The Judiciary Doesn't Rate-Set, But It's Topping QGI's Currency-Crisis Leaderboard

The standard macro anxiety report on the UK runs as follows: a structural current account deficit persisting for three decades; gilt yields at their highest in a generation; government debt tracking toward 96% of GDP; a pound that rose against the dollar while falling against the euro; debt service costs tripled in five years. That picture is real, and genuinely uncomfortable. It is also, as it turns out, the secondary story.

QGI's currency-crisis model ranks the United Kingdom first of 157 countries in 2026. The largest single driver of that ranking is not the gilt yield, not the current account deficit: it is judicial reform attempts. A courts-versus-executive confrontation that bond desks have treated as a constitutional sideshow is topping a model trained on historical currency crises across 157 countries and multiple decades.

The finding becomes more striking when the UK's nearest historical analogue is identified: the United Kingdom itself, in 1992, in the twelve quarters preceding Black Wednesday, with a match score of 0.975(effectively a perfect structural rhyme). The country's nearest structural comparator is itself before its most catastrophic post-war currency event.

A second independent data point reinforces the judicial thesis. The model's second-closest analogue is Italy 1992, at a match score of 0.9416: the same European monetary crisis, three days earlier, when the lira was suspended from the ERM before sterling followed. Italy 1992 carries secondary recipe tags in QGI's model for political_corruption and terrorist_attack, consistent with the Tangentopoli anti-corruption prosecutions simultaneously dismantling Italian political institutions, and the Mafia's assassination of judges Giovanni Falcone and Paolo Borsellino in May and July 1992 respectively. A currency crisis coinciding with the systematic murder of the country's most prominent judicial independence figures. Two independent historical data points (one from the UK's own past, one from a structurally similar contemporaneous case) both pairing currency crisis with acute institutional breakdown at the courts-executive boundary.

1. The Model and Its Honest Limitations

Before the analysis, the disclosure. QGI's currency-crisis model carries a cross-validated AUC of 0.7346 (moderate discriminative power).¹Cross-validated Area Under the ROC Curve for the currency-crisis recipe. In a randomly selected pair of country-years (one pre-crisis, one not), the model correctly identifies the pre-crisis case approximately 73% of the time. Materially better than chance (0.5) but not a precision instrument.Read more → On a scale where 0.5 is a coin flip and 1.0 is perfection, this model sits closer to the useful end of the spectrum but is not a precision instrument. What it produces is structural resemblance to historical precedents, not a forecast. A score of 100 means the UK's current indicator configuration is more similar to the pre-crisis configurations in QGI's historical evidence base than any other country on earth. It does not mean a crisis is imminent, inevitable, or even probable in a calendar sense. It means the structural fingerprint matches. Investors and analysts should treat this as a pattern-recognition flag, not a trade signal.

The model was scored at 2026-06-01T22:58:47 UTC, vintage te-fleet-run-2, across 157 countries. The UK is in the evidence base (meaning it appears in the training data as a historical case) and ranks first among all scored countries.

2. What Is Actually Driving the Score

QGI's framework decomposes each country score into its contributing indicators using SHAP values²SHapley Additive exPlanations: a technique derived from game theory that decomposes a model's output for a specific observation into the contribution of each input feature. A positive SHAP value means the feature is pushing the score upward toward crisis resemblance; a negative value would push it downward.Read more →, a technique that assigns each input variable a share of the model's output for that observation. For the UK in 2026, the driver ranking is:

Three things are worth holding in mind as you read this table.

First, all five SHAP values are positive: every driver is pushing the score upward. There are no offsetting signals in this model. The UK is not a country where economic strengths are partially cancelling institutional weaknesses. The configuration is uniformly crisis-resemblant across all five dimensions.

Second, the top three drivers together account for 82.5% of the signal. This is a concentrated, coherent profile, not a diffuse one.

Third, and most consequential for the analysis that follows: the lead signal is institutional, not macroeconomic. A judiciary indicator outweighs the current account deficit by seven percentage points.

