Hedging Discontinuous FX Risk: Evidence from the EUR CHF Floor Removal

Executive Insight

On 15 January 2015, the Swiss National Bank abandoned its three-and-a-half-year-old floor of 1.20 CHF per euro. The EUR/CHF exchange rate collapsed from 1.20 to below 0.90 within seconds—a 25% gap with no intervening tradable prices. Stop-loss orders placed at 1.15 executed between 0.90 and 1.15, depending on broker latency. Billions of dollars in currency-hedged positions were wiped out in what became the defining case study of discontinuous FX risk.

This paper provides a rigorous quantitative comparison of stop-loss orders versus option-based hedging for foreign-currency loan exposures. Using Bloomberg market data from 2012–2014 and the Garman–Kohlhagen pricing framework, we demonstrate that 1-year put options cost EUR 6,951 per EUR 100,000 of exposure—providing guaranteed protection—while a stop-loss strategy costs between EUR 4,696 (best case) and EUR 33,778 (worst case). The stop-loss’s apparent cheapness is an illusion: its cost is unbounded precisely when protection is needed most.

Stop-loss orders assume continuous price paths. The January 2015 EUR/CHF event proved conclusively that this assumption fails at the exact moments when hedging matters—during regime breaks, central bank interventions, and liquidity collapses.

The EUR/CHF Shock: Event Anatomy

The SNB established its 1.20 floor in September 2011, pledging to buy unlimited quantities of euros to defend it. For 3.5 years, EUR/CHF traded in an extraordinarily narrow band around 1.20, with implied volatility compressing to historic lows. On 15 January 2015, without warning, the SNB abandoned the floor. The resulting gap was one of the largest instantaneous moves in major-currency history.

The mechanics of the failure are instructive. Stop-loss orders are market orders triggered by a price level. They guarantee execution, but not execution price. When EUR/CHF gapped from 1.20 to below 0.90 with no intervening quotes, stop-loss orders at 1.15 could not execute at 1.15—they executed at whatever price was available. Many retail brokers reported fills at 1.00–1.05, well below the trigger level. Institutional desks fared little better: the ECN order book was empty between 1.20 and 0.92.

Put options, by contrast, pay max$(K - S_T, 0)$ regardless of the price path. A European put with strike 1.20 paid $1.20 - 0.98 = 0.22$ CHF per euro at settlement, exactly as contracted. The path taken to reach 0.98 is irrelevant to the option’s payoff.

Historical Precedent: GBP/DEM 1992

The study draws a direct parallel to Black Wednesday (16 September 1992), when the UK was forced out of the European Exchange Rate Mechanism. The GBP/DEM rate had been stable at 2.80–3.00 for two years, with a politically guaranteed floor at 2.773 (set at ERM entry in October 1990 at a rate of 2.95). On 16 September, the rate broke to ~2.70; by 17 September it was ~2.60; within weeks it reached 2.50.

The case study prices hedging instruments at historical data points. A 100,000 GBP foreign-currency loan denominated in DEM produces the following outcomes under three strategies:

In this case the stop-loss outperformed the put option—but only because the GBP/DEM decline was gradual (over days, not seconds), allowing the stop-loss to execute near its trigger level. The EUR/CHF event demonstrates what happens when the decline is instantaneous: the stop-loss cost explodes while the put option cost remains bounded.

The Garman–Kohlhagen Pricing Framework

Currency options are priced using the Garman–Kohlhagen model (1983), which extends Black–Scholes to accommodate two interest rates—domestic ($r$) and foreign ($r_f$):

where

$X(t)$ is the spot exchange rate, $K$ is the strike, $\sigma$ is the implied volatility, and $\Phi$ is the standard normal CDF. The model assumes continuous price paths—an assumption that the hedge comparison is designed to stress-test.

The Greeks provide additional insight: $\Delta_C = e^{-r_f T}\Phi(d_1)$ for the call delta, and $\text{Vega} = X(0)\,e^{-r_f T}\,\phi(d_1)\sqrt{T}$ for both calls and puts. The collapsing implied vol regime leading up to January 2015 made put options extraordinarily cheap, yet very few borrowers purchased them.

