NPL Resolution and Credit Channel Recovery in Azerbaijan: An ARDL Benchmarking Analysis Against the KAMCO Model
This paper employs an ARDL framework to quantify the long-run negative elasticity between non-performing loans and credit volume in Azerbaijan following the 2015 currency devaluation, revealing a slower recovery trajectory than the South Korean KAMCO model and proposing a three-pillar policy strategy involving a contingent asset management company, legal reforms for faster collateral enforcement, and a structured link between NPL disposal and SME lending.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the world of banking, money is supposed to flow like water, moving from those who have it to those who need it to build homes, start businesses, or buy goods. But sometimes, the pipes get clogged. When borrowers cannot pay back what they owe, these bad debts pile up, creating a blockage that stops new money from moving. This blockage is known as a non-performing loan. When too many of these bad loans accumulate, banks become afraid to lend again, even to people who are perfectly safe to borrow from. This fear creates a "capital crunch," where the entire economy slows down not because people stopped wanting to buy things, but because the banks have run out of the confidence and capital needed to say yes. Understanding how to clear these clogs is vital for any nation, because the speed at which a banking system recovers determines how quickly a country can return to normal life after a financial crisis.
In Azerbaijan, this clogging happened with startling speed and severity. For years, the country's banks lent out money in foreign currencies like the US dollar and the euro, while the people and businesses borrowing that money earned their income in the local currency, the manat. This arrangement worked fine as long as the exchange rate stayed steady. However, when global oil prices dropped and the value of the manat suddenly fell by 34 percent in early 2015, the debt for borrowers effectively doubled in value overnight. A loan that cost 7,800 manat suddenly became worth over 15,500 manat. The result was a wave of defaults that pushed the ratio of bad loans to 13.8 percent by 2017. In response, the banking system contracted sharply, with total credit shrinking by nearly half and 26 bank licenses being revoked. The question that researchers sought to answer was not just what happened, but how the relationship between these bad loans and the total amount of credit in the economy actually works, and whether Azerbaijan's path to recovery was as fast or effective as it could have been.
A team of researchers from universities in Azerbaijan and Turkey set out to map this relationship using a detailed record of the country's banking activity from 2010 through early 2024. They gathered 57 quarterly snapshots of the economy, looking at the total volume of credit, the percentage of loans that were not being repaid, the size of the non-oil economy, and the exchange rate. Instead of simply describing the events, they used a statistical method designed to find long-term patterns even when data is messy or incomplete. This approach allowed them to separate the permanent effects of bad loans from temporary fluctuations and to measure exactly how much credit would likely return if the bad loans were cleared away.
The analysis revealed a clear and strong connection between the health of the banks and the flow of credit. The researchers found that for every 1 percent drop in the ratio of bad loans, the total amount of credit available in the economy tends to rise by about 0.65 percent in the long run. This suggests that cleaning up the banking books is a powerful way to restart lending. However, the study also uncovered a critical detail: the exchange rate was actually a bigger driver of the credit collapse than the bad loans themselves. The sudden devaluation of the currency was the primary shock that broke the system. This means that even if a country successfully fixes its bad loans, it will not see a full recovery in lending unless it also addresses the mismatch between the currency banks lend in and the currency people earn in.
The speed at which the system corrected itself was another key finding. The researchers calculated that when the economy was out of balance, about half of the gap between the actual credit level and the ideal level was closed within a single quarter. While this sounds efficient, the study suggests it is actually slower than what happened in other countries facing similar crises. To understand this, the researchers looked at how South Korea handled a comparable crisis in the late 1990s. South Korea used a centralized, state-run agency to buy up bad loans quickly, price them fairly, and sell them off, which allowed their credit system to recover rapidly. In contrast, Azerbaijan's approach was fragmented, with different state entities handling parts of the problem without a unified strategy or market-based pricing. The data indicates that Azerbaijan's recovery was weaker and slower than the South Korean model, likely because the solution was not as systematic or well-organized.
Based on these findings, the authors propose a three-part plan to help Azerbaijan and similar economies prepare for future crises. First, they suggest creating a pre-designed legal framework for a specialized agency that would only activate if bad loans rise above a specific threshold, such as 8 percent. This agency would be run like a business, insulated from political interference, and would buy bad loans at market prices to ensure they are sold quickly. Second, they argue for legal reforms to speed up the process of seizing collateral from borrowers who default. Currently, it takes 18 to 24 months to enforce a claim in court; shortening this to 6 to 9 months would make it much easier for banks to recover their money and lend again. Finally, they recommend linking the money recovered from selling bad loans directly to new loans for small and medium-sized businesses. This would ensure that as the banks' balance sheets are repaired, the new money immediately flows to the parts of the economy that need it most to grow.
The study concludes that while Azerbaijan has made progress in reducing bad loans, the path to a fully recovered banking system requires more than just cleaning up the books. It demands a coordinated approach that addresses currency risks, speeds up legal enforcement, and uses a centralized, market-driven mechanism to handle distressed assets. The researchers are careful to note that their findings are based on observed patterns and comparisons with other countries, rather than a perfect simulation of cause and effect. They acknowledge that the data is limited and that the specific conditions of each country's crisis differ. However, the evidence strongly suggests that the way a country organizes its response to bad loans matters just as much as the severity of the crisis itself. By learning from the faster, more organized recoveries of the past, Azerbaijan can build a more resilient financial system for the future.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.