3. The Judicial Dimension: What the Model Is Measuring

The V-Dem v2jureform indicator³V-Dem (Varieties of Democracy) is an academic dataset produced by the V-Dem Institute at the University of Gothenburg, coding political and institutional indicators based on expert assessments. The v2jureform indicator captures attempts by the executive to reform the judiciary in ways that reduce its independence. It is expert-coded on a 0-4 ordinal scale then converted via Bayesian IRT to a continuous latent variable; lower values correspond to more aggressive executive attempts to constrain judicial independence. V-Dem coding carries a typical lag of one to two data-vintage years.Read more → is expert-coded on a 0–4 ordinal input scale, then converted via Bayesian item response theory to a continuous latent variable score; the output is not bounded to 0–4 and can take negative values (the UK raw value in this vintage is −0.045). Lower latent scores correspond to more aggressive executive attempts to constrain or reform the judiciary. The UK's trajectory in the data shows a striking pattern: relatively stable scores from 2021 through early 2023, then a sharp deterioration (a z-score drop from +0.55 to −2.98 at the inflection point), with partial stabilisation around −0.40 to −0.61 thereafter. The model's trajectory window registers this as a large, sudden, and incompletely reversed shock to judicial-executive relations.

The events underneath that inflection point are not obscure. In November 2023, the UK Supreme Court ruled unanimously that the government's Rwanda deportation policy was unlawful, finding Rwanda was not a safe third country and that the risk of refoulement was real [Human Rights Watch, Nov 2023; Euronews, Nov 2023]. The Sunak government's response was not to accept the ruling. It legislated around it. The Safety of Rwanda (Asylum and Immigration) Act 2024 directed judges to treat Rwanda as safe “regardless of any evidence to the contrary before them”, stripping courts of their factual review function in that domain. The UN Special Rapporteur and the Council of Europe Commissioner for Human Rights issued formal statements: the Bill constituted an “effective infringement of judicial independence” and amounted to “legislative usurpation of the judicial function” [OHCHR, March 2024; International Bar Association, 2024].

This was not an isolated episode. Running alongside it: a sustained debate about the UK's continued membership of the European Convention on Human Rights, which shifted from fringe to mainstream party politics by 2025–26 [Public Law for Everyone, July 2025; BIICL, 2026]; ongoing judicial review reform proposals narrowing the scope of court oversight of executive action [Law Society; Commons Library]; and a structural question about executive willingness to operate within judicial constraints that the Law Society described as the most acute courts-versus-executive confrontation in post-war British constitutional history.

The QGI model was not coded to find this. It was trained on pre-crisis structural configurations across 157 countries and multiple decades. When it identifies judicial reform attempts as the single largest driver of UK currency-crisis resemblance, it is reporting a pattern in the historical data: the pre-crisis period in currency crises has repeatedly been a period of elevated institutional stress, including stress at the boundary between executive and judicial authority. The mechanism is not opaque, even if the model does not specify it. Three transmission channels are consistent with the historical co-occurrence. First, judicial independence functions as a proxy for rule-of-law credibility: overseas capital markets assign a sovereign risk premium to jurisdictions where executive power over judicial outcomes is perceived to be expanding, not because courts set rates, but because the rule of law is the underlying guarantee of property rights and contract enforcement on which capital allocation depends. Second, executive-judicial conflict is a leading indicator of policy unpredictability: governments willing to legislate around court rulings signal a lower constraint on future executive action, raising the institutional discount that foreign investors apply to long-duration assets denominated in that currency. Third, and most structurally, executive overreach toward judicial institutions has historically been a coincident or leading indicator of the underlying political-economy deterioration that produces crises: governments under electoral and fiscal stress reach for institutional control tools precisely when macroeconomic pressure is building, making judicial attacks a symptom of the same underlying fragility. The model has learned this co-occurrence across 157 countries and multiple decades. The mechanism need not be direct to be real.

