The discount rate Lebanon’s deposit law never states

Seven years after Lebanon’s banking collapse, the size of the loss remains a range rather than a figure. BLOMINVEST’s reading of the Alvarez & Marsal preliminary audit, net of central bank offsets, produces $3.3 billion. The IMF assessment reported by the Financial Times in 2020 put it at $49 billion. A January 2026 assessment by the Tahrir Institute for Middle East Policy (TIMEP) reaches $60 billion. The government’s own 2022 estimate was $70 billion. The draft Financial Gap Law approved by cabinet in December 2025 works from $80 to $83 billion.

Figure 1. The loss, as estimated by institutions looking at the same country

Sources: Alvarez & Marsal preliminary BdL audit as read by BLOMINVEST (net-of-offsets interpretation); IMF assessment reported by the Financial Times, 2020; TIMEP, January 2026; Lebanese government estimate, 2022; draft Financial Gap Law, December 2025.

The estimates measure partly different things, which is itself the point. No authority has published a reconciliation between them. A loss that cannot be sized cannot be allocated, and a loss that cannot be allocated leaves no creditor senior to any other. The IMF has asked, so far without result, for the hierarchy of claims to be made explicit.

That gap is usually described as a political failure. It is more precisely a valuation failure, and it is soluble. The claim has been priced repeatedly, in venues that share no common mechanism, and those prices imply a discount rate that the draft law itself declines to state.

What the market has already decided the rate is

After October 2019 banks imposed capital controls that were never legislated. Withdrawals were rationed by circular. A single dollar-denominated line on a customer’s statement resolved into several distinct claims carrying materially different values, every one still displayed with the same currency symbol.

Figure 2. One deposit, seven simultaneous prices, in cents on the dollar

Sources: BdL Circulars 151 and 158; secondary “lollar” pricing as widely reported 2021–23; Tyre settlement reported by AFP and the Depositors’ Association, 4 October 2022; Eurobond pricing per Bank Audi research via L’Orient Today, July 2026.

The conventional treatment of the draft law assumes a discount rate and solves for present value, which invites the objection that the rate is arbitrary. The exercise runs better in reverse. The instrument offered to balances above $100,000 is described as a twenty-year obligation carrying a low coupon.

Taking a 2 percent coupon, its present value is P = c·a(r,20) + (1+r)⁻²⁰, where a(r,20) is the twenty-year annuity factor. Setting P equal to each observed price and solving for r gives the rate that venue has already applied to Lebanese central bank credit.

Table 1. The discount rate each venue has already applied

Author’s calculation. Rate solving P = c·a(r,20) + (1+r)⁻²⁰ at c = 2 percent for each observed price. Prices as sourced in Figure 2.

Three unrelated venues, a distressed sovereign bond market, a branch-level settlement reached under duress and a central bank’s own published liquidity, converge on a discount rate between roughly 12 and 18 percent. That convergence, rather than any single price, is the finding.

At 15 percent, the midpoint, the twenty-year instrument is worth 18.6 cents on the dollar. At the Eurobond-implied 12 percent it is worth 25.3 cents. At the rate implied by the central bank’s own coverage, 14.34 percent of foreign currency liabilities to banks, it is worth 19.5 cents.

Two tiers, one obligor, two prices

The draft law protects balances up to $100,000 across roughly 782,000 accounts, repaid over four years, and converts balances above that threshold into long-dated central bank instruments. Both tiers are claims on the same obligor. Neither is formally senior to the other. The economic distance between them is nonetheless the largest single number in the settlement.

Discounted at 15 percent, four equal year-end instalments are worth 71.4 cents. The twenty-year instrument is worth 18.6 cents. The spread is 52.7 cents, and it barely moves with the discount rate: 47.4 cents at 10 percent, 52.8 cents at 18.6 percent. The differential is produced by duration and coupon, and it is therefore invisible in any document that describes both tiers as repayment.

Table 2. Present value of the above-$100,000 instrument at a 15 percent discount rate, in cents on the dollar

Author’s calculation. Rate solving P = c·a(r,20) + (1+r)⁻²⁰ at c = 2 percent for each observed price. Prices as sourced in Figure 2.

The scale can be stated another way. For the protected tier to be worth what the market currently pays for the unprotected one, it would have to be discounted at roughly 93 percent. To close the gap from the other direction, the twenty-year instrument would need a coupon of about 10.4 percent rather than 2 percent. The draft therefore embeds a subordination of roughly eight percentage points of annual yield through structure alone, without ever ranking one depositor behind another.

None of this makes the law wrong. Restructurings are designed to impose losses, and long-dated instruments are a legitimate tool and the issue at the end of the day is disclosure. A statute describing a nineteen-cent instrument as full repayment has omitted a discount rate, and the omission is worth roughly eighty cents on every dollar above the threshold.

