1. Introduction

International institutions do not merely coordinate diplomacy. Following North (1990), they reduce uncertainty, lower transaction costs, and compress the price of protection against adverse states of the world. When those institutions lose credibility, markets need not crash immediately. Instead, the cost of calm can rise: the implicit put protection that institutional architecture once provided is withdrawn, and the market must purchase replacement insurance at market rates.

Between 2013 and 2019, the CBOE VIX regularly reached readings of 9–13 in tranquil episodes. Since 2022, comparable calm conditions produce readings of 15–19. The 252-day rolling 10th percentile of the Bloomberg CBOE VIX (hereafter VIX p10) rose from a mean of 12.74 in the 2012–2017 era to 15.02 in the 2022–2026 era. It did not return to pre-2020 levels even when Federal Reserve policy stabilized and credit spreads compressed. This is the empirical puzzle: the floor moved, and it stayed moved. The standard explanations — monetary tightening, pandemic aftermath, China slowdown — do not explain the timing. The WTO Appellate Body paralysis (December 2019) shifts the floor +2.59 points 26 months before Federal Reserve tightening began. This paper proposes that markets stopped treating institutional stability as permanent infrastructure and began pricing its fragility as a persistent risk factor.

The paper is organized as follows. Section 2 documents the institutional erosion episodes (WTO, NATO, OPEC) and provides a taxonomic note distinguishing rules-based governance institutions from coordinating cartels. Section 3 reviews the related literature, drawing on institutional economics (North, Keohane, Ikenberry), political risk pricing (Pástor-Veronesi, Baker-Bloom-Davis, Hassan et al.), the irreversibility tradition (Bernanke, Dixit-Pindyck, Higgs), and the variance risk premium (Bollerslev-Tauchen-Zhou, Bekaert-Hoerova). Section 4 develops the IIP conceptual framework with a simple formalization. Section 6 describes the data, the measurement approach, and the planned inter-rater reliability protocol for the IEI. Section 7 presents descriptive facts including the hero figure of regime-conditional VIX distributions. Section 9 reports empirical results: §§9.1–9.6 present the Bloomberg baseline OLS, structural breaks, domain regressions, quantile regressions, and horse race. Section 9.9 extends the Bloomberg panel results to rates volatility, term premia, inflation, and gold (proprietary data) across six asset classes. Section 9 addresses identification and presents three placebo tests. Section 10 concludes.

2. Institutional Context

2.1 The WTO Appellate Body

The WTO Appellate Body ceased to function as a binding second-instance arbiter on December 11, 2019, when U.S. blockage of new appointments left it without the quorum of three judges required to decide appeals.¹ Since that date, losing parties at the panel stage have been able to file appeals that go unresolved — effectively appealing into a legal void, in the language of the European Parliament Research Service (2024). Van den Bossche (2024) frames the loss precisely: the WTO dispute settlement system was valuable because it transformed trade conflict from retaliation into adjudication; its paralysis therefore changes not only legal procedure but the expected cost of trade conflict. WTO dispute filings fell to approximately one-third of their pre-2019 volume.² As Havertz (2026) observes, without a functioning appellate layer the system can no longer guarantee final resolution of trade disputes, weakening the insurance value of rules-based trade governance.

2.2 NATO and the Security Guarantee

The Trump administration's second term introduced explicit transactionalism into the NATO security commitment, conditioning U.S. guarantees on allied defense spending levels and publicly characterizing Greenland as a potential strategic acquisition. The deterrent value of Article 5 derives from its unconditional character; conditionality converts the security guarantee from insurance into a contingent contract — a qualitatively different instrument. European rearmament responses, including previously inconceivable spending targets, reflect recalibration of a lost put rather than its restoration.

2.3 OPEC and Supply Coordination

On April 28, 2026, the UAE — one of OPEC's largest producers — announced its departure, effective May 1, 2026, in what Reuters described as the largest oil-producer exit in the cartel's history.³ The announcement followed years of production-quota disputes with Saudi Arabia and the UAE's desire, in the words of Energy Minister Suhail al-Mazrouei, to operate "outside any constraint" (CNBC, April 28, 2026). Combined with prior coordination failures — the March 2020 price war and persistent non-compliance — the UAE exit materially weakens OPEC's capacity to buffer energy supply variance, raising the structural volatility of the commodity that underlies most macroeconomic forecasts.

2.4 A Taxonomic Note: Governance Institutions and Coordinating Cartels

One referee correctly observes that OPEC differs categorically from the WTO and NATO: it is a producer cartel without legal enforcement, not a rules-based multilateral institution. We retain OPEC as a fourth pillar of the IEI on a narrower, financial-economic ground that does not require equating it with governance institutions. The IIP framework conceptualizes any credible coordination mechanism that compresses the variance of a globally priced systemic input as a source of implicit put protection. OPEC's value to financial markets has never been governance legitimacy; it has been the reduction of realized oil-supply variance and the lowering of expected variance through credible quota signaling (Hamilton 2009). When that coordination weakens — through the 2020 price war, persistent compliance failure, and now the UAE's departure — the variance of oil supply rises and, by transmission, so does the variance of every macroeconomic forecast that takes oil prices as an input. The mechanism is functional, not institutional in the North-Keohane sense. We make this distinction explicit in the domain regressions of Section 8, which report the security, energy, and financial-architecture channels separately rather than aggregating them.

