## _wp09230 - 1. Euro-Dollar Forex Swap

## Source details

**Canonical URL:** [_wp09230 - 1. Euro-Dollar Forex Swap](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09230.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09230.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09230.pdf.json)

---

### I. Introduction — purpose and motivation
- IMF tasked to deepen work on systemic risk, identification of systemically important institutions and markets, and development of early warning signals of distress.
- Market uncertainty and investor risk appetite play a key role in asset values and shock propagation (for example, liquidity spirals).
- Abrupt changes in market conditions often trigger financial crises even when underlying vulnerabilities persist.
- Common measures (for example, credit default swap spreads and some reduced-form models) often cannot separate idiosyncratic credit risk from an intrinsic premium attached to overall market distress.
- Identifying regime changes (tranquil → medium/high volatility, or vice versa) is useful for assessing timing of exits from government support.

### Key market variables and proxies used
- VIX: proxy for market uncertainty.
  - Defined: "The VIX, the Chicago Board Options Exchange volatility index, is a measure of the implied volatility of the S&P 500 index options over the next 30 days and is calculated from a weighted average of option prices."
- TED spread: proxy for stress in the interbank market.
  - Defined: "The TED spread is the difference between the three-month LIBOR and the three-month treasury bill rate."
- Euro-U.S. dollar forex swap: indicator of U.S. dollar funding pressures; noted spillovers from interbank market to FX swap market caused deviations from covered interest parity.

### Analytical approach and periods of focus
- Regime-switching models (SWARCH) using VIX, three-month TED spread, and euro-U.S. dollar forex swap to proxy global market conditions.
- Two focal periods:
  - Period 1: events and signals leading to the peak of the global crisis following the Lehman collapse on September 15, 2008.
    - Observations: rapid spillovers to emerging markets, sharply increasing uncertainty across asset markets, scramble for U.S. dollars, breakdown of carry trade, need for institutions to refinance U.S. dollar positions.
    - Models indicate a move towards a high volatility state before the Lehman episode.
  - Period 2: period following massive government intervention starting in many countries (post-mid-2009).
    - Market conditions and economic activity appeared improved since mid-2009; models show substantial improvement since spring 2009 but still moderate pressures on certain market conditions:
      - VIX and the forex swap both signal a high probability of a medium-volatility regime.
      - TED spread strongly signals a high probability of being in a low-volatility state.

### Organization
- Section II: overview of systemic risk. Subsequent sections (III, IV, etc.) provide qualitative discussion and SWARCH methodology.

---

### Overview of systemic risk (qualitative)
- Systemic risk resists a single formal definition; typically reflected by a “broad-based” breakdown with many institution failures (usually banks).
- Episodes characterized by intensity and breadth across markets and countries.
- Arises from institution-level fragilities and contagion across markets and countries; contagion channels can be additional to fundamentals.
- Recent crisis-era "runs" include:
  - runs on wholesale funding (inability to roll over funding for conduits/SIVs),
  - runs on interbank lending,
  - runs on banks’ market equity values.
- Temporary liquidity shocks can become defaults; high market uncertainty or low investor risk appetite/global liquidity tightness can amplify shocks.
- Systemic events linked to increased correlations across assets; observed correlations must control for autocorrelated/heteroskedastic volatility.
- Systemic events may involve regime shifts in correlations and higher moments (skewness, kurtosis).
- Policy actions can reduce or increase systemic risk; propagation may occur via direct financial links or indirect links (portfolio rebalancing, deleveraging, repricing).
- Conclusion: systemic risk best viewed as a collection of measures; detecting switches into “systemic gear” is critical for policymakers.