Bloomberg Market Data: 2012–2014

The study calibrates put option prices using Bloomberg terminal data for three annual snapshots, covering the period when EUR/CHF borrowers could have hedged:

The decline in implied volatility from 9.375% (2012) to 4.780% (2014) reflects the market’s growing belief that the SNB floor was permanent—a belief that made put options progressively cheaper and, simultaneously, the gap risk progressively more dangerous. By January 2014, a one-year put at strike 1.20 cost just CHF 0.015 per euro—almost free insurance against a risk that materialized 12 months later.

For longer horizons, the 31 January 2012 data provides: 10-year implied vol 13.31%, 10-year put price CHF 0.2365, 20-year put price CHF 0.3025. Corresponding swap rates: 10Y EUR 2.2395%, 10Y CHF 1.0735%, 20Y EUR 2.5835%, 20Y CHF 1.3710%.

Quantitative Hedging Comparison

The central result of the paper is a comprehensive cost comparison for hedging a EUR 100,000 foreign-currency loan at a spot rate of 1.204 (31 Jan 2012). Five hedging variants are evaluated:

Gap Risk and the Limits of Stop-Loss Orders

The fundamental problem with stop-loss orders is that they conflate two independent properties: the trigger condition (price reaches a level) and the execution guarantee (the order fills at or near that level). In continuous markets, these properties are approximately equivalent. In discontinuous markets, they diverge catastrophically.

The EUR/CHF event is not unique. The source study documents comparable discontinuities in FX markets, including:

• Samsung Heavy Industries (2006): $2 billion in losses from discontinuous KRW/USD moves during the Korean won appreciation
• Brazilian Airlines (2008): Over $5 billion in losses from BRL/USD options positions during the commodity currency collapse
• Metallgesellschaft (1993): $1.3 billion in hedging losses from maturity mismatches in oil contracts, exacerbated by liquidity-driven discontinuities

In every case, the losses were amplified by the assumption that hedging instruments would perform as designed under continuous-market conditions. The Garman–Kohlhagen model itself assumes continuity (via the Brownian motion driver), so its prices must be interpreted as lower bounds on the true cost of gap risk protection.

Institutional Implications

For Corporate Treasurers

• Hedge instrument selection: Stop-loss orders should never be the sole hedging instrument for foreign-currency exposures with multi-year horizons. They are appropriate only as a supplement to option-based protection in liquid, continuously traded currency pairs.
• Cost framing: Put option premiums are not “costs” to be minimized—they are insurance premiums with guaranteed payoffs. The correct comparison is not put premium vs. zero, but put premium vs. the expected loss from an unhedged or stop-loss-hedged position under stressed conditions.

For Risk Managers and Boards

• Gap risk disclosure: Any hedging policy that relies on stop-loss mechanisms must include explicit stress scenarios under discontinuous price dynamics. Boards should require scenario analyses showing the cost of stop-loss failure at 1σ, 2σ, and gap-event levels.
• Implied volatility monitoring: Falling implied volatility makes put options cheaper but simultaneously signals growing complacency—the exact conditions under which gap events are most likely. The cheapest time to buy protection is often the most dangerous time to go without it.

For Regulators

• Foreign-currency loan products sold to retail borrowers should require mandatory disclosure of gap risk and the limitations of stop-loss mechanisms. The EUR/CHF event demonstrated that retail borrowers were systematically sold FX loans with stop-loss “protection” that failed to protect.

Methodology & Academic Foundation

The analysis synthesizes market data, pricing theory, and institutional case evidence:

• FX option pricing: Garman & Kohlhagen (1983, Journal of International Money and Finance); the model is the standard for currency option valuation and derivatives desk pricing
• Market data: Bloomberg terminal data for EUR/CHF spot, implied volatilities (1Y and 10Y), LIBOR rates, and swap curves from January 2012 through January 2014
• Event analysis: SNB floor establishment (September 2011) and removal (15 January 2015); Bank of England ERM exit (16 September 1992, “Black Wednesday”)
• FX failure modes: Metallgesellschaft ($1.3B, 1993), Samsung Heavy ($2B, 2006), Brazilian Airlines ($5B+, 2008)—all involving hedging assumptions violated by market discontinuities
• Discontinuous dynamics: Merton (1976, jump-diffusion); Cont & Tankov (2004, Financial Modelling with Jump Processes); the Garman–Kohlhagen framework provides a continuous-path lower bound on true gap-risk pricing

All put option prices are computed from the Garman–Kohlhagen formula using observed market inputs. No Monte Carlo simulation or proprietary models are required—the analysis is fully reproducible from publicly available Bloomberg data.