Shadow Debt
The Character of Arbitration Awards and Why a Discussion on Arbitral Debt Disclosure is Needed [1]
Introduction
There is a blind spot in the world of sovereign finance, hidden from many investors and their balance sheets: arbitral debt. Over the past decade, a growing number of countries have been hit with multibillion-dollar arbitration awards — judgments handed down after disputes with foreign investors — yet many of these same countries continue to treat the debt as if it does not fully exist. It is not disclosed in bond prospectuses. It is not factored into debt sustainability analyses. It often is not reflected in sovereign credit ratings.
The gap between reality and reporting about this hidden debt is the subject of this piece. Below, we identify the core themes and develop a practical framework (e.g., a traffic light system) for disclosure, relevant whether you're a policymaker, a bondholder, a rating analyst, or simply someone who follows sovereign risk.
This piece draws on a 2020 analysis of sovereign arbitration debt and has been updated to reflect how several of the cases it discussed have since played out — including a major real-world test of its central warning: Venezuela.
Theme One: Arbitration Debt is Bigger than it Looks
Arbitration debts are, in some cases, strikingly large. The Republic of Congo's Commisimpex award, at $1.3 billion and climbing, represents close to a tenth of that country's entire GDP. Russia's Yukos liability — once the largest arbitration award in history at roughly $50 billion — has now been fully confirmed through the courts, with interest pushing its value well past that figure. Both examples suggest the existence of a much larger pool of exposure in both countries’ debt profiles than currently reported.
Part of the problem is structural: arbitration proceedings are confidential by nature, and the major arbitration-specific databases that track investor-state disputes cannot capture everything. Disclosure depends to a large extent on the willingness of the disputing parties to make the existence of their dispute public. Some information about treaty-based investor-state disputes may be disclosed, although not always in a way that permits meaningful assessment of sovereign liability (either actual or potential). By contrast, information about contract-based investor-state disputes is less readily available, even though it too can give rise to material sovereign indebtedness. The Commisimpex award against Congo referenced above, and the P&ID award against Nigeria discussed in our 2020 piece, are both examples of this: each arose from an ordinary commercial contract rather than an investment treaty, and neither proceeded under ICSID rules. At the time of our 2020 analysis, the P&ID award, which had climbed to roughly $11 billion with interest, was still an unresolved threat to Nigeria's balance sheet. It has since played out: an English court set the award aside in 2023 after finding it had been procured through fraud and bribery. The episode cuts both ways for the disclosure argument. It shows how large a contract-based liability can grow, largely unnoticed, before it is finally resolved — but also that "genuinely contested" is sometimes exactly right, which is why our recommendations below tie the amber classification to an active, court-supervised challenge rather than to a sovereign's own assertion of fraud.
Arbitration awards may also have a compounding effect, depending on the circumstances. A single triggering event — a change in currency policy, an expropriation, a regulatory shift — can spawn dozens of nearly identical "doppelganger" claims from different investors, each one adding to a cumulative sovereign exposure. This pattern has already played out in Argentina, Venezuela and Spain, with cumulative exposures running into the billions in several of these situations.
Theme two: Arbitration Debt Behaves Differently than Bond Debt
It would be one thing if arbitration debt behaved like ordinary sovereign debt. It does not, for the following reasons:
It benefits from global enforcement. The New York and Washington (ICSID) conventions ensure that arbitration awards can be enforced globally in any of the signatory countries to those treaties. For context, 165 countries have signed the ICSID Convention, of which 158 have ratified it and hold full Contracting State status. Compare that to a defaulted bond, where a creditor typically has to win recognition of a judgment in a specific court before even beginning the slow, expensive process of chasing down assets abroad. Arbitration creditors who wish to engage in enforcement litigation have a faster, more global toolkit.
The grounds for contesting an award are narrow — but it can take time. States can pursue a set-aside or annulment action, and the grounds are limited and largely jurisdictional. The outcome of these proceedings is binary: either the award is annulled or it is confirmed. Once confirmed, the debt is real — it is no longer "contingent," nor "contested." From that point forward, governments and agencies should – but often do not – include that as part of the sovereign’s overall indebtedness. This is to be distinguished from subsequent enforcement litigation, which may be necessary to actually collect on the debt. In those situations, the debt itself is no longer contingent. All that is contingent at that point is whether a specific enforcement action, against a specific enforcement target, will succeed. Put another way, if – for whatever reason – a specific enforcement action was to fail, that would have no impact on the outstanding debt, which would remain fully payable.