---

### Global market conditions — qualitative observations and historical markers
- Early August 2007: interbank markets clogged; evidence of a "run for quality."
  - Gold spot price: rose from $660 per ounce in early August 2007 to $1002 around the Bear Stearns rescue and Fed announcement on March 16, 2008; after that announcement the gold spot price dropped 10 percent in a short time.
  - Lehman bankruptcy saw gold soar over 20 percent within a few weeks.
- Demand for 10-year U.S. Treasuries surged; yields “almost halved” between crisis onset in August 2007 and the Bear Stearns and Lehman episodes.
- Funding and liquidity stress:
  - Libor-OIS spread used as proxy for funding liquidity and interbank stress; humps at crisis onset and end-year effects (December 2007).
  - Lehman collapse heightened counterparty and liquidity risk; central banks injected liquidity and substituted for interbank markets.
  - Shortage of high-quality collateral; haircuts increased across Treasury securities and risky assets.
  - Post-Lehman failure of counterparties to deliver U.S. Treasuries in repo transactions “drastically soared.”
- Volatility and cross-market effects:
  - Structural break in the VIX since the Lehman collapse; similar patterns in other implied equity-volatility indices.
  - Volatility spilled into foreign exchange; carry trades rapidly unwound at end-September 2008.
  - Implied volatilities of major EM currencies reflected carry-trade breakdowns; high-yielding currencies depreciated sharply vs. the U.S. dollar; funding currencies (e.g., Japanese yen) appreciated.
  - Scramble for U.S. dollars evident in higher volatility of euro-U.S. Dollar swap rates.
  - Covered interest rate parity (CIRP) violated during the crisis; daily deviations jumped at the Bear Stearns rescue and “completely broke down” for various EM currencies after Lehman.
- Emerging markets:
  - EM equity markets peaked in November 2007 and were initially less affected.
  - Deterioration in late 2008 after Lehman: flows to EM equity and debt mutual funds turned negative; total foreign assets in EM equity mutual funds peaked in November 2007; assets in EM debt mutual funds began to fall rapidly beginning in September 2008.
  - EM corporate spreads (over Treasuries) escalated sharply after Lehman; similar for high-yield corporate credit in U.S. and Europe.
  - EM countries with large current account deficits and heavy reliance on foreign wholesale funding were hardest hit; financial spillovers morphed into real-sector problems (industrial production and GDP plunging).
  - Large fiscal stimulus plans implemented beginning in late-2008.
- Earlier warning episodes:
  - February 2007: sharp, short-lived sell-off in EM equity, fixed income, and currency markets.
  - Late-February 2007: Shanghai stock market correction; ABX (BBB) index began to decline; carry trades in Brazil, South Africa, Turkey unwound.
- Spring 2009: global market conditions improved since spring of 2009, supported by central bank interventions, fiscal stimulus, and nascent pickup in activity; volatility subsided but remained above pre-crisis levels.

---

### Methodology: Markov-regime switching analysis (SWARCH)
- Objective: use Markov-regime switching techniques to examine financial stress formally.
- Model choice:
  - ARCH Markov-Switching model (SWARCH) of Hamilton and Susmel (1994) differentiates volatility states (low, medium, high).
  - Univariate SWARCH models with variables in first differences to address non-stationarity.
- Model structure (as presented):
  - Let t y be a vector of observed variables and t s denote an unobserved random variable with values 1, 2, ..., K governing the conditional distribution of t y.
  - Transition probabilities p_{ij} combine into a K×K transition matrix with Prob(s_t = j | s_{t-1} = i) = p_{ij} (Equation (1)).
  - Mean equation: AR(1) process (Equation (2)): t t t y y ε φ α + + = −1
  - Time-varying variance parameterization (Equation (3)) with state-dependent ARCH parameters and scaling factor t S g normalized to unity for the low volatility regime.
  - Error term υ_t ~ N(0,1) i.i.d.; state set {1,2,3}; dummy 1−d_{t−1} defined by sign of ε̃_{t−1}.
- Modeling note: ARCH specification estimated because GARCH(p,q) is not nested within SWARCH due to implicit infinite lag representation.