3. Related Literature

3.1 Institutions as Uncertainty-Reducing Mechanisms

The foundational insight for our framework comes from North (1990, 1991), who defines institutions as "the humanly devised constraints that structure political, economic, and social interaction" and argues that they exist primarily "to create order and reduce uncertainty in exchange."⁴ In North's account, the economic value of institutions is not principally normative — it is informational and structural: institutions lower the variance of expected outcomes by establishing credible rules that constrain behavior. Applying this logic to international finance, we argue that the WTO, NATO, and OPEC generated measurable financial value by compressing uncertainty in trade, security, and energy domains. Their credibility loss should therefore be visible in asset prices — specifically in implied volatility floors.

Keohane (1984, 1988) extends the institutional logic to international regimes, demonstrating that cooperation can persist even after hegemonic decline because regimes "reduce transaction costs of legitimate bargains," maintain informational infrastructure that lowers asymmetries, and raise the cost of defection. The IIP framework applies Keohane's transaction-cost logic to option pricing: when regimes lose credibility, the cost of insuring against the disorder they previously prevented rises. Ikenberry (2001) further argues that the postwar liberal order was built on institutions that "lock in" hegemonic commitments, lowering the variance of expectations for both leader and follower states. Our framework adopts this insight: the value of institutional lock-in is partly financial, expressed in the price of insurance against the volatility that unlocked commitments would generate. Mearsheimer's (1994/95) skepticism about institutional efficacy provides the natural null hypothesis: if institutions are mere reflections of underlying power and have no independent constraining force, their formal erosion should not move asset prices conditional on the underlying balance of power. The IIP framework is testable in part because it disagrees with this null.

3.2 Political Uncertainty and Implicit Put Protection

The most direct financial-economics precedent for our argument is Pástor and Veronesi (2013), who show that "political uncertainty commands a risk premium" and — crucially — that "political uncertainty reduces the value of the implicit put protection that the government provides to the market."⁵ We borrow this language of implicit put protection and extend it to the international institutional order: WTO enforcement, NATO deterrence, OPEC coordination, and dollar-system neutrality collectively functioned as implicit puts on global tail risk. Their degradation is consistent with withdrawal of that put, raising the market's self-insurance cost.

Baker, Bloom, and Davis (2016) provide the empirical bridge between policy uncertainty and market outcomes, demonstrating that policy-related uncertainty raises stock-market volatility and depresses investment. The IEI is conceptually distinct from EPU: EPU measures policy noise; the IEI measures the erosion of the institutional mechanisms that previously contained policy conflict. Bloom (2009) further shows that uncertainty shocks can generate real effects by raising the option value of waiting. Institutional erosion differs from these shocks: it is a slow-moving deterioration in the rule architecture that determines how future shocks will be resolved, not a one-off impulse with predictable mean-reversion. Hassan et al. (2019) provide micro-level evidence of this distinction by constructing firm-level political risk from earnings-call transcripts and showing that it commands a return premium — confirming that the political-uncertainty channel is priced at the firm as well as the aggregate level.

3.3 Irreversibility, Regime Uncertainty, and the Demand for Insurance

The microfoundation for why institutional erosion should affect the volatility floor rather than the unconditional mean comes from the irreversibility literature. Bernanke (1983) shows that under irreversible investment with uncertainty, agents have an incentive to delay action — the "option to wait" — and that this incentive rises with the variance of the relevant state. Dixit and Pindyck (1994) formalize this as a real-options framework: any decision with irreversible commitment under uncertainty becomes more valuable to defer when uncertainty rises. Institutional credibility loss is precisely the kind of structural uncertainty that this literature addresses: unlike a transitory news shock, the disappearance of an adjudicator or a security guarantee is not expected to mean-revert. The option value of insurance against such uncertainty is therefore persistent, which translates into a persistent shift in the floor of variance pricing rather than a temporary spike.

Higgs (1997) introduces the concept of regime uncertainty: the prospect that the institutional rules governing property rights, contract enforcement, and the cost of doing business may be altered unpredictably. Higgs originally applied this concept to the New Deal era to explain delayed recovery; we apply it to the post-2019 international order. When the WTO Appellate Body ceases to function, when Article 5 becomes conditional, when OPEC quotas no longer bind, and when reserve currency access can be revoked, agents face a regime-uncertainty problem at the international level. The IIP framework is essentially regime uncertainty priced in implied volatility space.

3.4 Variance Risk Premia and the Price of Insurance

The VIX is not merely a forecast of realized volatility — it embeds a variance risk premium (VRP): the compensation investors require to warehouse volatility risk. Bollerslev, Tauchen, and Zhou (2009) show that the difference between implied and realized variation "explains a non-trivial fraction of stock market returns," establishing the VRP as a priced risk factor across asset classes.⁶ Bekaert and Hoerova (2014) decompose the VIX into a conditional variance component and a risk premium component, demonstrating that the latter is the more informative predictor of future returns and bears the signature of risk aversion shifts. The IIP framework targets precisely this second component: institutional erosion is hypothesized to operate through the risk-aversion / risk-premium channel, not through expected realized variance. Corradi, Distaso, and Mele (2013) document that the VRP is "strongly countercyclical," rising in adverse economic states. The IIP hypothesis extends this logic: institutional erosion may raise the floor of the VRP even in the absence of acute crises — making calm structurally more expensive. Prior BIS analysis found that "the reduction in volatility represents to a considerable extent the consequence of improvements in the functioning and structure of financial markets" (BIS 2006); we hypothesize the reverse operates with institutional deterioration.⁷

At the cross-country level, Hartwell (2018) provides direct evidence that "more advanced institutions help to dampen financial sector volatility," supporting the view that institutional quality is a determinant of volatility levels. Our paper applies this logic to the international institutional order and, specifically, to implied volatility floors.

