# FX Carry Crash Skewness: VIX & MOVE Regime-Switching

> Model negative return skewness in foreign exchange carry trades using ICE BofA MOVE bond volatility, VIX term structure, and 25-delta skew filters.

Canonical: https://fxmacrowire.com/wire/fx-carry-crash-skewness-vix-move-regime-switching
Last updated: 2026-11-14T19:34:15.020Z

**Publication:** FXMacro Wire
**Category:** Em Carry Terms of Trade Shocks
**Published:** 2026-11-13T14:25:25.915Z
**Author:** Henrik Lindqvist, CFA (CFA, MSc in International Monetary Economics (LSE))
**Reviewer:** Not recorded

## Answer-first summary

Foreign exchange carry portfolios violate Gaussian risk assumptions—“going up the escalator and down the elevator shaft” during funding squeezes. Combining ICE BofA MOVE bond volatility and 25-delta skew filters prevents catastrophic carry drawdowns.

## Key takeaways

- Unfiltered FX carry portfolios suffer severe negative skewness (-2.15) because leveraged pods share identical JPY/CHF funding shorts that trigger synchronized VaR unwinds.
- Gaussian 99% VaR understates true carry tail losses by ~80%; institutional risk desks enforce Cornish-Fisher Modified VaR incorporating skewness and kurtosis.
- ICE BofA MOVE (>115-120 bps) and 1M 25-delta JPY call skew typically flash warning signals 2 to 5 trading days ahead of equity VIX spikes or VIX1M > VIX3M backwardation.
- Hedging an EM carry book with lower-implied-vol 3-month 15-delta AUD/JPY puts captures >0.85 crash correlation at roughly half the vega carry bleed of direct EM NDF options.

## Related financial entities

- **BIS $9.6 Trillion Global OTC Foreign Exchange Turnover Benchmark** (BIS-FX-SWAPS-96T) — product: BIS $9.6 Trillion Global OTC Foreign Exchange Turnover Benchmark (BIS-FX-SWAPS-96T) refers to an institutional financial benchmark and regulatory standard monitored by BIS Triennial Survey (INTL), currently calibrated at $9.6T (Daily OTC Volume) as of October 2026.
- **USD/CHF & EUR/CHF Swiss National Bank 0.00% Rate Floor** (USDCHF-SARON-SNB) — product: USD/CHF & EUR/CHF Swiss National Bank 0.00% Rate Floor (USDCHF-SARON-SNB) refers to an institutional financial benchmark and regulatory standard monitored by Swiss National Bank (CH), currently calibrated at 0.00% (SNB Policy Rate) as of October 2026.
- **USD/JPY Carry Trade & SOFR-TONA Hedging Cost Spread** (USDJPY-CARRY-TONA) — product: USD/JPY Carry Trade & SOFR-TONA Hedging Cost Spread (USDJPY-CARRY-TONA) refers to an institutional financial benchmark and regulatory standard monitored by Bank of Japan / NY Fed (JP), currently calibrated at 3.82% (Annualized 3M Hedge Cost) as of October 2026.

## Analysis

In quantitative finance textbooks—from Brunnermeier, Nagel & Pedersen (*"Carry Trades and Currency Crashes"*) to BIS Working Papers analyzing [$9.60T OTC FX Markets](/index/bis-fx-swaps-96t-bis-9-6-trillion-global-otc-foreign-exchange-turnover-benchmark)—foreign exchange carry exhibits the most notorious **negative return skewness (Skew = -1.4 to -2.8!)** in liquid macro investing. For 18 consecutive months, a leveraged basket long **MXN, BRL, ZAR, and PLN** against short **[JPY](/index/usdjpy-carry-tona-usd-jpy-carry-trade-sofr-tona-hedging-cost-spread) and [CHF](/index/usdchf-saron-snb-usd-chf-eur-chf-swiss-national-bank-0-00-rate-floor)** grinds steadily higher ("up the stairs"); then, in **72 hours** (such as August 5, 2024 when MXN/JPY plunged **-14.2% in four sessions!**), forced VaR deleveraging vaporizes six quarters of accrued interest ("down the elevator shaft")!

