The 2018 Punjab National Bank (PNB) fraud remains the clearest demonstration of how India’s public-sector banking system can be systematically drained by colluding insiders and well-connected borrowers, with the true cost multiplying for years while recovery crawls. Between 2011 and 2017, rogue employees at PNB’s Mumbai Brady House branch issued fraudulent Letters of Undertaking (LoUs) that were never recorded in the bank’s core banking system. Because SWIFT messaging was not integrated with the core ledger, overseas branches of other Indian banks honored the guarantees and released trade credit to entities controlled by Nirav Modi and Mehul Choksi. The principal exposure reached roughly $2 billion (about ₹13,500 crore at the then-prevailing exchange rate). What began as a localized operational failure at one branch quickly revealed itself as a multi-year structural vulnerability that allowed billions to leave the formal banking system without appearing on any central risk register. The absence of real-time reconciliation between domestic ledgers and international payment messaging created a perfect blind spot that was exploited with methodical consistency. Senior management at PNB remained unaware of the growing contingent liability for the better part of a decade, illustrating how concentrated authority at the branch level, combined with technological silos, can undermine an entire public-sector institution. The case stands as a textbook example of how institutional design failures, along with individual greed, enable the scale of such heists. Even the initial conversion of the dollar exposure into rupees at the 2018 rate of roughly ₹65 already locked in a domestic figure that would later prove far too optimistic once exchange-rate reality asserted itself.
The fraud was elementary in design and prolonged by institutional blindness. A single branch could issue multi-hundred-crore guarantees without collateral or central visibility for seven years. When a new clerk finally demanded the missing cash margin after a colluding official retired, the entire structure collapsed. Markets reacted instantly; exposed lenders saw sharp single-day declines. The Reserve Bank of India banned fresh LoUs and ordered real-time integration of messaging systems with core banking. Those reforms closed one specific loophole, but they could not reverse the capital already gone. The simplicity of the method—manual issuance of LoUs outside the core banking solution—exposed a deeper cultural and technological complacency within public-sector banks at the time. Internal audits, concurrent audits, and external inspections all failed to detect the growing volume of off-book liabilities. The eventual discovery was accidental rather than the result of robust surveillance systems. Even after the RBI’s emergency measures, the episode left a lasting imprint on market confidence and forced a wholesale re-examination of how trade-finance instruments were monitored across the banking system. The reforms that followed were necessary, yet they arrived only after the damage had already been inflicted on the balance sheets of multiple lenders.
What no reform could undo was the subsequent interaction between the unrecovered dollar principal and the steady depreciation of the rupee against the dollar over the following eight-plus years.
The perpetrators left India before enforcement could act effectively. Nirav Modi and Mehul Choksi used wealth and networks to secure refuge abroad. Their corporate structures filed for bankruptcy protection in the United States, converting a domestic recovery effort into multi-jurisdictional litigation. In India, the Enforcement Directorate later attached assets valued at approximately $1.78 billion—around 89 percent of the original principal on paper. Yet actual cash returned to the banking consortium by late 2026 stood at only ₹3,617.9 crore (roughly $431 million), or about 21.5 percent of the principal. The rest remains trapped in illiquid real estate, diamonds, wind assets, and foreign accounts subject to endless appeals and Mutual Legal Assistance Treaty processes. The speed with which the principal accused were able to exit the country highlighted serious gaps in real-time monitoring of high-value economic offenders. Once abroad, the combination of dual citizenship arrangements, complex corporate layering, and aggressive legal challenges in foreign courts turned what should have been a straightforward recovery process into a protracted international contest. Domestic attachment orders, while impressive in nominal terms, have repeatedly encountered practical barriers when the underlying assets are either depreciating, difficult to value, or located in jurisdictions with their own procedural safeguards. The result is a persistent gap between the paper value of seized assets and the liquid capital that actually returns to the banks that suffered the original loss. That gap has been widened further by the fact that the same dollars, when eventually recovered, buy significantly fewer rupees of real economic value than they would have in 2018.
The arithmetic of delay is the real indictment. Unrecovered principal of roughly $1.57 billion has, under conservative commercial lending rates of 8 percent, generated forgone interest of about $1.31 billion on a simple-interest basis. Under compounding default terms typical of banking practice, the interest penalty rises to approximately $1.89 billion. Total economic loss to the banking system therefore exceeds $3.46 billion (over ₹32,980 crore at 2026 exchange rates).