The V-Dem lag caveat⁴The trajectory values for v2jureform in the QGI model are z-scores relative to the cross-national distribution. A z-score persisting across two periods may indicate genuine stabilisation or may reflect the data vintage not yet incorporating the most recent assessments. This distinction is the most material empirical uncertainty in this report.Read more → deserves full treatment here, not a footnote. V-Dem scores carry a typical lag of one to two years in their expert-coding process. The trajectory values in the model's window (2024–26) likely reflect predominantly 2022–24 events, specifically the Rwanda Act period under the Sunak government. Whether the Starmer government's materially different posture has already changed the underlying score is unconfirmed in available data. The plateau in the z-score trajectory (−1.0085 for two consecutive periods, then −1.1605) may reflect data vintage lag as much as genuine stabilisation. There are two possible interpretations with meaningfully different implications. If the plateau is genuine (meaning the Starmer government's less confrontational approach has arrested the deterioration), the lead signal should recede in the next two V-Dem coding cycles, and the QGI score should fall from its current rank-1 position. If the plateau is artefactual (meaning the coding lag means the 2025 assessments have not yet been incorporated and the underlying institutional trajectory has not reversed as fully as the current government's posture suggests), the signal may remain elevated or worsen. A country specialist pressing on this question would find it genuinely unresolved in the current data vintage. It is the single most important caveat on the lead finding.

4. The Macroeconomic Layer: The Conventional Story Is Also Real

Stripping out the judicial signal, the remaining 65.5% of the QGI score is macroeconomic, and it maps directly onto the standard anxiety report.

Current account balance (27.2% of signal).The UK's current account deficit ran at £86 billion (3.0% of GDP) in 2024, narrowing to £74 billion (2.4% of GDP) in 2025 [ONS, Q4 2025]. The quarterly pattern is not reassuring: Q2 2025 widened to 3.2% of GDP, Q3 narrowed to 1.4%, Q4 widened again to 2.4%, oscillating rather than converging. The QGI raw value reads −2.98% of GDP, improving directionally from −3.58% but still structurally negative. The trajectory z-score of +1.598 means the improvement is modest relative to historical variation in the evidence base.

The UK has run a structural current account deficit for most of the past three decades. Its external position depends continuously on capital inflows: the “kindness of strangers” formulation that Mark Carney put to Parliament in 2016 retains its accuracy. That dependence creates a structural vulnerability to confidence shifts in overseas capital markets that a surplus country does not face.

Imports (20.8% of signal).The raw value of $1.174 trillion (up from $1.132 trillion in the prior vintage) carries a trajectory z-score change of +2.456, the steepest upward move of any indicator in the UK's entire profile. This is the variable moving fastest, relative to its own historical distribution, in the UK's current configuration, and it is moving in the wrong direction.

Rising imports, in the model's historical pattern, are associated with the pre-crisis period through a specific mechanism: domestic demand running ahead of productive capacity, currency strength masking external fragility, and a trade position that can deteriorate rapidly when the exchange rate moves. The sequence in crisis episodes typically runs as follows. An import surge, sustained by consumer demand and a relatively strong currency, widens the current account deficit in a way that is self-obscuring while the currency holds, with sterling strength in 2025 making the current account look manageable even as the underlying import trend worsened. When the exchange rate moves under external pressure, the import bill does not contract immediately; it typically lags by several quarters, meaning the external position worsens through the initial depreciation before improving. In the 1992 UK analogue, the Lawson Boom import surge (current account reaching −4.9% of GDP by 1989) was precisely this mechanism: demand-led overheating, currency-masked fragility, rapid reversal on ERM entry and subsequent squeeze. The +2.456 z-score on imports today is the strongest upward signal in the UK's current configuration. It warrants equivalent attention to the judicial dimension: it is one-fifth of the model's total output, and the mechanism connecting it to currency crisis risk is well-documented in the historical evidence.

GDP nominal level (14.1%) and GDP deflator (3.5%). These are structural positioning variables. At $3.686 trillion, UK GDP places the country in a size cohort (large, liquid, internationally referenced) that matches the historical currency-crisis evidence base in specific ways. Large, liquid currencies are subject to large, liquid speculative attacks, as 1992 demonstrated. The GDP deflator at 3.62%, declining from a recent peak of 6.34%, shows that the post-pandemic inflation spike is unwinding, but the trajectory z-score of +1.990 means the level relative to history remains elevated.

