Thrift Banks Are Growing Faster Than Their Credit Infrastructure Can Handle

Philippine Thrift Banks Are in a Growth Squeeze

Thrift banks in the Philippines have been growing SME lending portfolios aggressively. BSP data shows the sector expanded commercial loan books by double digits through 2025, driven by government policy encouraging financial inclusion and the structural retreat of universal banks from smaller borrower segments.

That growth came with a quiet tradeoff: most thrift banks added borrowers without adding proportional credit monitoring capacity.

The result is a sector where loan officers are stretched across larger portfolios, annual review cadences have not changed despite portfolio growth, and early warning signals for borrower deterioration arrive too late to prevent classification downgrades.

The Structural Problem: Manual Review Does Not Scale

A thrift bank with PHP 8 billion in SME loans and 400 active borrowers can manage annual reviews manually. The same bank at PHP 14 billion with 700 borrowers cannot — not without either hiring significantly more credit staff or accepting that some portion of the portfolio goes unmonitored between review cycles.

Most thrift banks chose a third option: keep the same team, prioritize the largest exposures, and hope the smaller borrowers stay healthy between reviews.

BSP loan classification rules require banks to assess borrower capacity to service debt on an ongoing basis. The expectation is timely detection — not discovering the problem during the next scheduled annual review when the borrower is already 90 days past due.

For thrift banks operating with lean credit teams, "timely detection" requires infrastructure that does not currently exist at most institutions.

Three Patterns Showing Up Across the Thrift Banking Sector

Pattern 1: Portfolio growth outpaced hiring.

Between 2023 and 2025, many thrift banks grew loan books 40-60% while credit teams grew 10-15%. The math does not work. Each officer is now responsible for more borrowers, which means less time per borrower, which means more reliance on lagging indicators (missed payments) rather than leading indicators (financial deterioration before default).

Pattern 2: Annual review cycles create a 12-month blind spot.

If a borrower's financial condition deteriorates in February and the annual review is scheduled for November, the bank has nine months of exposure to a credit that should have been flagged. For a PHP 50 million exposure, that is nine months of accruing interest on a loan that may need to be written down.

The BSP does not prescribe a specific review frequency. But examiners ask: "When did you know?" If the answer is "at the next annual review, nine months after deterioration started," that is a governance finding.

Pattern 3: Board reporting relies on stale data.

Thrift bank boards receive portfolio quality reports based on the last completed review cycle. If only 30% of the portfolio has been reviewed in the current quarter, the board is making risk appetite decisions based on data that is 6-12 months old for 70% of exposures.

This creates a specific governance vulnerability: the board approved a growth strategy based on portfolio quality metrics that did not reflect current borrower conditions.

What Continuous Credit Intelligence Changes

The thrift banks addressing this gap share a common approach: they separate the monitoring function from the review function.

Monitoring is automated, continuous, and covers 100% of the portfolio. Financial condition ratings are computed from submitted financial statements. Deterioration signals are generated when key ratios breach thresholds. The credit team receives alerts rather than conducting manual surveillance.

Review remains human, judgment-based, and triggered by signals rather than calendar dates. Officers spend time on borrowers that need attention rather than cycling through the entire book on a fixed schedule.

The practical difference: instead of 400 annual reviews consuming 1,600 officer-hours per year, the monitoring system flags the 40-60 borrowers showing stress signals. Officers conduct deep reviews on those credits and periodic validations on the rest. Total officer time decreases while coverage increases from annual to continuous.

BSP Examination Readiness

BSP examiners evaluating thrift banks are increasingly asking about monitoring infrastructure — not just policies, but systems. Specifically:

• How does the bank detect borrower deterioration between review cycles?
• What is the average time between onset of financial stress and bank awareness?
• Can the bank demonstrate that classification decisions are based on current data?
• Does the monitoring system cover the full portfolio or only top exposures?

Thrift banks that can answer these questions with system-generated evidence — automated ratings, timestamped deterioration alerts, portfolio-wide coverage reports — have materially different examination outcomes than banks relying on manual review logs.

The CTB Convention Question

The Chamber of Thrift Banks Convention on July 15 will bring together executives who are all facing the same growth-versus-infrastructure tension. Every thrift bank in the room grew their loan book over the past two years. Very few added monitoring infrastructure proportionally.

The question is no longer whether continuous credit monitoring is necessary. BSP expectations, portfolio growth rates, and lean staffing models all point in the same direction. The question is how to implement it without hiring 30% more credit staff — which is what manual continuous monitoring would require.

CRDX provides the credit intelligence infrastructure that thrift banks need to monitor growing SME portfolios continuously without proportional staff increases. Automated financial condition ratings. Portfolio-wide deterioration alerts. BSP examination-ready documentation.

If your thrift bank is navigating the growth-versus-monitoring tradeoff, we can walk through how other institutions in the sector are solving it.

Book a call: https://calendly.com/lia_creditbpo/client-discovery-call