3.5 Weaponized Interdependence and Financial Architecture

The financial architecture component of the IEI draws on Farrell and Newman (2019), who demonstrate that "states are able to weaponize interdependence" when global economic networks contain hubs and chokepoints, using asymmetric structure to gather information or deny access to adversaries. The freezing of Russian central bank reserves in 2022 represents precisely this: what functioned as neutral settlement infrastructure became a source of conditional access risk. Drezner, Farrell, and Newman (2021) document how "the infrastructure of globalization" becomes "a source of strategic leverage." In the IIP framework, the moment the dollar system's neutrality becomes conditional, any holder of dollar-denominated assets must price that conditionality — raising the floor of the FX, rates, and equity volatility insurance they require.

3.6 Geopolitical Risk: Distinction from the IEI

Caldara and Iacoviello (2022) construct a news-based measure of "adverse geopolitical events and associated risks" that predicts lower investment and employment. The IEI is conceptually distinct: GPR measures the shock; the IEI measures the erosion of the shock absorber. A trade war is a GPR event; the paralysis of the WTO Appellate Body is an IEI event. We include the Caldara-Iacoviello GPR as a control in horse-race specifications (Section 9.5). For the energy domain specifically, Hamilton (2009) shows that oil price uncertainty operates as a quasi-monetary tax on real activity; if the IEI's energy channel captures the loss of OPEC's variance-compression role, we should observe its effect partly via the same channel Hamilton documents — supply uncertainty raising downstream volatility. Our domain regression result — that all four channels are individually significant in the Bloomberg panel, with trade the weakest by coefficient size and finance/energy the strongest — is consistent with markets sector-pricing systemic institutional risk through corporate repricing, while harder-to-hedge systemic channels represent residual structural exposures that raise the broad VIX floor.

4. Conceptual Framework: The Institutional Insurance Premium

4.1 Institutions as Systemic Risk Absorbers

In standard option pricing, implied volatility reflects both the expectation of realized volatility and the market price of insurance against variance risk. Following North (1990) and Keohane (1984), international institutions reduce both components: they lower the realized variance of outcomes in their domain, and they lower the price of protection against the remaining variance. A party operating under a credible WTO Appellate Body has less reason to purchase insurance against trade retaliation, because the institutional mechanism provides a cheaper alternative — adjudication. When that mechanism disappears, the insurance price rises, even if the probability of actual trade conflict is unchanged. Pástor and Veronesi's (2013) implicit put protection concept captures exactly this mechanism at the national level; the IIP generalizes it internationally.

4.2 A Simple Formalization

Let θ_t ∈ [0, 1] denote institutional credibility at time t, where θ = 1 corresponds to full credibility and θ = 0 corresponds to complete collapse. Let ξ_t denote an exogenous tail-risk shock with conditional variance σ²_ξ. Following the Bernanke (1983) / Dixit-Pindyck (1994) irreversibility logic, the institutional architecture absorbs a fraction θ_t of the institutionally-insurable component of the shock's variance. Crucially, however, even under full credibility (θ = 1) a residual variance σ²₀ > 0 remains — the irreducible, non-insurable background variance of an open system (Hartwell 2018; BIS 2006). Total realized variance facing the market is therefore:

This formulation is realistic: institutions reduce the fraction of variance that is institutionally insurable, but do not eliminate variance. The market's risk-neutral demand for variance insurance, under standard CRRA preferences with risk-aversion coefficient γ, is proportional to the conditional variance plus a premium for variance risk itself (Bollerslev, Tauchen and Zhou 2009). Implied variance then satisfies:

where λ > 0 is the variance risk premium loading and the third term is the regime-uncertainty premium — the market's compensation for not knowing the future trajectory of θ. The IIP framework predicts that institutional erosion operates through two credibility-related terms: a fall in θ raises the conditional variance, and the irreversibility of the erosion (Higgs 1997) raises the variance of the perceived future θ. The latter is the channel that explains why the floor rises specifically: even in states where the realized shock ξ_t is small or zero, the regime-uncertainty premium remains elevated because it does not mean-revert when no shock arrives. Differentiating with respect to θ:

This is the testable comparative static: institutional credibility erosion should compress the lower tail of the implied volatility distribution upward (because the regime-uncertainty premium survives the absence of shocks), while its effect on the unconditional mean depends jointly on (a) the realized variance channel and (b) any change in the arrival rate or persistence of shocks. The narrow Pástor-Veronesi (2013) intuition predicts that ∂(IV_mean)/∂θ is small in calm states; a more general reading recognizes that institutional erosion may also raise the frequency of resolution-pending events, in which case the entire distribution shifts. The empirical evidence in Section 9 finds support for the floor prediction in the cumulative-IEI specification, a broad-based distributional shift in the quantile regression (consistent with the more general reading), and clear support in the Chow break tests.

4.3 Stock vs. Flow: Why Cumulative IEI, Not Monthly Flow

The IEI admits two natural specifications: monthly event flow (IEI_t) and cumulative stock (Σ_s≤t IEI_s). Under the IIP framework, the relevant variable is the stock: what markets price is the accumulated loss of credibility, not the most recent event. A single tariff announcement is information; the cumulative weight of four years of tariff escalation, WTO blockages, and producer-cartel exits is a regime change. This is why Equation (2) takes θ as a slow-moving state variable rather than a flow shock. Empirically, the cumulative IEI is the specification that enters the baseline regression with β = 0.0898***, p < 0.001; the AR(1) spec yields β = 0.007, p = 0.195 (AR1); the contemporaneous monthly flow is not separately significant. This attenuation is consistent with the interpretation that VIX-floor persistence dominates short-run variation once lag structure is introduced. The AR(1) result (p = 0.195) is best treated as a persistence artefact rather than a refutation of the stock-channel reading, though it should be interpreted cautiously.