The repayment profile for each tier should be published in full. The present value of each instrument should be stated alongside its face value at a disclosed rate. Where the state’s own costings and independent readings diverge, the assumptions should be published so competing estimates can be tested against identical terms. Analysis reported by EnterpriseAM puts first-year funding for the protected tier near $9.5 billion against $11.53 billion of liquid reserves, while Ishac Diwan of the Carnegie Endowment reads the same balance sheets and concludes assets of that magnitude are broadly available. Both are serious readings, and the divergence has not been reconciled.

Why nothing flagged it

Three arrangements allowed a concentration of this size to accumulate without registering anywhere.

Under the Basel framework as applied in Lebanon, exposures to the sovereign and to the central bank in local currency carried a zero risk weight. By 2019, per TIMEP, commercial banks had placed roughly $90 billion with Banque du Liban, around 75 percent of depositors’ foreign currency, while lending about a quarter of balance sheets to the private sector. Every dollar moved from a risk-weighted private loan to a zero-weighted central bank placement improved the reported capital ratio while worsening the actual position.

Second, the central bank published no profit and loss account and did not apply international financial reporting standards. Losses in economic substance were carried on the asset side under headings including seigniorage and deferred open-market operations. Its most recent published balance sheet still carries a valuation adjustment account and a deferred open-market operations line running into the hundreds of trillions of lira, classified as assets. An institution with no published P&L cannot report a loss, because there is no statement on which one would appear.

Third, timing. Banks were required to retain the profits generated by the 2016 operations as lira reserves ahead of the 2018 adoption of IFRS 9, which replaced incurred-loss with expected-loss provisioning. Fixing the treatment of those gains before the new standard applied reads as procedural and functioned as decisive.

The same preference for administrative instruments over legislative ones governed the restrictions themselves. Circulars 151, 158 and 166 rationed withdrawals at successive conversion rates. No capital control law was ever passed. A formal control would have required a parliamentary vote and created a legal event with a date attached, against which claims could be measured.

Where the claim is still enforceable

Two legal clocks are running, and both operate outside Lebanese jurisdiction.

Local banks sold more than $6 billion of Lebanese sovereign Eurobonds into foreign hands, much of it after the crisis began and some after the March 2020 default. Foreign funds are reported to hold north of $17 billion and to control blocking positions in around 40 percent of outstanding series. Because collective action clauses vote series by series at a 75 percent threshold, a blocking stake in 40 percent of series functions as a veto over any restructuring. The country’s negotiating leverage was sold, at distressed prices, to fund the exits.

Meanwhile the statute of limitations on the Eurobonds runs ten years on principal and five on interest from the 9 March 2020 default. Absent a restructuring, principal claims prescribe in March 2030 while judgment creditors accumulate outside the collective action framework. And enforcement has already succeeded once in a foreign forum, specifically after a London court ruled for depositor Vatche Manoukian, lawyers reported banks pre-emptively closing accounts, affecting more than fifty British savers. The claim proved enforceable precisely where the jurisdiction was not Lebanese.

The architecture is still standing

Figure 3. The architecture has not been dismantled

Sources: TIMEP, January 2026, for the 2019 placement share; Lebanon’s consolidated commercial bank balance sheet for May 2026 via BLOMINVEST.

Lebanon’s consolidated commercial bank balance sheet for May 2026, as reported by BLOMINVEST, shows total banking-sector assets of $100.6 billion, of which $75.8 billion sat with the central bank and $4.3 billion represented claims on resident customers. Measured against total assets rather than foreign currency deposits, the concentration is tighter than it was in 2019.

Since August 2023 the central bank has stopped financing the government, lifted the reserve-consuming subsidies and sharply limited money creation. Foreign reserve assets rose roughly $3.3 billion over that period and fell only $420 million across the first seven months of 2026, a period containing a regional war and a closed Strait of Hormuz. Banking secrecy was narrowed in April 2025 with ten-year retroactive access, and a bank resolution law passed that July.

A funnel that is no longer fed cannot inflate. It can only drain, and it is draining: financial sector deposits at the central bank fell 4.8 percent year on year, resident foreign currency deposits 4.4 percent. Between three and four billion dollars of claims disappears annually through attrition, write-off and settlement rather than repayment, which is the same slow default continuing by other means.

A repeat of 2019 is not available. There is no peg left to break, no unfrozen deposit left to freeze and no Eurobond left to default on. What remains scheduled is narrower and dated.

The FATF plenary concluding Lebanon’s two-year action plan at end-2026, where a downgrade would degrade the correspondent banking relationships through which $6.5 to $7 billion of annual remittances arrive, and the parliamentary vote that conditions any IMF programme and the Eurobond restructuring behind it.

Five of the six figures worth monitoring are diagnostic of what has already happened. Only one measures whether banks have resumed the function banks exist to perform, and lending to resident customers sits at 4.3 percent of assets. The question facing Lebanon is at what rate the next dollar in Lebanon will be discounted, and whether anyone will be told.