---

### 1. SWARCH Model based on Euro-Dollar Forex Swap

### Regime dynamics and timeline
- Euro-U.S. Dollar forex swap moved from low to medium volatility regime beginning of August 2007, then entered high volatility state right after the Lehman collapse in September 2008, remaining there until end-November 2008.
- Move into high volatility on September 15, 2008 coincides with sharp increase in counterparty risk from Lehman’s failure and a sizeable dollar shortage with margins and haircuts increasing.
- Non-U.S. banks, especially European ones, faced U.S. dollar funding shortages beginning summer 2007 and increasingly used FX swap and cross-currency swap markets (euro and sterling used as funding currencies).

### Findings (post-Lehman and 2009)
- U.S. dollar funding pressures declined dramatically since end-2008, largely due to facilities and swap arrangements between the Fed and several central banks.
- Since early-July 2009:
  - Probability of a low-volatility regime has increased to over 0.4.
  - Probability of a medium-volatility state, though declining, still remains at around 0.6.
- Last observation in the estimation is July 23, 2009.

---

### 2. SWARCH Model based on the VIX

### Historical episodes of high volatility
- VIX SWARCH shows highest probability of high volatility during:
  - Russian Crisis and LTCM default in 1998,
  - period surrounding WorldCom scandal and Brazil’s election in 2002,
  - beginning of the subprime crisis in fall 2007,
  - period following the Lehman collapse.
- Brief high-volatility entry at Shanghai crash and first ABX (BBB) price decline in late-February 2007.
- During the Bear Stearns rescue, VIX more likely in high than medium volatility.

### Findings (post-Lehman and 2009)
- Most recent peak probability of being in high-volatility reached end-September 2008 and lasted until mid-March 2009, at close to 1.
- Thereafter the probability of high-volatility declined rapidly to reach less than 0.05 by end-April 2009 and has remained at around the same low level since then.
- Since early-May 2009 the VIX has remained at a high probability (nearly 1) of a medium-volatility state.
- Probability of a low-volatility regime is nearly zero.

---

### 3. SWARCH Model based on TED Spreads

### Historical episodes and signals
- Three-month TED spread moved into a high volatility regime beginning of August 2007 and remained until the Bear Stearns rescue; the Lehman collapse again triggered high volatility.
- TED spread SWARCH also picks up Russia and LTCM default in 1998 and the September 11 shock.
- Shanghai correction and first ABX (BBB) price declines in late February 2007 appear as potential warning signals.

### Findings (post-Lehman and 2009)
- After peaking late in 2008, probability of being in a high-volatility state based on TED spreads declined dramatically since early-April 2009 and has remained low since then.
- Since late May 2009:
  - Probability of being in a low-volatility state has been higher than 0.9 and has remained there.
  - Probability of being in a medium-volatility regime is quite low (less than 0.1).
- Temporary spike in probability of a high-volatility state for TED spread observed in late-March 2009, likely related to Federal Reserve announcement to start directly buying U.S. Treasury bonds.

---

### 4. Synthesis: systemic risk signals and policy implications

### Aggregate conclusions from models
- Lehman failure was a watershed, but high volatility periods were present before Lehman (e.g., Shanghai crash and ABX (BBB) decline in February 2007, beginning of subprime in August 2007, Bear Stearns rescue in March 2008).
- Lehman aggravated an already observable crisis characterized by persistent signs of a high volatility state.
- High volatility states can be viewed as a potential manifestation of systemic risk.
- Global market indicators do not always remain in high-volatility state for long, with exceptions such as the TED spread or the VIX.

### Policy-relevant findings (mid-2009)
- Massive government interventions in late 2008 and early 2009 significantly eased interbank stress, though effects were smaller the more prolonged the crisis.
- By mid-2009:
  - TED spread strongly signals a high probability of being in a low-volatility state, suggesting interbank pressures had subsided dramatically since spring 2009.
  - VIX and forex swap market both signal a high probability of a medium-volatility regime, indicating continued moderate uncertainty.
- Mixed signals imply caution for policymakers considering exit strategies from financial-system support.

