Ring-fence Bangladesh's NBFI sector before a single failure cascades into a Reliance or IL&FS-style collapse

Diagnosis

The curated risk note characterizes the problem as "PLFS / Reliance / IL&FS-style NBFI cascade risk." That framing is precise and it is the heart of the matter. Each of those names is shorthand for the same failure mode: a non-bank financial institution that funds long-dated, illiquid, often related-party assets with short-dated wholesale money, conceals the gap through ever-greening and weak disclosure, and then fails in a way that freezes funding for every other institution that looked similar. The danger is not one weak NBFI. The danger is that the first visible default triggers a run on the whole class, because creditors cannot tell the solvent from the insolvent in time.

Bangladesh sits exposed to exactly this dynamic. NBFIs here borrow from banks and depositors, lend into property and term projects, and are interconnected through common funders and directors. The current_state indicator is null, which is itself the first finding: there is no live, trusted measure of sector fragility to act on. The data status is flagged as needing a collector. A regulator that cannot see cross-institution exposure in near real time cannot pre-empt a cascade, it can only react after the run has started. The Ministry of Finance (MoF) is the named lead body, supported by Bangladesh Bank, the Bangladesh Securities and Exchange Commission (BSEC), the General Economics Division (GED), and the Internal Resources Division (IRD).

Recommended actions

1. Stand up a live NBFI exposure monitor (close the visibility gap first). Owner: Bangladesh Bank under MoF direction, via a supervisory data circular mandating standardized monthly reporting of interbank/inter-NBFI exposures, large related-party loans, and asset-liability maturity gaps. Observable signal: a populated fragility dashboard replacing the current null state, with named institutions ranked by liquidity gap.
2. Triage and resolve the clearly insolvent NBFIs, do not keep them on life support. Owner: MoF with Bangladesh Bank, using a structured resolution process (purchase-and-assumption or orderly wind-down) backed by a ring-fenced resolution fund. Observable signal: at least the worst-classified institutions placed into resolution rather than rolling regulatory forbearance.
3. Ring-fence contagion channels. Owner: Bangladesh Bank, via prudential limits on bank lending to NBFIs and caps on related-party and single-borrower exposure inside NBFIs. Observable signal: declining concentration of NBFI funding from any single bank source, reported monthly to MoF.
4. Tighten disclosure and audit on listed NBFIs and their securities. Owner: BSEC, through enhanced continuous-disclosure rules and forensic review of provisioning. Observable signal: restated provisions and clean independent audits filed for listed NBFIs.
5. Pre-position a contingency liquidity backstop with hard conditions. Owner: MoF with IRD on fiscal capacity and GED on macro framing, via a standby emergency liquidity facility available only to solvent institutions against good collateral. Observable signal: a published facility with eligibility rules in place before, not during, a crisis.

Sequencing (first 12 months)

Start with the monitor (action 1), because without exposure data every other action is guesswork and resolution decisions are contestable. The monitor unlocks credible triage (action 2): once Bangladesh Bank can rank institutions by liquidity gap, MoF can defend resolving the worst few without being accused of arbitrary targeting. Triage in turn makes ring-fencing (action 3) safe, because limits bite hardest when the weakest borrowers are already being removed. Disclosure reform (action 4) runs in parallel from month one. The liquidity backstop (action 5) is built last, deliberately, so it rescues the solvent rather than subsidizing the insolvent already identified.

Risks and constraints

The binding constraint is political: NBFI resolution exposes connected directors and the banks that funded them, and forbearance is the path of least resistance. The fiscal constraint is real too, since a resolution fund and any backstop draw on scarce public capacity, which is why IRD and GED must size them before commitments are made. The supervisory constraint is capacity: standardized real-time reporting demands staff and systems Bangladesh Bank must resource. The largest hidden risk is signaling: heavy-handed action can itself spark the run it aims to prevent, so resolution must move quietly and the solvent-only backstop must be visible before any name is touched.

Bottom line

The PLFS / Reliance / IL&FS pattern is a known, repeatable failure mode, and Bangladesh's NBFI sector currently has no live fragility reading to act on, which is the single most fixable weakness. MoF should make Bangladesh Bank build the exposure monitor first, then resolve the clearly insolvent and ring-fence the rest before a single failure becomes a sector-wide run.