PH Banks' NPL Ratio Fell to 3.29%. Bad Loans Grew ₱54 Billion.

Key takeaways: Philippine banks' gross non-performing loan ratio fell to 3.29% in June 2026, a six-month low, from 3.44% in May and 3.34% a year earlier. Over those twelve months the peso value of soured loans rose 10.31%, to ₱584.943 billion. The ratio improved because lending grew faster than bad loans did — not because fewer borrowers deteriorated.

Both of those statements come from the same BSP data release, reported 27 July 2026. A credit committee reading only the first one would conclude the portfolio is getting healthier. A credit committee reading both would conclude something narrower and more useful: the portfolio is getting bigger faster than it is getting worse. Those are not the same finding, and they do not lead to the same monitoring decision.

What did the June 2026 BSP data actually show?

Four figures, all from the same release.

The gross NPL ratio stood at 3.29% in June, improving from 3.44% in May and from 3.34% a year earlier. It was the lowest reading in six months — since 3.07% in December 2025 — and matched the ratio recorded in March.

Soured loans fell 2.74% month on month, to ₱584.943 billion from ₱601.41 billion in May.

Year on year, soured loans rose 10.31%, from ₱530.292 billion. That is ₱54.65 billion of additional non-performing exposure carried by the banking system compared with June 2025.

The NPL coverage ratio — allowances set aside against potential losses — edged up to 92.53% in June from 88.92% in May, but slipped from 95.4% a year earlier.

Why does a falling ratio not mean falling risk?

The NPL ratio is a fraction. Non-performing loans sit on top; the total loan portfolio sits underneath. Both moved over the last twelve months, and they moved in the same direction.

The numerator grew 10.31%. The ratio still fell slightly, from 3.34% to 3.29%. That combination is only arithmetically possible if the denominator grew faster — and it did. The banking system's total loan book reached ₱17.78 trillion in June, up from ₱17.48 trillion in May and ₱15.88 trillion a year earlier: growth of nearly 12% over the twelve months.

So the banking system absorbed ₱54.65 billion of additional soured loans, and the headline ratio still improved, because new lending outpaced the deterioration. That is a real and reasonable outcome. It is also a fragile one: the ratio's improvement depends on loan growth continuing at that pace. If growth slows and the stock of bad loans holds, the ratio moves against you without a single new borrower defaulting.

For a credit committee, both numbers are true and they answer different questions. The ratio answers how does the portfolio look relative to its size. The peso figure answers how much money is currently not performing. Only the second one has to be provisioned, worked out, or written off.

What does the coverage ratio add to the picture?

Coverage at 92.53% is a healthy level by any reasonable standard, and the month-on-month move was upward. The twelve-month comparison is the one worth noting: 92.53% against 95.4% a year earlier means slightly less allowance per peso of bad loan, set against a bad-loan stock that is ₱54.65 billion larger.

Neither figure signals distress. Together they describe a system provisioning proportionately rather than defensively. That is a planning input, not an alarm — and it is exactly the kind of directional read that gets lost when a board pack carries one ratio and no context.

Why is MSME lending still at 4.73% against a 10% benchmark?

Total loans to micro, small and medium enterprises reached ₱574.8 billion as of end-March 2026 — 4.73% of the banking industry's ₱12.1 trillion loan portfolio net of exclusions, which is the base the allocation was measured against. That base is narrower than the ₱17.78 trillion total loan book above: it strips out interbank loans, repo transactions, loans under special financing programmes and other categories set out in BSP's Manual of Regulations for Banks. The two percentages are not directly comparable.

The Magna Carta for MSMEs (Republic Act No. 6977, as amended by RA 9501) required banks for ten years — from 17 June 2008 to 16 June 2018 — to allocate 10% of their loan portfolio to the sector: 8% to micro and small enterprises, 2% to medium ones. That mandatory allocation expired in June 2018 and has not been renewed, though the BSP still tracks MSME exposure through quarterly reports and the 8% and 2% levels remain the reference benchmarks. The industry has sat below the combined level for years, and the shortfall is almost entirely at the small end: micro and small enterprises drew ₱238.5 billion, roughly 2% against the former 8% benchmark, while medium enterprises drew ₱336.4 billion, or 2.8% — above their former 2% benchmark.

The usual explanation is credit appetite. The more precise explanation is unit economics of assessment.

BSP Circular No. 855 and the Internal Credit Risk Rating System requirements under Circular No. 439 already oblige banks to spread audited financial statements and run ratio and credit-risk analysis on borrowers. Philippine banks do this, and do it properly. The constraint is not whether the analysis happens — it is what the analysis costs per account.

A ₱3 million exposure and a ₱300 million exposure take broadly similar analyst hours to spread by hand. At the small end, the assessment cost consumes the margin. So the file gets reviewed once at origination, refreshed annually if at all, and the portfolio's small-ticket tail goes unwatched between cycles — not through negligence, but because manual review capacity is finite and gets allocated to the largest exposures first.

That allocation is rational per file and expensive in aggregate. It is also the mechanism by which a ₱54.65 billion year-on-year increase in soured loans accumulates quietly underneath an improving headline ratio.

What would a credit team change on this evidence?

Three things, in order of how quickly they can be done.

Report the stock next to the ratio. Any portfolio pack that carries the NPL ratio should carry the peso value of non-performing exposure and the year-on-year change beside it. This costs nothing and prevents the single most common misreading of an improving ratio.

Match monitoring cadence to deterioration speed. An annual refresh assumes a borrower's condition changes about once a year. It does not. Between two annual reviews there are four quarters of filings, payment behaviour and trade references that either confirm the original rating or contradict it. The gap between those two review dates is where the ₱54.65 billion accumulated.

Standardise the rating itself. Manual spreading varies by analyst, by workload, and by the week it was done. Two analysts reading the same audited statements can land on different risk grades and both defend the result. Consistency across the portfolio is what makes a re-rating comparable to the rating it replaces.

This is the specific problem CRDX was built for: spreading and rating a borrower's audited financials in minutes rather than hours, applying the same quantitative method to every file, so that re-rating a portfolio between annual cycles becomes a scheduling decision rather than a headcount decision. One BSP-supervised thrift bank we work with moved its small-ticket SME book from an annual refresh to a quarterly one without adding analysts, because the per-file cost of the refresh fell far enough to make the cadence affordable.

The reading that matters this quarter

June's 3.29% is a genuinely good number, and it deserves to be reported as one. It is also a ratio whose improvement was delivered by the denominator while the numerator grew by ₱54.65 billion.

The useful question for the next portfolio review is not whether credit quality improved. It is whether your bank would know if it had not — and how many months would pass before the file said so.

Sources: Bangko Sentral ng Pilipinas banking sector data, reported 27 July 2026 (NPL ratio 3.29%, soured loans ₱584.943B, coverage ratio 92.53%, total loan book ₱17.781 trillion); BSP MSME lending data as of end-March 2026 (₱574.8B, 4.73% of the ₱12.1 trillion loan portfolio net of exclusions); BSP Manual of Regulations for Banks Sec. 332 (mandatory MSME credit allocation, 17 June 2008 to 16 June 2018, and the exclusions defining the compliance base); BSP Circular No. 855 and Circular No. 439 (ICRRS); Republic Act No. 6977 as amended by RA 9501. Last updated: August 2026.