Even if every attached asset were liquidated at full book value tomorrow, the proceeds would still fall short of the interest that has already accrued. The original crime has been overtaken by the cost of time. This compounding effect is not an abstract accounting exercise; it represents real lending capacity that public-sector banks could have deployed into productive commercial credit over nearly nine years. Every year that the principal remains unrecovered, the opportunity cost continues to grow, reducing the banks’ ability to support genuine economic activity. The mathematical reality is stark: the interest that has already accrued now exceeds the residual unrecovered principal, meaning the system is effectively paying a permanent premium for the failure to achieve timely recovery. This dynamic transforms a one-time fraud into a continuous fiscal drain that compounds with each passing quarter. When these dollar figures are converted at the October 2026 market rate of approximately ₹95.34 per dollar, the domestic-currency magnitude of both the unrecovered principal and the interest penalty expands dramatically, turning what was already a severe loss into a far heavier burden on rupee balance sheets.
Currency depreciation amplified the damage. At ₹65 per dollar in 2018 the loss was ₹13,500 crore; at ₹95-plus in 2026 the same unrecovered dollars require nearly ₹15,000 crore simply to restore the principal, before interest. Indian depositors and taxpayers effectively pay a rising premium on a foreign-currency obligation that the original borrowers never intended to repay. Public-sector banks, whose capital ultimately rests on sovereign support and household deposits, absorb the hit. The rupee’s structural weakening over the intervening years has acted as a silent multiplier of the original loss. What was recorded as a $2 billion exposure in 2018 has, in domestic-currency terms, expanded simply because the unit of account has lost value. This currency-drag effect is particularly punishing for public-sector institutions that must ultimately report and provision in rupees. The combination of unrecovered foreign-currency principal and a depreciating domestic currency creates a double burden that falls squarely on the Indian fiscal and banking system.
Depositors who place their savings in these banks, and taxpayers who periodically recapitalize them, end up financing both the original theft and the subsequent exchange-rate loss. By October 2026 the exchange rate had reached ₹95.34 per dollar. Applying this rate to the unrecovered principal of $1.569 billion produces a domestic liability of approximately ₹14,960.9 crore—already higher than the entire original 2018 principal of ₹13,500 crore—before a single rupee of interest is added. When the compounded interest penalty is included, the total local liability climbs to roughly ₹32,981 crore. A further modest weakening to ₹97 per dollar would push the same obligation above ₹33,550 crore. In practical terms, the pure currency effect alone has added several thousand crore to the cost of an unrecovered dollar obligation that never should have existed. This is equivalent to the public system paying a substantial additional premium simply for the passage of time and the trajectory of the exchange rate, a premium that benefits neither depositors nor the productive economy.
This pattern is not unique to PNB. RBI data covering two decades show bank frauds totaling more than ₹5.66 lakh crore, with recoveries a small fraction of that sum. High-profile cases—Vijay Mallya’s Kingfisher loans, DHFL’s diversion of tens of thousands of crores, Bhushan Power and Steel, Rotomac, and others—follow the same arc: large exposures to politically or socially connected borrowers, weak internal controls, delayed detection, flight or prolonged litigation, and recovery rates that leave most of the loss on public books.
Employee-enabled frauds continue to surface, from branch-level siphoning of fixed deposits and PPF accounts to more recent systems breaches. The cumulative scale of these frauds over twenty years reveals a structural rather than episodic problem. Recovery rates remaining in the low double digits across the entire portfolio of reported frauds indicate that the enforcement and resolution machinery has consistently lagged behind the speed and sophistication of diversion. Each successive high-profile case has produced temporary public outrage and incremental regulatory tightening, yet the underlying pattern of concentrated large exposures, insider facilitation, and protracted recovery has proven remarkably durable. The persistence of employee-driven frauds, even after major technological upgrades, further underscores that human and governance factors remain central vulnerabilities. When many of these older frauds also involved foreign-currency or trade-finance elements, the same exchange-rate dynamics observed in the PNB case have quietly inflated their domestic-currency cost over time, compounding the original shortfall.