4.4 Testable Implications

The IIP framework generates five testable predictions: (i) the cumulative IEI, not the monthly flow, should be positively associated with the VIX floor conditional on macrofinancial controls; (ii) the effect should be concentrated in domains with harder-to-hedge systemic channels (security, energy, financial architecture) rather than in trade governance where sector-level repricing is feasible; (iii) effects should be persistent rather than transitory, reflecting credibility loss rather than event spikes; (iv) the effect should not replicate in domestic-only political uncertainty series; (v) the IEI should retain explanatory power conditional on the Caldara-Iacoviello GPR and the BBD EPU, indicating the institutional channel is distinct from event-based geopolitical risk and policy noise. We test predictions (i)–(iv) in Section 8 and add a horse-race test of (v) in Section 9.5.

where VIX_p10_t is the rolling 252-day 10th percentile of the CBOE VIX, IEI_Cumul_t-1 is the lagged cumulative Institutional Erosion Index, and X_t contains the Federal Funds Rate, its change, CPI inflation, and the HY spread. We use Newey-West standard errors (12 lags) throughout, following Newey and West (1987).

6. Data and Measurement

6.1 Volatility Floor Measure — Bloomberg CBOE VIX

Why the rolling 10th percentile? The rolling 252-day 10th percentile — not the spot VIX or its conditional mean — is the appropriate measure for the IIP hypothesis. Under the regime-uncertainty channel, institutional credibility loss raises the lower bound of the volatility distribution: the price of calm in normal states. This is mechanically distinct from raising the mean (which would require more frequent crises) or the upper tail (which would require larger crises). The 10th percentile captures the structural floor: the VIX level that markets treat as the baseline cost of insurance when nothing obvious is wrong.

Why 252 trading days? The 252-day window corresponds to one trading year — long enough to smooth over short-duration crisis episodes (which inflate the mean) while remaining sensitive to regime shifts that persist for multiple quarters. A 63-day window would be too noisy; a 504-day window would be too slow to detect the 2019–2022 regime break. The 252-day specification is standard in the realized-volatility literature and is not cherry-picked: robustness to 126-day and 504-day windows yields qualitatively similar regime shifts (available in the econometric appendix).

Why not the spot VIX mean? The spot VIX is dominated by acute stress episodes that reflect geopolitical events (GPR) and policy surprises (EPU) — channels the IIP hypothesis explicitly distinguishes from. Using the mean would conflate the institutional insurance channel with the event-based risk channels the paper is designed to separate. The floor is the institutional layer; the spike is the event layer.

We use the Bloomberg CBOE VIX series (5,699 daily observations, 2004-07-21 to 2026-05-25). The rolling p10 is computed as the calendar-day 10th percentile over the trailing 252 trading days. Bloomberg is the unified data source; FRED VIXCLS produces quantitatively similar floor estimates and is used as a public-source cross-check where noted.

6.2 Institutional Erosion Index

We hand-code 43 events (v2 of the IEI, after correcting the UAE/OPEC date from May 2025 to April 2026 following external review) across four domains, each scored 1–3 on a procedural / substantive / structural ladder. The cumulative IEI tracks the accumulated stock of institutional credibility erosion. The full event list is in Appendix A and the coding protocol is in Appendix B. To distinguish IEI from GPR (Caldara and Iacoviello 2022) and EPU (Baker, Bloom, and Davis 2016), we note that GPR codes adverse events, EPU codes policy uncertainty, while the IEI codes erosion of mechanisms. An invasion is a GPR event; a tariff escalation is an EPU spike; the failure to reconstitute a judicial body is an IEI event.

Coding circularity and inter-rater reliability. Hand-coding by the same researcher who develops the framework raises a legitimate concern that scores may be assigned with knowledge of subsequent market outcomes, biasing the IEI upward in periods of observed volatility. We address this in three ways. First, the coding protocol (Appendix B) specifies ex-ante criteria — quorum loss, formal withdrawal, public conditionalization of treaty obligations, cartel-member departure — that are observable independent of market reactions. Second, we plan a second independent coding pass by an external collaborator and will report Cohen's κ in a subsequent revision; if κ falls below 0.7 we will recalibrate the protocol. Third, the placebo test using domestic-only political events (Section 9.2, Placebo 3) provides indirect validation: if the IEI were primarily picking up ex-post-justified narrative coding, a domestic political index coded by the same researcher with the same retrospective awareness should also predict the VIX floor, and it does not (β = 0.126, p = 0.707).

6.3 Controls

Fed Funds Rate (FRED FEDFUNDS, Board of Governors; public domain; to control for monetary policy regime). HY OAS (FRED BAMLH0A0HYM2; daily May 2023–May 2026; annual averages pre-2023; proxies for credit cycle and financial conditions). CPI inflation (BLS year-on-year). Caldara-Iacoviello GPR (available at policyuncertainty.com; included in horse-race specifications in Section 9.5 to separate the IEI from event-based geopolitical risk). Baker-Bloom-Davis EPU (policyuncertainty.com; included in the same horse race to separate the IEI from policy noise).

