural models alone. Instead, they combine them with traditional statistical approaches. Financial ratios and accounting metrics capture important information about earnings quality, operational efficiency, and leverage that pure structural models might miss. By using both approaches together—what researchers call a "hybrid" framework—lenders get a more complete picture.

For instance, a company might show a high probability of default on a structural model because it's very volatile, but its financial ratios might suggest it's generating strong cash flows and deleveraging. The hybrid view lets a lender understand why the risk profile looks the way it does and make better judgments about whether risk is genuine or temporary.

This is particularly valuable in developing and emerging market lending, where the relationship between volatility, leverage, and actual default might differ from markets where models were originally developed. By combining multiple information sources—market signals, accounting fundamentals, macroeconomic variables—lenders can build risk models that are both theoretically grounded and practically robust.

How Cyte Applies These Insights

Platforms like Cyte use dual probability of default (PD) models that bring together structural and traditional approaches. The structural component captures the forward-looking implications of asset volatility and leverage. The traditional component anchors the model in accounting realities and historical performance. Together, these generate a PD estimate that reflects both the market's view of a borrower's financial health and the company's actual operational performance.

For developing and emerging country financial institutions, this dual approach helps answer the right questions: not just "What's the historical default rate for companies like this?" but "Given what we know about this company's assets, liabilities, volatility, and earnings, what's the probability it won't meet its obligations?" The first question is historical; the second is prospective. In volatile, data-sparse markets, the second question is often more useful.

Why This Matters Now

Structural credit risk models have been around for 50 years, yet they're more relevant today than ever. As developing and emerging country financial systems mature and integrate with global markets, borrowers face more complex financing structures, more volatile operating environments, and more pressure to refinance frequently. Traditional credit scoring, while still useful, isn't sufficient on its own.

The good news is that the theory is well-understood, the computational methods are efficient, and the frameworks are flexible enough to adapt to local market realities. Lenders who understand these models—and who use platforms that implement them properly—are better positioned to make faster, more accurate credit decisions even in challenging markets.

For developing and emerging country financial institutions aiming to grow responsibly and compete globally, understanding structural credit risk models isn't an academic exercise. It's a practical necessity.