The institutional environment makes such outcomes predictable. Public-sector banks operate under dual pressures of commercial lending and informal political expectations. Maker-checker controls, audit trails, and SWIFT integration were incomplete for years. Once fraud is detected, the combination of slow domestic courts, multi-tier appeals, and the practical difficulty of liquidating seized luxury assets or foreign holdings turns recovery into a multi-year war of attrition. The Fugitive Economic Offenders Act and asset-attachment powers exist on paper; converting attachments into cash that restores bank capital remains the exception. The dual mandate under which public-sector banks function—expected both to operate commercially and to respond to broader developmental or political signals—creates inherent tensions in credit appraisal and monitoring. When large corporate exposures sour, the same institutions that extended the credit must then navigate a judicial system that prioritizes procedural thoroughness over commercial speed. Illiquid assets such as high-end residential properties, specialized industrial equipment, or overseas holdings require prolonged valuation, marketing, and legal clearance before they can be converted into usable capital. The result is that even successful attachment often yields only partial and delayed recovery, while the banks continue to carry the provisioning burden. During those years of delay the exchange rate has moved decisively against the rupee, so that any eventual recovery of dollar-denominated value translates into fewer rupees of capital restoration than would have been the case had the same recovery occurred in 2018 or 2019.
Meanwhile the proceeds of the original fraud have not sat idle. Funds diverted into diamonds, real estate, overseas accounts, and other assets continued to generate returns for the beneficiaries long after the banks booked the loss. The interest the banks should have earned on legitimate commercial loans is instead earned, in one form or another, by those who extracted the capital.
The asymmetry is stark: the public balance sheet carries compounding liability; private actors retain the upside of the diverted capital. Once capital leaves the formal banking system through fraudulent channels, it does not disappear; it is typically reinvested into assets that themselves produce income or capital appreciation. Luxury real estate, high-value gemstones, and diversified overseas holdings can continue to generate returns for the original beneficiaries or their associates even while domestic banks are still provisioning against the original default. This creates a perverse transfer: the opportunity cost of the lost capital is borne by the public banking system and, ultimately, by depositors and taxpayers, while the economic benefits of that same capital continue to accrue privately. The longer the recovery process takes, the greater this asymmetry becomes. Because many of the diverted assets were held in hard-currency or hard-currency-linked forms, their holders have been partially insulated from, or even benefited by, the very rupee depreciation that has inflated the domestic cost of the unpaid obligation for Indian banks.
The social contrast is unavoidable. Hundreds of millions of Indians rely on subsidized food grain and other welfare transfers while the same fiscal system repeatedly recapitalizes banks that have been looted through insider collusion and connected lending. Depositors’ savings and taxpayer capital effectively underwrite the risk of large, poorly monitored corporate exposures. When recovery fails, the cost is socialized; when the fraudsters retain residual wealth, the gain remains privatized. The political and moral dimension of these losses cannot be separated from their financial arithmetic. At a time when large sections of the population depend on targeted public distribution and income-support schemes, the repeated need to inject capital into public-sector banks to cover fraud-related losses represents a direct diversion of scarce fiscal resources. Household savings that flow into the banking system are meant to intermediate productive investment; when a portion of those savings is instead lost to fraud and only partially recovered years later, the social contract that underpins financial intermediation is strained. The visible contrast between the lifestyle of those who benefit from large-scale diversions and the constrained circumstances of welfare-dependent citizens sharpens public perception of systemic unfairness. The additional thousands of crore imposed by rupee depreciation between 2018 and 2026 represent resources that could otherwise have strengthened the very welfare architecture on which hundreds of millions depend, or reduced the need for repeated bank recapitalizations funded by the same taxpayers.