### Recommended use of the approach
- Regime-switching (SWARCH) models can help policymakers evaluate:
  - when market conditions make even a relatively small shock capable of triggering systemic events,
  - when financial institutions and markets can become distressed due to unstable market conditions,
  - when conditions have improved sufficiently to begin withdrawing government support.
- Models are most useful when combined with other tools to detect systemic crises and their abatement.

### Caveats and future research directions
- Methodology caveats:
  - Approach was applied univariately; multivariate SWARCH models combining various factors coherently and forward-lookingly are recommended for future research.
  - States are dataset-dependent; once high volatility periods exit the dataset, a “high volatility” state would signify less actual volatility compared to the crisis period.

*Source: IMF staff estimates and analysis in "SWARCH Model based on Euro-Dollar Forex Swap", figures and text as provided in the source content.*

### 1. Euro-Dollar Forex Swap ..............................................................................................

### _wp09230 - 1. Euro-Dollar Forex Swap

### I. INTRODUCTION — purpose and motivation
- The IMF has been called upon to deepen work on systemic risk, identification of systemically important institutions and markets, and development of early warning signals of distress.
- Market conditions—especially levels of market uncertainty and investors' risk appetite—play a key role in the value of financial institutions' assets and in the propagation of shocks (for example, liquidity spirals).
- Abrupt changes in market conditions often act as the trigger for financial crises even when underlying vulnerabilities persist for long periods.
- Common measures of financial fragility (for example, credit default swap spreads and some reduced-form models) often cannot separate idiosyncratic credit risk from an intrinsic premium attached to overall market distress.
- Identifying regime changes (tranquil → medium/high volatility, or vice versa) is useful for assessing when exits from government support might be considered.

### Key market variables and proxies used
- VIX: used as a proxy for market uncertainty.
  - Definition provided: "The VIX, the Chicago Board Options Exchange volatility index, is a measure of the implied volatility of the S&P 500 index options over the next 30 days and is calculated from a weighted average of option prices."
- TED spread: used as a measure for stress in the interbank market.
  - Definition provided: "The TED spread is the difference between the three-month LIBOR and the three-month treasury bill rate."
- Euro-U.S. dollar forex swap: used as an indicator of U.S. dollar funding pressures in international financial markets.
  - Note: spillovers from the interbank market to the foreign exchange swap market have led to periods when foreign exchange swap prices deviated from that implied by covered interest parity conditions.

### Analytical approach
- Regime switching models are adopted using the VIX, the three-month TED spread, and the euro-U.S. dollar forex swap to proxy key global market conditions.
- The analysis focuses qualitatively and formally on two periods:
  - Period 1: events and market signals leading to the peak of the global crisis following the Lehman collapse on September 15, 2008.
    - Observations: rapid spillovers to emerging market economies, sharply increasing uncertainty across asset markets, a scramble for U.S. dollars with the breakdown of the carry trade, and the need for financial institutions to refinance U.S. dollar positions.
    - Regime switching models indicate a move towards a high volatility state before the Lehman episode, consistent with elevated systemic risks.
  - Period 2: the period following massive government intervention starting in many countries.
    - Context: global market conditions and economic activity appeared to have improved since mid-2009, raising the question of when exit strategies can be contemplated.
    - Findings from the models: market conditions improved substantially since the spring 2009, but regime switching models still signal moderate pressures on certain market conditions.
      - The VIX and the forex swap both signal a high probability of a medium-volatility regime.
      - The TED spread strongly signals a high probability of being in a low-volatility state.

### Organization of the paper (as stated)
- The paper is organized with Section II providing an overview of systemic risk (further sections not included in the supplied excerpt).