The gilt market as corroborating signal.The QGI model does not include gilt yields directly, but the bond market's own verdict is consistent with the model's reading. In January 2025, 30-year gilt yields reached 5.27%, their highest level since 1998 [Bloomberg, 7 Jan 2025]. The Lords held an emergency debate on 14 January 2025 specifically titled “Sterling: Rise in Yields on 30-Year Gilts” [Hansard]. By May 2026, yields had returned to those levels [Bloomberg, 5 May 2026]. The National Institute of Economic and Social Research published a blog in 2026 titled “Could a Gilt Market Shock Derail the Economy in 2026?” [NIESR, 2026]. The UK's 10-year gilt currently carries the highest yield in the G7. The government plans to issue £297 billion of bonds in this fiscal year, the second-highest on record [Bank of England Insights, 2026]. These are not the numbers of a country whose external financing position is stable.

5. Black Wednesday: The Self-Rhyme

The nearest historical analogue in QGI's model is the United Kingdom in 1992, with a match score of 0.975, effectively a perfect structural rhyme. Understanding what happened then is inseparable from understanding what the model is saying now.

The Run-Up: 1988–1990

The conditions were assembled across a decade. The Lawson Boom of the late 1980s (UK GDP growth running well above the 2.5% long-run trend, a housing market surge feeding consumer demand) produced a current account deficit of 4.9% of GDP by 1989 and inflation reaching 9.5% by 1990, roughly three times Germany's rate [Economics Help; Economics Observatory]. These are the 1992 analogues of the current account deficit at −2.98% and imports running at +2.46 standard deviations above historical mean in today's data.

Margaret Thatcher had resisted ERM membership for over a decade, viewing a fixed exchange rate as a surrender of monetary sovereignty. On 8 October 1990, with John Major as Chancellor, the UK joined the ERM at a central rate of £1 = DM2.95, with a ±6% fluctuation band [John Major Archive, Oct 1990]. Major stated publicly he was “confident that a central rate of DM2.95 is sustainable.” Critics argued the rate was overvalued (some preferred DM2.50), and that the overvaluation would structurally worsen the current account [Economics Help; OMFIF].

The Squeeze: 1990–1992

Two forces converged to make the peg untenable. The Bundesbank raised German interest rates aggressively to counteract inflationary spending from German reunification. By August 1992, German short-term rates stood at approximately 9.9% (their highest in roughly 60 years), while US rates sat at 3.4%, their lowest in 30 years [Economics Observatory; NBER Working Paper w29488]. The ERM's design required member states to import German monetary tightness regardless of their own cyclical position.

The UK was in recession. High interest rates (maintained to defend the peg) inflated mortgage costs precisely as the housing market collapsed. Default rates rose. Unemployment climbed. The political trap was complete: defend the peg (maintain high rates that were breaking the domestic economy), cut rates (abandon the peg), or devalue (political humiliation). George Soros began building a short position on sterling in late 1991 as the structural contradiction became visible to sophisticated market participants [MoneyWeek; IG UK]. By summer 1992 the position had grown to $10 billion against the pound.

The political-institutional dimension ran in parallel. The Maastricht Treaty ratification became a constitutional crisis: Conservative “Maastricht Rebels” (including figures backed by Thatcher and Norman Tebbit) opposed the European Communities (Amendment) Bill at multiple stages [Wikipedia, Maastricht Rebels; Mile End Institute, QMUL]. Separately, the Factortame litigation had established by 1990 that UK courts could disapply Acts of Parliament conflicting with EU law, a rupture in the Diceyan doctrine of parliamentary sovereignty that ran directly through the ERM/Maastricht era [House of Lords, Factortame, 1999]. In July 1993, William Rees-Mogg brought a judicial review challenge to the Maastricht Treaty's ratification (calling it “the most important constitutional case for 300 years”), which was dismissed but demonstrated that the courts were being drawn into the centre of a political-monetary crisis [Wikipedia, R v Secretary of State ex p Rees-Mogg].

This is the dimension the QGI model is now measuring in 2026 through the v2jureform indicator. The judicial-constitutional tension of 1992–93, parliament versus courts, executive versus legal constraints on sovereignty, co-occurred with the currency crisis. The model learned that this co-occurrence is a structural pattern, not a coincidence.