Sample correction note (v6.1). An earlier version of this paper inadvertently dropped four observations (Jan–Apr 2023) due to a merge artifact between the daily HY-OAS series (which begins in May 2023 in our primary source) and the monthly VIX/FFR series. The corrected sample of N = 173 monthly observations is used throughout the present version. The HY spread values for Jan–Apr 2023 use FRED's BAMLH0A0HYM2 monthly averages (4.20%, 4.10%, 4.85%, 4.65%), with the March 2023 spike reflecting the SVB/banking-stress episode. All headline results survive the correction: the cumulative IEI coefficient on the VIX floor is β = 0.0898*** (p < 0.001) in the baseline OLS on the full real-data sample, and the Chow break-test significance ordering is preserved.

6.4 Bloomberg Multi-Asset Panel

Sections 9.7–9.9 add three cross-asset robustness extensions. FRED Treasury series: ACM term premium (THREEFYTP10, 9,079 daily obs) and T5YIFR (5,852 daily obs), used in §§9.7–9.8. Bloomberg Multi-Asset Panel: 5,699 daily Bloomberg observations (2004-07-21 to 2026-05-25) across 15 series, aggregated to N = 173 monthly observations (January 2012 – May 2026). Series include: CBOE VIX, Bloomberg MOVE Index, ACM and Kim-Wright 10Y term premia, USD 5y5y inflation swap (InfSwap5Y5Y), 5y5y breakeven inflation, Gold spot $/oz, TIPS 10Y real yield, Treasury 2Y/10Y/30Y, FCI, Treasury bid-ask spread. All floor measures use the rolling 252-day 10th percentile from the full daily series — same methodology as the FRED analysis. The Bloomberg panel is analyzed in §9.9.

8. Hard Data — Empirical Facts Across Asset Classes

The following statistics are computed directly from Bloomberg market data and document the raw level shifts that motivate the formal analysis. These are facts, not estimates. (5,699 daily observations, 173 monthly, January 2012 – May 2026). These are not "stylized" approximations — they are the actual numbers from the data that motivated this paper.

Several patterns stand out. The VIX floor rose from 12.73 to 15.02 — a shift of +2.29 points — and has not returned to pre-2020 levels even during 2023–2024, when the Federal Reserve held rates steady and HY spreads compressed to multi-year lows. The MOVE floor rose by roughly 43% from 60.1 to 86.0, indicating that the rates volatility floor shifted dramatically after the 2022 institutional repricing. Gold rose from a mean of $1,336 to $2,620/oz, consistent with rising demand for insurance outside the institutional system. The ACM term premium turned negative in the transition era and has partially recovered, reflecting the competing forces of flight-to-safety demand and sovereign risk repricing. Crucially, the 5-year forward inflation swap remained anchored at 2.53% throughout — weighing against a simple inflationary-uncertainty narrative for the VIX and MOVE floor shifts. These facts are the empirical backbone of the IIP hypothesis.

The persistence of the VIX and MOVE floors during periods of macro calm — and the simultaneous rise in gold — is the primary empirical motivation for the IIP hypothesis. The following sections test whether cumulative institutional erosion (IEI) can account for these patterns after controlling for monetary policy, inflation, and credit spreads.

9. Empirical Results

The IIP signal is strongest not where macro expectations are revised, but where insurance against regime uncertainty is priced: equity volatility floors, rates volatility, and stores of value outside the institutional architecture.

9.1 Baseline OLS — Descriptive Benchmark

Table 1 reports the baseline OLS regression with Newey-West (1987) standard errors (12 lags). We present this as a descriptive benchmark, not a causal identification strategy; the preferred strategy is a cross-asset panel with time fixed effects that absorbs the common monetary shock. The IEI cumulative (lagged 1 month) enters with β = 0.0898*** (p < 0.001). Controls: 10Y UST yield (β = -1.9755***, p < 0.001) and 5y5y inflation swap (β = 4.4362***, p < 0.001) are significant. R² = 0.435, N = 173.

9.2 Structural Break Tests

As a preliminary structural-break exercise, Table 3 reports Chow F-statistics at six candidate dates corresponding to known institutional and policy episodes. This is not a full Bai and Perron (1998) procedure — which estimates the number and timing of breaks endogenously — but a more limited exercise that tests pre-specified candidate dates one at a time, with the same controls held constant. A formal Bai-Perron multiple-break estimation remains for the next revision. With that caveat: the December 2019 WTO paralysis is associated with the largest unconditional floor shift (+2.59 VIX points; F = 14.43, p < 0.001) among the candidate dates we test. The February 2022 Ukraine invasion is also significant (+2.33 points; F = 9.92, p < 0.001). Importantly, the December 2019 break is economically significant even before the coincident Fed tightening cycle — consistent with a pre-monetary-cycle floor shift. The significance of multiple dates is consistent with the IIP framework's prediction of cumulative, multi-episode institutional erosion.

9.3 Domain Regressions — All Four Institutional Channels Significant

Table 4 reports separate Bloomberg-panel regressions for each IEI domain. In the unified Bloomberg panel, all four institutional domains are individually significant at p < 0.001. Finance and energy carry the largest coefficients (β = 0.5311*** and 0.5038*** respectively). Security is intermediate (β = 0.3570***). Trade is significant but the weakest channel by coefficient size (β = 0.1909***). The broad-based significance across all four channels is consistent with the IIP framework: institutional erosion operates through multiple systemic risk channels simultaneously, not through any single domain.

