Bitter harvest: The rise of bad loans in the agriculture sector

THe RBI released its biannual Financial stableness cover (FSR) on dec 31. Information on badness loans, or non-performing assets (NPA), shows that bank balance sheets are healthier than they have ever been in the last decade, with gross NPAs as a share of total advances falling to just 2.2% in September 2025.However, there is one statistic in the FSR which should set off alarm bells for policy observers. Agriculture accounts for the highest share of bad loans and this share has been rising for quite some time now. A further examination also shows that agriculture’s share in bad loans is disproportionately larger than its overall share in bank credit in the country.The Centre for Monitoring Indian Economy (CMIE) database has sector-wise NPA share data from the quarter ending March 2015. Back then, industry and services accounted for more than 80% of the total NPAs of scheduled commercial banks (SCBs). The situation remained broadly unchanged for five years but started changing steadily after that.As of September 2025, the latest period for which this data is available, industry and services accounted for only 45% of NPAs while more than 50% of it is now in the agricultural and retail loan category. Agriculture alone accounts for 36% of SCB’s NPAs, the highest this number has ever been for the period for which we have this data. The NPA share of retail loans (17.7%) is almost the same as that of industry (20.1%). NPAs in services have been flat in the recent past after falling and then rising again in the post-2015 period. This trajectory has also meant that the share of “large borrowers” – they are defined as one who has aggregate fund-based and non-fund-based exposure of ₹5 crore and above with any bank – in total NPAs has fallen from almost 90% in 2015 to just about one-third by September 2025. (See charts 1A, 1B) What is troubling about farm NPAs is the rise in their relative shareA particular sector’s rising share of NPAs on its own need not cause a concern. It can very well be a result of an increase in credit to that particular sector. This would mean that even for the same probability of a particular loan ending up as a bad loan, the sectoral NPA share can see an increase because of rise in overall credit. However, this is not the case for the current rise in agricultural NPAs in India. Agriculture’s relative share in NPAs – its share in NPA divided by its share in total advances – has seen a very sharp rise in the past four-five years and is significantly higher than other sectors.Retail loans, on the other hand, show the lowest relative share in NPAs which suggests that their rising share in overall NPAs is more a function of increase in credit allocation rather than distress or delinquency. That personal loans have seen a steadily rising share in overall credit in India is a well-known fact. It is important to underline that not all bad loans are the result of delinquent behaviour which is more likely to apply to the category of wilful defaulters. Which of these two factors is driving agricultural NPAs is difficult to say at the moment from publicly available data. (See chart 2) Not all loans are to very poor farmers, but most are of smaller ticket sizeWhat class of agricultural credit takers are driving the rising share of NPAs in agriculture? Because the RBI does not publish a cross-tab of sector and ticket-size of loan for NPAs, we cannot answer this question conclusively.However, we can check the overall break-up of agricultural loans by ticket size and see how it compares with other sectors. Not all agricultural credit by banks is meant for the proverbial poor farmer, but the share of smaller ticket-sized loans in agriculture is larger than it is in industry and services. Even in the case of personal loans, the share of credit disbursed with a ticket size of less than ₹10 lakh is much lower—30% of total personal loans advanced—than in agriculture (74%). This would suggest that smaller ticket-sized agricultural loans are bound to have contributed more to the rising share of agricultural NPAs in the country. (See chart 3) Does this signal growing agrarian distress? Are these factors structural or cyclical? Will it lead to greater political pressure for yet another farm loan waiver, something India has seen in the past? These are all questions which require more granular data than is available in the public realm to get conclusive answers. Given the importance of agriculture for mass livelihoods and the political pressure a crisis in this sector can generate for loan waivers, economic policy must take urgent notice of the situation and get to the root of the problem.