Black Wednesday: 16 September 1992

“If I had to pick one day in history to symbolise the supremacy of markets over states and democracies, Black Wednesday would be the one... Black Wednesday killed off any idea that markets could be tamed by democratic means.”

The Italian lira had been devalued on 13 September and suspended from the ERM, signalling wider fragility [Wikipedia; IG UK]. On 16 September, the Bank of England was buying £2 billion of sterling per houragainst an unstoppable tide of selling. The base rate was raised from 10% to 12%, then to 15%, within a single trading day. Speculators judged, correctly, that 15% interest rates were politically unsustainable against a backdrop of recession, collapsing house prices, and heavy variable-rate mortgage exposure: the government could not inflict 15% rates on an already-broken domestic economy long enough to outlast a position of $10 billion. Selling continued. By evening, Chancellor Norman Lamont announced Britain's suspension from the ERM. The loss on currency intervention was later confirmed by declassified Treasury papers at £3.3 billion [IG UK; UK Stockbrokers]. Soros made an estimated $1 billion profit [Wikipedia; MoneyWeek].

The political consequence was structural and lasting. Black Wednesday destroyed the Conservative Party's reputation for economic competence, legitimised euroscepticism, and made future UK participation in the euro politically impossible [tutor2u; IG UK; The Conversation/Stenfors]. The institutional consequence was equally consequential: the crisis triggered the adoption of inflation targeting and ultimately the Bank of England's operational independence in 1997 [Economics Observatory; IG UK], one of the most significant institutional changes in UK economic governance in the twentieth century.

The Secondary Analogues

Italy 1992 carries a match score of 0.9416, the second-closest structural comparator [QGI, te-fleet-run-2]. Italy also exited the ERM on 13 September 1992, three days before the UK. The secondary recipe tags (political_corruption and terrorist_attack) point to Tangentopoli and to the assassinations of Falcone and Borsellino. The Italian case is not a simple monetary parallel to 1992: it is a case in which a currency crisis and the systematic destruction of judicial independence through assassination unfolded simultaneously. A cross-national model picking up Italy 1992 as the UK's second-nearest analogue, through the judicial reform indicator, is doing something analytically meaningful: it is confirming the judicial-institutional co-occurrence from a structurally independent case that shares neither the UK's legal tradition nor its monetary history.

Spain 1993 is the third analogue, at a match score of 0.7727; the Spanish peseta underwent multiple devaluations in the ERM aftershock period. All three analogues cluster in the same European monetary crisis of 1992–93.

The fact that the UK's nearest analogue is itself (not a country with different institutions, different currency, different history) is analytically striking. The model is not drawing an analogy across different systems. It is saying the UK's current indicator configuration, across five structural dimensions, is nearly identical to its own configuration in the twelve quarters before its most severe post-war currency event.

6. What Is Different This Time: How the Crisis Would Manifest

The self-rhyme is structural, not mechanical. There are important differences between 1992 and 2026 that the model's resemblance score does not fully capture, and intellectual honesty requires stating them, but they modify the form of vulnerability, not its existence.

No fixed exchange-rate constraint. Sterling floats. The ERM peg was the precise mechanism through which the 1992 crisis manifested: the government was obligated to defend a rate that markets believed was wrong, and that obligation was both finite (bounded reserves) and politically catastrophic to abandon. Today, sterling can move. A currency crisis in a floating-rate regime looks different: it is typically slower, more diffuse, expressed through sustained depreciation pressure and the cost of external financing rather than a single dramatic peg collapse.

The Liz Truss precedent is instructive about what that looks like. In 45 days in autumn 2022, sterling fell to an all-time low of $1.035, 30-year gilt yields rose over 100 basis points, the Bank of England intervened in the gilt market to prevent a pension fund liquidity crisis, and a Prime Minister was removed [Wikipedia; Bank of England]. The mechanism was not an exchange rate peg: it was bond-market repricing combined with sterling depreciation triggered by a loss of fiscal credibility in overseas capital markets. That is the template for how a crisis manifests in a floating-rate regime with the UK's external position. The 2022 episode resolved rapidly because the fiscal shock was reversed; a crisis that combined the fiscal position with the external financing volume of £297 billion in planned issuance, and with an already-elevated yield environment, would be less quickly resolved.