9.4 Quantile Regressions — Honest Assessment

Following Koenker and Bassett (1978), quantile regressions support the view that the IEI is positively and significantly associated with VIX at all quantiles (τ = 0.10 to 0.90; all p < 0.001). Coefficients range from 0.156 to 0.277. A narrow reading of the IIP that would require the lower tail to be the most affected is not supported: β(τ=0.90) = 0.229 exceeds β(τ=0.10) = 0.182. The pattern in the data is a broad-based upward shift in the entire VIX distribution, not exclusively in the calm-state floor. The revised formalization in Equation (3) accommodates this: institutional erosion is associated with a higher floor through the regime-uncertainty premium channel, while its association with the mean and the upper tail depends jointly on the residual variance channel and on whether realized shocks become more frequent or persistent. The hero figure (Figure 2) is consistent with this reading visually: the entire distribution shifts to the right, with the floor moving from VIX ~9–11 to VIX ~13–14, while the upper-tail mass also thickens. The honest characterization is that the data are consistent with a generalized IIP — both channels active — rather than with a floor-only version.

9.5 Horse Race — IEI vs. GPR vs. EPU

A natural concern is that the IEI is simply re-labeling what existing geopolitical and policy uncertainty indices already capture. Table 5 reports a horse-race specification in which the Caldara-Iacoviello (2022) GPR index, the Baker-Bloom-Davis (2016) EPU index, and the cumulative IEI enter the regression jointly, with the same macrofinancial controls. The cumulative IEI retains a positive coefficient (β = 0.062, p < 0.05) conditional on GPR; notably, controlling for GPR sharpens the IEI estimate, consistent with GPR partially suppressing the institutional channel. GPR itself enters with a negative and weaker coefficient (β = 0.011, p = 0.094), consistent with its event-based construction picking up acute shocks rather than persistent regime shifts. EPU enters with a small positive coefficient that is not statistically significant once the IEI is included (β = 0.009, p = 0.241). The point estimates and the relative R² gain from adding the IEI (ΔR² = +0.043) are consistent with the IEI capturing a distinct dimension of uncertainty: the slow-moving erosion of mechanisms, as opposed to the flow of geopolitical events (GPR) or policy noise (EPU).

9.5b Anti-Trend Robustness — Is the IEI Just a Time Trend?

Because the cumulative IEI rises monotonically over the sample, a natural concern is that it is acting as a generalized post-2019 time trend that picks up any persistent upward movement in the VIX floor. Table 9.5 reports four specifications designed to discriminate the institutional signal from a generic time trend.

The IEI coefficient remains strong in the Bloomberg baseline and time-trend specifications (specs 1–3). The Bloomberg AR(1) model (spec 4) absorbs much of the VIX-floor persistence, attenuating the IEI to β = 0.0074 (p = 0.195); this is consistent with the institutional channel operating at the regime level rather than through monthly event flows. First-difference and flow specifications (5–6) are negative and do not provide a positive institutional signal, supporting the stock-vs-flow channel: markets appear to price accumulated credibility loss, not monthly institutional news flow.

9.6 Cross-Country Equity Volatility Panel — Identification Strategy in Progress

The cleanest identification strategy for the IIP framework is a cross-country panel exploiting heterogeneous institutional exposure with time fixed effects that absorb the common monetary shock. Following the logic of Hartwell (2018), we expect that small open economies more dependent on multilateral trade rules (Korea, Netherlands, Singapore) should display a larger post-2019 implied-volatility floor shift than large, more closed economies (United States, Brazil) — provided the IIP channel operates through institutional dependence rather than through generic global risk. Constructing this panel requires daily IV surfaces or country-ETF option data (e.g., EWY for Korea, EWN for Netherlands, EWS for Singapore, SPY for the US, EWZ for Brazil), and exposure scores derived from public administrative data:

Constructing exposure scores from these public administrative sources removes the subjectivity that would arise from researcher-assigned weights. The expected sign of the interaction Exposure_i × IEI_Cumul_t is positive: countries with higher exposure to a given institutional pillar should experience larger floor shifts when that pillar erodes. We treat this panel as the primary identification strategy for a future revision.

9.4 A Research Agenda: From Exploratory Framework to Empirical Programme

The limitations above define a tractable research agenda. Each represents a next step rather than a fundamental obstacle.

The current paper establishes the IIP framework and presents exploratory time-series evidence. The agenda above maps the path from exploratory hypothesis to a credible empirical programme.

9.7 Cross-Asset Extension I — MOVE and Treasury Term Premium

We test the IIP hypothesis on the Treasury term premium (FRED THREEFYTP10, 9,079 daily observations, rolling 252-day p10). In baseline OLS, the IEI is not significantly associated with the term premium (β = −0.005, p = 0.462). Quantile regressions, however, reveal a highly significant distributional asymmetry: the IEI is associated with a lower floor (τ = 0.10: β = −0.014, p < 0.001) and a higher ceiling (τ = 0.90: β = +0.005, p = 0.002). Institutional erosion widens the term premium distribution — consistent with amplified flight-to-safety dynamics: in calm states, Treasuries are bid harder; in stress states, risk premia spike more violently.

The Ukraine structural break (February 2022) shifts the term premium p10 upward by +0.166 points (t = 4.42, p < 0.0001), implying even calm-state term premium is higher post-invasion — consistent with persistent sovereign risk repricing. The WTO break (December 2019) produces a term premium p10 shift of −0.167 points (p = 0.0002), reflecting intensified safe-haven demand before the 2022 monetary repricing. Together, these directional shifts explain the OLS null: the net level effect is near zero, but the distributional change is large.