It gets more expensive the longer it's ignored. Many large awards carry interest rates, sometimes compounding. These interest rates are often set – intentionally – at above market rates. The rationale is that setting post-award interest above market rates makes it economically irrational for award debtors to resist payment. Put another way, it should be cheaper for an award debtor to borrow money on the debt markets and use the proceeds to service its arbitration debt. Regrettably, this assumption – that sovereigns will always act in an economically rational way – is not always borne out in practice. There are numerous examples of sovereigns – perhaps poorly advised – adding to their own indebtedness unnecessarily by refusing to pay arbitration awards rendered against them. Some argue, without evidentiary support, that not all sovereigns can access the debt capital markets in this way, or at least not in a way that can deliver cheaper capital to pay arbitration creditors. The evidence points the other way: many sovereigns with clear access to affordable capital still act irrationally, failing to get a grip on their arbitral indebtedness. The Kingdom of Spain is a good example of this phenomenon. Spain’s access to cheap capital on the international debt capital markets is indisputable, yet it persists in resisting the payment of dozens of arbitration awards that carry interest at above market rates.
Unlike bonds, it doesn't get restructured. Sovereign bonds are routinely renegotiated, however messily, as part of debt workouts. Arbitration debt hasn't yet been systematically folded into that process. That leaves arbitration creditors free to pursue enforcement unilaterally, even while a country is trying to negotiate a broader restructuring with its other creditors — and without the collective action mechanisms that help bring bondholders to the table. Accordingly, arbitration creditors, often unencumbered by restructuring arrangements, can secure more favorable recoveries than bondholders.
Theme three: A New Market for Arbitration Debt is Emerging
Arbitration awards are increasingly being treated as tradable assets in their own right. A class of specialized investors now buys claims and awards from successful claimants, often at a discount, giving the original investor an exit after years of costly litigation while the buyer waits out (or actively pursues) collection. This mirrors how the market for distressed sovereign loans developed in the 1990s. In short, arbitration debt is quietly becoming its own distinct asset class, with its own buyers, sellers, and pricing dynamics — another reason it can no longer be treated as a footnote.
Theme Four: The System Currently “Rewards” Non-disclosure
Perhaps the most uncomfortable theme in the underlying analysis is the misalignment of incentives. Governments can benefit from not acknowledging arbitration debt in ways that would be inconceivable for other categories of debt. For example, a government that chooses not to disclose material arbitral indebtedness in a sovereign bond offering may face no sanction from the investment community, whereas failure to disclose a loan of equivalent value might constitute an event of default. There is no justification for this dichotomy. Debt is debt, regardless of the form it takes. If anything, arbitral debt, with its worldwide enforceability, is arguably more impactful on a sovereign’s ability to pay its other creditors, and full disclosure is therefore necessary to ensure the accuracy of its credit profile. The solution is a process that distinguishes genuinely contested awards (i.e. those that are subject to annulment and thus remain contingent) from awards that have survived annulment and are no longer contingent. It is imperative that sovereigns and creditors alike begin to take the latter category into account in their debt sustainability analyses and all the various arrangements that are predicated on the accuracy of such analyses.
Several forces contribute to the status quo:
1. Institutional silos. The attorneys general who litigate arbitration cases often don't communicate well with the finance ministries and debt offices that manage sovereign borrowing, so the debt simply doesn't make it onto the right desk.
2. Political discomfort. Acknowledging a multibillion-dollar liability can be politically costly, creating a strong incentive to keep calling an award "contingent" long after the legal contest as to the validity of the debt – as opposed to enforcement of that debt – is definitively over.
3. Permissive reporting standards. IMF debt-reporting guidelines allow low-income countries to exclude arbitration debt where there's a genuine dispute over validity or amount — a carve-out that becomes an easy loophole once a country simply asserts that the award remains disputed.
4. Rating agencies rely on self-reporting. If a country doesn't disclose an award, the agencies rating its debt generally won't know about it either, which means credit ratings can be built on an incomplete picture.
The consequence is that creditors — bondholders, investors, and other lenders — may be pricing sovereign risk without full information, and countries themselves may be planning their borrowing without a clear-eyed view of their own liabilities.
Our Recommendations
The case for change rests on a simple idea: acknowledging and disclosing arbitration awards does not have to guarantee payment. Sovereigns concerned about conceding an advantage need not do so, but disclosure does prevent nasty surprises for everyone involved. In other words, acknowledging and disclosing arbitration awards helps, rather than hurts, a sovereign's overall credit standing. Here's what's proposed to close the gap.
Build capacity within governments. Finance ministries, debt management offices, and attorneys general need better internal coordination and dedicated systems for tracking arbitration exposure as part of overall debt management — not as a separate, siloed legal matter.