The judicial dimension is new, and its unresolved lag is the most important caveat. The 1992 analogue does not have an equivalent of the Rwanda Act as a direct causal precursor to the currency crisis. The constitutional tensions of 1992–93 were real and contemporaneous, but they were not the lead signal in contemporaneous analysis. The fact that the QGI model now identifies judicial reform attempts as the strongest signal in the UK's 2026 configuration is either evidence that the model is detecting a cross-national structural pattern that contemporaneous analysis missed, or evidence that the UK's current institutional configuration is genuinely outside historical parameters on this dimension. The V-Dem coding lag (discussed in the judicial section above) means the critical empirical question (has the Starmer government already changed the underlying score?) remains unresolved. This is the single most material uncertainty in the analysis.

Scale of external financing.£297 billion in planned gilt issuance (the second-highest on record [Bank of England Insights, 2026]) means the UK's dependence on ongoing overseas demand for its debt is exceptionally high. Gilt yields at 27-year highs, with the 10-year spread the widest in the G7, suggest that market is pricing in elevated risk even before any acute event.

A forward-looking peer worth watching.The three identified analogues are all co-contemporaneous crisis participants, not countries in an early stage of resemblance. A country currently showing an early-stage combination of judicial independence pressure and current account deficit that rhymes with the UK configuration (at a lower QGI score today, but on a trajectory) would complete the analytical picture. QGI's scoring of Poland warrants attention in this context: judicial independence deterioration has been the defining political-economy story of the past decade, the current account position has oscillated, and the country is in the large-external-financing-dependent cohort of EU member states. This is an early-stage watch, not a scored finding, and will be revisited in future vintage updates.

7. The Forward Call

QGI's framework requires a falsifiable forward statement.

The structural configuration identified (judiciary-first lead signal, current account deficit, import elevation at +2.456 standard deviations, nominal GDP size placing the UK in the large-liquid-currency cohort) is consistent with elevated vulnerability to an external financing shock that manifests in gilt yields and sterling rather than a peg collapse. The specific crisis form most consistent with the historical analogue and the Truss 2022 precedent: a loss of fiscal credibility in overseas capital markets, leading to sustained gilt yield elevation, currency depreciation pressure, and a government forced to tighten monetary or fiscal policy at a cyclically inopportune moment.

The falsifiable test has three observable components. First: if the UK's V-Dem judicial reform score stabilises and reverses materially over the next two V-Dem coding cycles, reflecting a structural improvement in judicial-executive relations confirmed by the Starmer government's legislative record, the QGI score should fall substantially from rank 1. Second: if the current account deficit narrows below −1.5% of GDP for two consecutive quarters, the macroeconomic layer recedes with it. Third, the model's flag is validated if, before the next vintage update, the UK advances ECHR departure legislation AND gilt yields sustain above 5% on 30-year paper AND sterling trades below 1.20 against the euro for more than 30 consecutive trading days, the combination of renewed institutional confrontation, elevated financing costs, and currency pressure that constitutes an external-shock-meeting-structural-vulnerability in a floating-rate regime.

Conversely, the combination most likely to produce a score retreat: the Starmer government legislatively restores judicial review scope that was narrowed in the Rwanda era, the current account deficit converges toward −1% of GDP, and sterling holds its trade-weighted position through the autumn gilt issuance period. If all three conditions are met and the QGI score persists at rank 1 through two vintage updates without an acute event, that would represent evidence against the model's discriminative validity at this end of the distribution.

This is a structural resemblance finding, not an event prediction. The model's honest AUC of 0.7346 means roughly that in a two-country comparison between a high-scorer and a low-scorer, it calls the crisis country correctly about 73% of the time at the population level. It does not speak to timing, magnitude, or the probability of any specific event in the next twelve months. Investors and risk desks should use it as one input among many, specifically as a cross-indicator structural flag that surfaces a configuration not visible from any single macro variable.