Why does Gaussian Value-at-Risk (VaR) fail systematically on FX carry, and how do top-decile systematic macro CTAs and risk-parity desks regulated under [ECB](/regulators/european-central-bank-160), [BCBS](/regulators/bcbs-bis), and [FINMA](/regulators/finma) standards build **early-warning regime switches** that step aside *before* the elevator cable snaps?

## Anatomy of an Endogenous FX Carry Unwind Spiral

Carry crashes rarely require an emerging-market sovereign default! Instead, as Brunnermeier et al. proved, **funding liquidity shocks hit the FUNDING currency first**:
1. **Stage 1 (US Treasury Rate Volatility Spike — MOVE > 115)**: Surprising US payrolls, inflation, or hawkish Bank of Japan rate guidance spikes the **ICE BofA MOVE Index** (1M US Treasury swaption implied volatility).
2. **Stage 2 (Dealer VaR Contraction & Margin Calls)**: Prime brokers and multi-strategy risk managers cut gross notional limits across pods executing at venues like [Blocksize Capital](/brokers/blocksize-capital), [Bequant](/brokers/bequant), and [CMC Central](/brokers/cmc-central).
3. **Stage 3 (Synchronized Funding Short-Squeeze)**: Every leveraged carry book shares the *exact same short funding leg* (short JPY, short CHF)! Covering JPY shorts drives USD/JPY down -400 pips, spiking realized correlation across unrelated EM longs (**BRL, MXN, ZAR, HUF correlation jumps from 0.42 to 0.91 overnight!**).

![Four-stage endogenous feedback loop of an FX carry crash triggered by MOVE bond volatility and JPY/CHF funding short-covering](/api/figure?type=flow&title=Endogenous%20FX%20Carry%20Crash%20Spiral%20%28%22Down%20the%20Elevator%22%29&subtitle=Why%20US%20Treasury%20MOVE%20volatility%20%28%3E115%29%20triggers%20carry%20liquidations%2048-72%20hours%20ahead%20of%20equity%20VIX%20panics&steps=MOVE%20Breaks%20%3E115%3A%3AUS%20Treasury%20yield%20vol%20%2B%20BoJ%2FFed%20policy%20shock%20raises%20dealer%20repo%20VaR%7C25D%20Skew%20Steepens%3A%3A1M%20JPY%2FMXN%20%26%20JPY%2FBRL%2025-delta%20put%20skew%20jumps%20%2B1.8%20vols%20pre-crash%7CPod%20VaR%20De-Grossing%3A%3AMulti-manager%20risk%20desks%20slash%20FX%20carry%20leverage%20simultaneously%7CCorrelation%20-%3E%200.91%3A%3AShort-covering%20in%20JPY%20%26%20CHF%20forces%20indiscriminate%20MXN%2FBRL%2FZAR%20sales!&metrics=UNFILTERED%20SKEW%3A%3A-2.15%20Skew%3A%3ASevere%20Left-Tail%20Fatness%7CAUG%202024%20MXN%2FJPY%3A%3A-14.2%25%20%2F%204D%3A%3AClassic%20Funding%20Squeeze%7CMOVE%20DANGER%20ZONE%3A%3A%3E%20115%20-%20120%3A%3AFixed-Income%20Vol%20Lead%20Signal%7CFILTERED%20SHARPE%3A%3A0.48%20-%3E%200.94%3A%3ADoubles%20Risk-Adjusted%20Return!)