Create a tiered disclosure framework. A "traffic light" system could classify awards by status: red for awards that are final and payable (requiring full disclosure and inclusion in debt sustainability analyses), amber for awards still subject to annulment, and green for claims still pending. Pairing this with a materiality threshold — say, awards exceeding a set percentage of GDP — would keep the framework focused on what actually matters for creditworthiness. The trigger for moving an award from amber to red should differ by enforcement track. ICSID awards should move to red once the Convention's own annulment window closes: because ICSID sits inside the World Bank Group, its public case register gives debt offices and rating agencies a ready-made, low-cost checkpoint, and enforcement in any of the Convention's 158 Contracting States follows automatically once that window shuts. Awards enforced under the New York Convention — the contract-based claims discussed above — warrant a longer amber runway, but only for as long as a court at the seat of arbitration is actively adjudicating a genuine fraud or validity challenge, not merely because a sovereign has asserted one. Once such a challenge is resolved, whichever way it goes, reclassification should follow immediately rather than lapsing back into indefinite ambiguity.
Extend due diligence practices. Rating agencies and country advisors should be systematically checking existing arbitration databases as part of their standard risk assessment, rather than relying solely on what a sovereign chooses to report.
Develop market-based settlement structures in advance. Argentina's settlement with Repsol — where roughly $5 billion in government bonds were issued to satisfy a $10 billion claim, which Repsol could then sell into the market — is held up as a workable template. Countries with significant exposure and reasonable market access could build similar structures — equity swaps, securitization, or settlement-in-kind — and state-contingent instruments like GDP-linked bonds could help cushion the shock of a large award.
Rebalance incentives within the arbitration system itself. Two changes at ICSID could help: introducing cost consequences for investors who bring inflated or weak claims, and requiring arbitrators to have basic training in sovereign debt dynamics so that awards are set with some awareness of a country's actual capacity to pay.
The Test Case: Venezuela
Slaney Advisors’ 2020 analysis floated a hypothetical: perhaps Venezuela would be the country where a sovereign debt restructuring collided head-on with undisclosed arbitration liabilities. That is now playing out. In May 2026, Venezuela's transitional government launched a formal restructuring of its external sovereign debt and the obligations of state oil company PDVSA, the first such process since the country's 2017 default. The numbers involved illustrate almost every disclosure issue discussed throughout this piece.
Estimates of Venezuela's total external liabilities range from $150–170 billion up to a reported $240 billion, depending on the source and what is counted. Arbitration awards and court judgments stemming from Chávez-era expropriations are estimated at more than $20 billion, excluding accrued interest — and researchers tracking the figure caution that even this total may not capture every claim. More than 50 companies have pursued cases against Venezuela and PDVSA, and some debts never went to arbitration or court at all, making them harder to trace through public records. Venezuela has not published comprehensive debt statistics for some time, and many years of sanctions further isolated it from the reporting norms of Western capital markets.
One research group studying the Venezuela restructuring has noted that arbitral awards are routinely left out of sovereign debt inventories, and Venezuela is far from the only country where that happens. In Venezuela’s case, the exclusion of arbitral debt has bite. In the US, for example, a court ruling has permitted arbitration and litigation creditors to pierce the corporate veil and pursue the sale of assets belonging to the state-owned oil company, PDVSA, to satisfy their claims. Simply put, money used to pay those creditors is money that may not be available in the broader restructuring.
For anyone who found the 2020 argument theoretical, Venezuela is its manifestation: a sovereign heading into one of the largest debt restructurings in history without a clear, audited picture of what it owes arbitration creditors, or how those claims will be weighed against bondholders and other lenders. It is a clear demonstration of exactly why the "traffic light" disclosure framework and earlier recognition matter — not as an academic nicety, but as groundwork that, if it had existed years ago, would make today's negotiations considerably more tractable.
Why This Matters Now
None of this is a purely academic concern. As arbitration debt grows in scale and starts trading as a genuine financial asset, the risk is that a sovereign debt restructuring gets blindsided by a liability that nobody accounted for, either properly or at all — or that the arbitration debt itself becomes large enough to trigger a crisis on its own. The solution is not complicated conceptually: treat confirmed arbitration awards as the real, disclosable debt that they are. The harder part is building the institutional habits and reporting frameworks to make that the norm rather than the exception — and, as Venezuela's restructuring is now demonstrating, doing it before the crisis forces the issue is often more effective than doing it during one.
[1] This Slaney Insight was written by Tobias Griffin, Senior Analyst at The Slaney Group.