9.8 Cross-Asset Extension II — Inflation Expectations (Useful Null)

We test the IIP hypothesis on 5-year, 5-year forward inflation expectations (FRED T5YIFR, 5,852 daily observations). The IEI is not significantly associated with inflation expectations in any specification: baseline OLS β = −0.007 (p = 0.135); rolling p10 p = 0.362; rolling p90 p = 0.216. The WTO Appellate Body break (December 2019) produces zero inflation response (Δ = −0.045, p = 0.356), while producing the largest VIX-floor shift among the candidate dates tested (+2.59 pts, p < 0.0001). The Ukraine invasion (February 2022) produces a strong inflation shift (+0.149 pts, p < 0.0001) — but this reflects the energy and supply shock channel, not the institutional erosion channel.

This null result is informative for identification. If the IEI were a generic macro uncertainty proxy, it would predict both equity volatility floors and inflation expectations. It predicts the former but not the latter. The specificity of the result for variance insurance pricing — and not for expected inflation levels — is evidence that the IIP captures a distinct channel. The clean dissociation between the WTO break's equity signal and its inflation non-signal is the most direct test of this specificity available in the current data.

9.9 Cross-Asset Results — Bloomberg Multi-Asset Panel

The unified Bloomberg panel allows cross-asset tests of the IIP hypothesis across five additional series of 5,699 daily observations across 15 financial series, aggregated to N = 173 monthly observations (January 2012 – May 2026). The Bloomberg data provides stronger proprietary-data evidence consistent with the IIP framework.

Panel A — VIX: Bloomberg VIX p10 β = 0.0898*** (SE = 0.0172, p < 0.001, R² = 0.435). Bloomberg baseline: β = 0.0898*** (p < 0.001, R² = 0.435), reflecting the Bloomberg daily series' greater precision in computing the rolling floor. The WTO Appellate Body paralysis (Dec 2019) shifts the VIX p10 by +2.59 points (F=14.43, p<0.001).

Panel B — MOVE: OLS null (MOVE p10: β = 0.1047, p = 0.500). But the structural breaks are among the largest in the paper: WTO break +16.3 points (t = -5.64, p < 0.0001); Ukraine break +31.9 points (t = -12.03, p < 0.0001). The OLS null plus discrete structural jumps is consistent with the IIP framework for mean-reverting series: rates volatility reprices institutionally in regime shifts, not gradual drift.

Panel C — Term Premium: Bloomberg ACM p10 β = -0.0128** (p = 0.025); Kim-Wright β = -0.0048*** (p = 0.005, R² = 0.757). Quantile regressions reveal that the IEI is uniformly negative and significant across all quantiles (τ = 0.10 to 0.90, all p < 0.001). Unlike the FRED bipolar pattern (floor falls, ceiling rises), Bloomberg shows pervasive safe-haven compression: institutional erosion uniformly reduces term premia at all distributional states.

Panel D — Inflation: 5y5y USD inflation swap β = -0.0070*** (p < 0.001, R² = 0.815). The sign is negative: institutional erosion is associated with lower inflation expectations. This inverts the FRED T5YIFR null (β = −0.007, p = 0.135) to a significant result. The mechanism is risk-off: as institutional credibility deteriorates, flight-to-safety flows depress growth and inflation forecasts. This is the opposite of a supply-shock uncertainty index, supporting the IIP specificity claim.

Panel E — Gold: Gold p10 β = 18.51*** (SE = 2.39, R² = 0.829) — the strongest OLS result in the paper. Each 1-unit increase in cumulative IEI is associated with a $18.51 rise in the gold floor. Over the 2016–2025 period (IEI 0→81), this implies an IEI-attributable gold floor contribution of approximately $1,499/oz, or 115% of the 2016 gold price of ~$1,300. Gold is the cleanest proxy for "insurance outside the institutional system." Note: Bloomberg is the unified market-data source. Selected FRED series are retained as public-source cross-checks for rates and inflation.

10. Identification Challenges and Placebo Tests

9.1 The Monetary Policy Confound and Bad Controls

The most serious alternative explanation is the Fed's post-2022 tightening cycle. We acknowledge that controlling for the Federal Funds Rate may be a "bad control" in the Angrist-Pischke sense: the post-2022 tightening cycle was itself partially a response to the inflation shock triggered by the Russian invasion, which is also an IEI event in our coding. Mechanically conditioning on the FFR may therefore absorb part of the IEI's true effect through the inflation channel. We address this in three ways. First, we present specifications with and without the FFR; the IEI coefficient survives both. Second, the December 2019 WTO break shows a floor shift of +2.59 VIX points more than two years before the 2022 tightening began, and this break is the largest Chow F-statistic in the sample — providing the cleanest pre-monetary-cycle evidence. Third, the domain regressions find significance in the security and energy channels, which have no direct monetary transmission. The ECB (2017, 2025) documents that policy uncertainty can decouple from financial volatility — particularly when equity momentum is strong — consistent with a delayed or floor-specific institutional repricing.

9.2 Three Placebo Tests

9.3 What the Evidence Cannot Claim — Explicit Limitations

We summarize the principal limitations of the current evidence, in descending order of importance for future revision.

11. Conclusion

This paper proposes the Institutional Insurance Premium (IIP) framework: the hypothesis that markets increasingly price institutional fragility as a persistent structural risk factor, not as a transient shock to be mean-reverted. The evidence is consistent with an upward shift in the global volatility floor associated with the cumulative erosion of the WTO, NATO, OPEC, and the neutrality of dollar-system access. The framework draws on North's (1990) view of institutions as uncertainty-reducers, Keohane's (1984) regime-based transaction-cost logic, and Pástor and Veronesi's (2013) implicit put protection concept — generalizing all three from domestic to international institutional order, and from equity returns to volatility floor pricing. The mechanism is microfounded in the Bernanke (1983) / Dixit-Pindyck (1994) tradition of irreversibility under regime uncertainty (Higgs 1997): when institutional credibility is permanently lost, the option value of self-insurance rises and does not mean-revert in the absence of new shocks.

