Why Stronger Buffers Did Not Mean Tighter Credit for Everyone

The Colombian reform produced two different kinds of result, and keeping them separate is the key to interpreting the evidence. It clearly increased provision coverage for the affected maturity tiers, but the study finds no robust average change in interest rates, loan amounts, or loan-to-value ratios. This distinction shows how a prudential rule can strengthen institutions’ loss-absorbing position without broadly changing the terms of newly originated loans—and why the average result still cannot tell us how every lender responded.

The rule moved the buffer first

The direct test of the reform is whether the additional maturity factor changed the provisions attached to covered long-term credit. The study measures this through the provision coverage ratio, which indicates the relationship between provisions and the relevant credit exposure. If coverage rises after the rule for the treated maturity groups relative to comparable shorter loans, the reform has achieved its most immediate resilience objective even if no borrower-facing contract term moves.

That is what the estimates show. Under the strictest exact-matching specification, coverage increased by about 0.166 percentage points at the 72-month threshold and about 0.86 percentage points at the 108-month threshold. The estimates remain positive and statistically significant across the alternative matching specifications. The larger result for the longest tier is consistent with the regulatory design: the additional factor is 40 percent above 108 months, compared with 10 percent for loans above 72 and up to 108 months.

These are treatment effects on the coverage ratio, not a claim that the entire banking system suddenly held 0.166 or 0.86 percentage points more coverage in every portfolio. The comparison is tied to newly originated loans in the relevant maturity groups and to the matched control movement. That specificity is a virtue: it connects the result to the intervention. It also prevents the result from being inflated into a statement about every existing consumer loan or every financial institution in Colombia.

The positive sign is nevertheless economically meaningful. Before the rule, an institution could carry a long-maturity exposure with the expected-loss calculation that applied under the earlier framework. After the rule, the affected new loan requires a larger recognition of potential loss. If the observed coverage ratio rises in the treated group relative to the control group, the accounting buffer is responding in the direction the policy intended. The result is therefore a direct test of the first step in the transmission chain, not merely a correlation between long loans and high coverage.

The numerical difference should not be read as a forecast of future losses. It says that the rule changed the buffer associated with new loans in the study’s matched samples. The larger increase in the longest tier is consistent with a larger mechanical provisioning increment, but it does not prove that those loans will default more often or that the institutions will ultimately suffer larger losses. Provisioning is preparation for a possible event, not evidence that the event has already occurred.