Legal process itself becomes part of the incentive structure. Endless appeals, valuation disputes, and international cooperation delays reward those with the resources to litigate indefinitely. A system that cannot convert attached assets into liquid recovery within a reasonable time effectively signals that the expected cost of detection, after years of delay and currency effects, is lower than the private benefit of the fraud. That is the definition of a structural vulnerability. When the expected timeline for final resolution stretches into the better part of a decade, rational actors factor that delay into their risk-reward calculations. Wealthy defendants can afford prolonged legal contests; the banks and the public treasury cannot indefinitely absorb the carrying cost of unrecovered assets. Multi-jurisdictional proceedings further compound the problem, as each foreign court applies its own standards of evidence, due process, and asset-protection rules. The cumulative effect is that the legal system, intended to enforce accountability, can inadvertently function as a mechanism that protects residual value for the original beneficiaries while the public loss continues to compound. The parallel movement of the exchange rate during these prolonged proceedings has meant that every additional year of litigation has not only increased the interest burden but also raised the rupee cost of the underlying dollar principal, further tilting the economic outcome against the public system.
Reforms after 2018—SWIFT integration, the Fugitive Economic Offenders Act, tighter reporting of large exposures—closed specific gaps. They have not eliminated the deeper problems of governance culture inside public-sector banks, the political economy of large corporate lending, or the glacial pace of asset realization. Newer frauds of smaller individual size but high aggregate volume, including insider theft of customer deposits and digital-channel abuses, continue to appear. Post-PNB measures addressed the particular technical vulnerability that had been exploited, and the legislative framework for dealing with fugitive economic offenders was strengthened. Yet the broader ecosystem of incentives, accountability, and enforcement speed has proven more resistant to change. Large corporate credit decisions still occur within an environment where political and social connections can influence both sanctioning and subsequent forbearance. Simultaneously, a parallel stream of smaller but cumulatively significant insider frauds—targeting fixed deposits, dormant accounts, and digital channels—demonstrates that control weaknesses persist at multiple levels of the banking hierarchy. Incremental regulatory tightening has reduced certain categories of risk, but it has not yet produced a decisive shift in either recovery rates or the frequency of high-impact events. None of the post-2018 reforms altered the fundamental exposure of unrecovered foreign-currency obligations to subsequent exchange-rate movements, leaving the system still vulnerable to the quiet inflation of losses through currency depreciation.
The PNB case therefore functions as a stress test that the system has only partially passed. Detection eventually occurred. Assets were frozen in large nominal amounts. Laws were amended. Yet nearly nine years later the banking consortium still holds a fraction of the original principal in cash, while interest and currency effects have more than doubled the economic loss. The lesson is not that individual criminals were unusually clever; it is that the combination of weak real-time controls, slow enforcement, and the time value of money turns a finite theft into an open-ended fiscal drain. The episode exposed both the capacity of the state to respond after the fact and the enduring limitations of that response. Symbolic victories in the form of large attachment figures have not translated into proportionate restoration of bank capital. The gap between nominal success and economic recovery remains wide. This partial passage of the stress test leaves the system still vulnerable to the same combination of factors—concentrated insider access, delayed detection, capital flight, and protracted realization—that produced the original loss. The 2026 exchange rate of roughly ₹95.34 per dollar, compared with ₹65 in 2018, provides a concrete measure of how severely the passage of time has inflated the domestic cost: the unrecovered principal alone now demands more rupees than the entire original fraud was booked at, and the addition of compounded interest pushes the total well beyond double the initial figure.
Until recovery mechanisms move at something closer to commercial speed, and until internal controls and accountability for large exposures become genuinely deterrent, the pattern will recur. Public banks will continue to intermediate household savings into large private risks. When those risks crystallize through fraud or willful default, the residual loss will again fall on the same depositors and taxpayers who already fund welfare transfers for the poorest. The sovereign bleed is not an accident of one branch in 2018; it is the predictable result of a system that still allows time, currency, and legal friction to compound private gains at public expense. Meaningful progress requires more than technical fixes to messaging systems or the addition of new statutes. It demands a sustained shift in the speed of asset realization, the culture of accountability within public-sector banks, and the political willingness to treat large-scale banking fraud as a continuous fiscal emergency rather than a series of isolated scandals. Until that shift occurs, the arithmetic of delay will continue to favor those who extract capital over those who must ultimately bear the cost of its loss.
The comparison between the 2018 and 2026 exchange rates makes the stakes explicit: every year of unresolved recovery does not merely preserve the original hole in the balance sheet; it enlarges that hole in domestic-currency terms through the quiet but relentless mechanism of currency depreciation, ensuring that the public ultimately pays more—often substantially more—than the criminals originally took.