Four findings stand out. First, the cumulative IEI is significant in the baseline OLS (β = 0.0898***, p < 0.001) and in the horse-race specification controlling for GPR and EPU (β = 0.062, p < 0.05). This pattern is supported by three placebo exercises. Second, candidate-date Chow tests find the largest conditional floor shift at the December 2019 WTO break (+2.59 points), predating the 2022 monetary tightening cycle; a full Bai and Perron (1998) multiple-break estimation remains for the next revision. Third, a horse-race specification against the Caldara-Iacoviello GPR and the Baker-Bloom-Davis EPU finds the cumulative IEI retains incremental explanatory power (β = 0.062, p < 0.05) when GPR is controlled, consistent with the IIP capturing a distinct dimension of uncertainty — the slow-moving erosion of mechanisms rather than the flow of events or policy noise. Fourth, domain regressions show that all four institutional channels are individually significant in the Bloomberg panel (p < 0.001), with finance and energy carrying the largest coefficients (β = 0.5311*** and 0.5038***) and trade the weakest but still significant channel (β = 0.1909***).

A cross-asset extension to Treasury term premium and inflation expectations provides identification-relevant evidence on the specificity of the IIP channel. The term premium (FRED THREEFYTP10) is not significantly associated with the IEI in OLS (p = 0.462), but quantile regressions reveal a distributional bifurcation consistent with institutional erosion amplifying flight-to-safety: the floor falls (τ = 0.10: β = −0.014, p < 0.001) while the ceiling rises (τ = 0.90: β = +0.005, p = 0.002). More importantly, 5-year, 5-year forward inflation expectations (FRED T5YIFR) are not significantly associated with the IEI in any specification (p = 0.135–0.362). The WTO Appellate Body break produces zero inflation response (Δ = −0.045, p = 0.356) despite producing the largest VIX-floor shift among the candidate dates tested (+2.59 pts, p < 0.0001). This dissociation weighs against the interpretation that the IEI is merely capturing generic macro uncertainty and supports the specificity of the IIP as a variance-insurance pricing channel.

Cross-asset tests on the Bloomberg panel (5,699 daily obs, N=173 months) confirm the IIP signature across asset classes: Gold is the strongest OLS result (β = $18.51***, R² = 0.829), consistent with institutional erosion driving demand for insurance outside the system. The MOVE Index OLS is null but the structural breaks are the largest in the paper (+16 and +32 points at WTO and Ukraine dates). Most importantly, the 5y5y inflation swap is significantly negative (β = -0.0070***), consistent with a risk-off/deflation channel rather than a supply-shock story — weighing against the generic uncertainty interpretation.

We present these findings as preliminary evidence consistent with the IIP framework, not a definitive causal demonstration. The paper's principal contribution is conceptual: it names a mechanism, embeds it in three established literatures (institutional economics, political risk pricing, variance risk premia), provides a simple formalization (Equations 2–3), and documents patterns in real data that motivate the cross-country panel identification strategy described in Section 9.6.

The market is not pricing the end of the international order. It is pricing the loss of free insurance once provided by that order.

Appendix A — Full IEI Event List (43 Events, v2 Corrected)

Severity: 1=procedural; 2=substantive; 3=structural. ✎ = corrected from prior version (UAE/OPEC: date changed from May 2025 to April 2026; score 2→3; source: Reuters/Al Jazeera/CNBC, April 28, 2026).
TTrade  SSecurity  EEnergy  FFinance

Appendix B — IEI Coding Protocol and Planned Inter-rater Reliability

Coding criteria. An event qualifies for inclusion in the IEI only if it satisfies all three of the following: (a) it represents a discrete, dated change in the binding force or operational capacity of an international institution or coordination mechanism listed in Table A1 (WTO, NATO/Article 5 framework, OPEC, the dollar-clearing system, sanctioned-reserve framework); (b) it is sourced to a primary public document (treaty text, press release, formal notification, or accredited news report) cited in Appendix A; (c) it is independent of subsequent financial market reactions in the assignment of its severity score.

Severity scoring. Score 1 (procedural) — administrative or rhetorical erosion without binding effect on mechanism operation (e.g., a tariff threat not implemented). Score 2 (substantive) — operational degradation that limits but does not eliminate the mechanism's binding force (e.g., a tariff round implemented; a temporary OPEC compliance breakdown). Score 3 (structural) — mechanism ceases to operate as designed (e.g., WTO Appellate Body quorum loss; UAE departure from OPEC; freezing of reserve assets converting the dollar system from neutral infrastructure to conditional access).

Inter-rater reliability protocol. A second independent coding will be performed by an external collaborator with no involvement in the framework design. Cohen's κ will be computed across all 43 events on the binary (include/exclude) and ordinal (1/2/3 severity) dimensions. The publication threshold is κ ≥ 0.7 on both dimensions. Events where the two coders disagree will be adjudicated by a third reader and reported in a revised appendix. We plan a third validation step: a GDELT-based event-density count for each institutional domain to provide a machine-coded alternative IEI; the correlation between hand-coded IEI and GDELT-density will be reported.

Robustness with alternative coding. All headline results will be re-run with: (i) the second-coder IEI, (ii) the GDELT-density IEI, (iii) an IEI that drops the top-five events by score (test of leverage to specific events), and (iv) an IEI that uses binary (any event vs. no event) rather than 1/2/3 scoring. Results that depend on a single coding decision will be flagged as such.