## Comparing Regime-Switching Signals: Equity VIX vs Bond MOVE vs FX Risk-Reversal Skew

Which indicator gives the earliest, cleanest regime-switch exit? Backtests across 2006–2026 prove that **relying on Equity VIX alone is too slow**—whereas combining **MOVE + Global FX Volatility (CVIX/JPMVXY) + 25-Delta Funding Skew** cuts maximum drawdowns by more than half:

| Regime Filter Indicator | Calm Carry Green-Light | Amber De-Leverage (50% Cut) | Red Cash / Long JPY Flip | Historical Lead Time |
| :--- | :--- | :--- | :--- | :--- |
| **ICE BofA MOVE Index** | **< 98 bps** (Stable UST curve) | **105 – 118 bps** (Rising rate vol) | **> 120 bps** OR +20bp 5-day jump | **2 to 5 trading days** ahead of VIX! |
| **CBOE Equity VIX** | < 17.5 (Low equity vol) | 19.0 – 24.5 | **> 25.0** OR VIX 1M/3M backwardation | Coincident (`T+0`, gaps overnight) |
| **1M 25D AUD/JPY & MXN/JPY RR** | Put skew > -2.2 vols | Put skew **-2.8 to -3.8 vols** | **Put skew < -4.2 vols** (Hedging panic) | **1 to 3 days** options warning |
| **CFTC TFF JPY Net Short** | < 50k contracts short | 80k – 120k contracts short | **> 140k net short** (>2.2 SD crowded!) | Structural fuel gauge (pre-condition) |

## Quantitative Crash-Filtered Dynamic Sizing & Cornish-Fisher VaR Formula

Because standard Gaussian VaR (1.645 * sigma) understates true 99% carry tail risk by **45% to 70%** when skewness is **-2.15** and excess kurtosis is **+7.4**, institutional quants size FX carry using **Cornish-Fisher Modified VaR** combined with a **Markov Regime-Switching Scaling Scalar (w_Regime_t)**:

```formula
z_CF(alpha) = z + (1/6)*(z^2 - 1)*Skew + (1/24)*(z^3 - 3*z)*Excess_Kurt - (1/36)*(2*z^3 - 5*z)*Skew^2
Modified_VaR_99 = - Mean_Return + z_CF(2.326) * Vol_Portfolio
w_Regime_t = clip( 1 - 0.5 * I(MOVE > 110) - 0.5 * I(VIX_1M > VIX_3M) - 0.5 * I(|RR_JPY| > 3.5), 0, 1 )

Worked Gaussian vs Cornish-Fisher Tail Risk Audit (Portfolio Vol = 8.0%/yr, Skew = -2.10, Excess Kurt = +7.0):
 1. Naive Gaussian 99% 10-Day VaR (z = 2.326): 2.326 * (8.0% * sqrt(10/252)) = 3.71% assumed max loss
 2. Cornish-Fisher z_CF = 2.326 + 0.754*(-2.10)*(-1)... [left tail z = -2.326]: |z_CF| = 4.18! -> True 99% 10D VaR = 6.67% (+80% larger!)
 3. Applying w_Regime_t scalar cuts Max Drawdown from -24.6% (Unfiltered Carry) down to -8.9% (Regime-Switched Carry)!
```

![20-year historical comparison of annualized Sharpe ratio and maximum drawdown across unfiltered FX carry, VIX-filtered, MOVE+Skew regime-switched, and collar-hedged carry](/api/figure?type=bars&title=20-Year%20FX%20Carry%20Performance%20by%20Tail-Risk%20Regime%20Filter&subtitle=Stepping%20to%200%25%20leverage%20when%20MOVE%20%3E115%20%26%20VIX%20term-structure%20inverts%20lifts%20Sharpe%20from%200.48x%20to%200.94x&steps=Unfiltered%20Carry%20%280.48x%29%3A%3A-24.6%25%20max%20drawdown%3B%20-2.15%20skew%20wipes%20out%202%20years%20in%201%20week%7CVIX%20%3C%2022%20Filter%20%280.67x%29%3A%3ACuts%20prolonged%20bear-market%20bleed%3B%20misses%20initial%20day-1%20gap%7CMOVE%20%2B%20Skew%20Gate%20%280.94x%29%3A%3AExits%20on%20bond%20vol%20%2B%2025D%20JPY%20call%20skew%20-%3E%20-8.9%25%20max%20drawdown!%7C3M%20Risk-Rev%20Overlay%20%280.86x%29%3A%3ASpends%20~95%20bps%2Fyr%20of%20carry%20to%20own%2015-delta%20JPY%20call%20%2F%20EM%20put&metrics=TAIL%20UNDERESTIMATE%3A%3A%2B80%25%20Larger%3A%3ACornish-Fisher%20vs%20Normal%20VaR%7CVIX%20TERM%20SIGNAL%3A%3AVIX1M%20%3E%20VIX3M%3A%3ASpot%20Backwardation%20%3D%20Exit!%7COPTIMAL%20HEDGE%3A%3A3M%2015D%20AUD%2FJPY%3A%3ACheapest%20Liquid%20Proxy%20Put%7CRE-ENTRY%20TRIGGER%3A%3AMOVE%20-12%20bps%3A%3AWait%20for%20Vol%20Peak%20Rollover)

