पिछले कुछ वर्षों में एनपीए तेज़ी से बढ़े हैं और यह प्रवृत्ति जारी है। भाजपा बैंकिंग क्षेत्र में एनपीए घटाने के लिए आवश्यक क़दम उठाएगी। साथ ही, भाजपा निवेशकों की रक्षा के लिए ग़ैर-बैंकिंग वित्तीय कंपनियों हेतु एक मज़बूत नियामक ढाँचा बनाएगी।

The 2014 manifesto's line on bad loans is short and correctly diagnosed: "NPAs have increased sharply over the past few years and the trend continues. BJP will take necessary steps to reduce NPAs in Banking sector." It adds a second commitment on non-banking financial companies: "a strong regulatory framework… to protect the investors."

The number to test against is the gross non-performing assets ratio of scheduled commercial banks, and the Ministry of Corporate Affairs quoted the Reserve Bank's June 2026 Financial Stability Report to the Lok Sabha on 20 July 2026: the ratio "has declined from its peak of 11.2% in March 2018 to a low of 1.8% in March 2026."

Read the dates. The peak was in March 2018 — four years into the term that promised reduction. NPAs did not fall during the 2014-19 term; they roughly doubled and then began to come down only at the very end of it.

That rise, though, is not a straightforward failure, and this fact-check will not present it as one. From 2015 the Reserve Bank ran asset quality reviews that forced banks to classify as non-performing a large stock of loans they had been evergreening. Much of the increase from 2014 to 2018 is recognition of bad debt that already existed, not new bad debt created. Cleaning a balance sheet makes the number worse before it makes it better, and a government that allows that to happen is doing the harder thing.

What followed is the part the promise was aiming at. The Insolvency and Bankruptcy Code, enacted in 2016, gave creditors a time-bound process for the first time: "The IBC has consolidated all these fragmented laws and resolved 1419 companies as on 31st March 2026. These cases have led to a recovery of ₹4.32 lakh crore to the creditors." Public sector banks were recapitalised on a large scale from 2017. The ratio fell to 1.8 per cent by March 2026.

Two cautions on that number. It is a ratio, and it falls when the numerator is written off as well as when loans are recovered — the answer reports no write-off figure, and write-offs have been substantial. And IBC realisation itself is modest against admitted claims: the year-wise table shows realisation by financial creditors at 54 and 55 per cent in the first two years, then 27, 27, 24, 39, 28, 37 and 20 per cent in 2025-26. Creditors typically recover a fifth to two-fifths of what they are owed; the recovery is real, and so is the loss.

On the NBFC half of the promise, the framework was substantially rebuilt — scale-based regulation, tighter liquidity norms, and a resolution framework — but that came after the IL&FS failure of 2018, which is to say after the risk the promise was made about had already materialised.

A correctly diagnosed problem, a term in which the measured number got worse for defensible reasons, and a recovery that arrived in the following term. Our verdict is partly accurate.