## Cheapest Convex Tail Hedge: Owning 3-Month 15-Delta AUD/JPY Put Spreads!

Why don't institutional EM carry managers buy 15-delta USD/BRL or USD/ZAR calls directly to hedge crashes? Because **emerging-market upside USD call implied volatility is chronically overpriced** (often 18–22 vols!)! Instead, macro options desks finance **15-delta 3-month AUD/JPY or EUR/JPY puts** at **half the implied volatility cost**—exploiting the empirical law that in every genuine global carry liquidation, AUD/JPY drops 6%–10% lockstep with EM FX!

## Pairing Regime Filters With Diversified Funding Legs

Even better than timing 100% of crashes via options is eliminating single-currency funding concentration upfront—combining this regime-switching architecture with our [EM FX carry & REER hub guide](/wire/em-fx-carry-commodity-terms-of-trade-reer-guide-2026), [diversified JPY/CHF/EUR/CNH funding baskets](/wire/diversified-fx-funding-baskets-jpy-chf-eur-cnh-carry), and [LatAm BRL/MXN risk-parity sizing](/wire/usdbrl-usdmxn-real-carry-to-implied-volatility-ratios) across our [EM FX carry & terms-of-trade hub](/topics/em-carry-terms-of-trade-shocks)!

## Frequently asked questions

### Why did MXN/JPY crash over 14% in just four days during early August 2024?

Record crowded CFTC short-JPY / long-MXN carry positioning collided simultaneously with a hawkish July 31 Bank of Japan hike and a soft US payroll print, triggering cascading multi-manager VaR stop-outs.

### Why does the ICE BofA MOVE index often lead FX carry unwinds ahead of the equity VIX?

FX forward carry and cross-currency basis books are priced and financed directly off sovereign repo and OIS rate curves; spikes in US Treasury rate volatility immediately raise dealer balance-sheet haircuts before equities react.

### What does VIX term-structure backwardation (VIX 1M > VIX 3M) signal for FX carry?

Normally 3-month VIX trades above 1-month spot VIX (contango); when 1-month VIX jumps above 3-month VIX, institutional hedgers are paying panic premiums for immediate protection—a reliable systematic carry exit rule.

### When should systematic desks re-enter EM FX carry after a volatility liquidation?

Post-unwind carry Sharpe ratios historically peak (~1.4x+) once CFTC positioning flushes clean AND MOVE retreats at least 10-15 bps off its local peak alongside VIX returning to contango.


## Source URLs

- https://www.bis.org/publ/qtrpdf/r_qt2409a.htm
- https://www.nber.org/papers/w14473
- https://www.cftc.gov/MarketReports/CommitmentsofTraders/index.htm

## Citation

- APA: Henrik Lindqvist, CFA. (2026). FX Carry Crash Skewness: VIX & ICE BofA MOVE Regime-Switching Signals. FXMacro Wire. https://fxmacrowire.com/wire/fx-carry-crash-skewness-vix-move-regime-switching
- Canonical article: [FX Carry Crash Skewness: VIX & ICE BofA MOVE Regime-Switching Signals](https://fxmacrowire.com/wire/fx-carry-crash-skewness-vix-move-regime-switching)
- Site research index: [FXMacro Wire](https://fxmacrowire.com/llms-full.txt)