*Source: _wp09230 - 1. Euro-Dollar Forex Swap*

### Section III discusses the role of global market conditions and systemic risk from a qualitative

### _wp09230 - Section III discusses the role of global market conditions and systemic risk from a qualitative

### Overview of systemic risk (Section II)
- Systemic risk resists a single formal definition and quantification; it typically reflects a “broad-based” breakdown in the financial system realized ex-post by a large number of failures of financial institutions (usually banks).
- Systemic episodes are often characterized as extremely acute financial crises whose severity is difficult to measure.
- Systemic risk is measured not only by intensity but also by the breadth of its reach across markets and countries.
- Systemic risk arises from both institution-level fragilities and contagion across markets and countries; contagion channels can be additional to market fundamental mechanisms present in non-crisis periods.
- Classic contagion example: bank panics from information asymmetries (deposit runs).
- Recent crises introduced other types of “runs,” including:
  - runs on wholesale funding (including inability to roll over funding for banks’ off-balance sheet special purpose vehicles),
  - runs on interbank lending,
  - runs on banks’ market equity values.
- Temporary liquidity shocks can metamorphose into default events that further affect liquidity; high market uncertainty (implicit volatility) or low investor risk appetite/global liquidity tightness can amplify small shocks into large effects.
- From risk managers’ perspective, systemic events are linked to increased correlations across assets (reducing hedging benefits); observed high correlations must control for autocorrelated/heteroskedastic volatility during crises.
- Systemic events may involve regime shifts: changes in correlations and in higher moments (skewness, kurtosis) of the data-generating process.
- Policy actions can alter systemic outcomes: some reduce systemic risk (guarantees, capital injections), while others (including lack of comprehensive crisis resolution) can increase systemic risk.
- Systemic risk sometimes denotes events causing loss of economic value or confidence across a substantial portion of the financial system with significant potential effects on the real economy; propagation can occur via direct financial links (parent-subsidiary, cross-border flows, trade finance) or indirect links (portfolio rebalancing, deleveraging, repricing of relative risks).
- Conclusion: systemic financial risk may be best viewed as a collection of measures; knowing when the system switches into “systemic gear” is critical for policymakers.

### Global market conditions and systemic risk: qualitative observations (Section III)
- Early August 2007: interbank markets across advanced economies became clogged, evidencing a “run for quality.”
  - Gold spot price rose from $660 per ounce in early August 2007 and reached $1002 around the Bear Stearns rescue by JP Morgan and the Fed’s announcement of the Primary Credit Dealer Facility on March 16, 2008; after that announcement the gold spot price dropped 10 percent in a short time.
  - The bankruptcy of Lehman Brothers saw the price of gold soar over 20 percent within a few weeks.
- Demand for 10-year U.S. Treasuries surged; yields “almost halved” between the onset of the crisis in August 2007 and the Bear Stearns and Lehman episodes.
- Funding and liquidity stress:
  - Libor-OIS spread served as a proxy for funding liquidity and interbank stress; experienced humps at crisis onset and end-year effects (December 2007).
  - Lehman collapse heightened counterparty and liquidity risk; many central banks injected liquidity and effectively substituted for interbank markets.
  - Shortage of high-quality collateral; haircuts increased across Treasury securities and risky assets.
  - Post-Lehman failure of counterparties to deliver U.S. Treasuries in repo transactions “drastically soared.”
- Volatility and cross-market effects:
  - Structural break in the VIX since the Lehman collapse; other implied equity-volatility indices showed similar patterns.
  - Volatility spilled into foreign exchange markets; carry trades rapidly unwound at end-September 2008.
  - Implied volatilities of major emerging market (EM) currencies reflected carry-trade breakdowns; high-yielding currencies depreciated sharply vs. the U.S. dollar; funding currencies (e.g., Japanese yen) appreciated from repatriation of funds.
  - Scramble for U.S. dollars evident in higher volatility of euro-U.S. Dollar swap rates.
  - Covered interest rate parity (CIRP) was violated during the crisis; daily deviations from CIRP jumped at the Bear Stearns rescue and “completely broke down” for various EM currencies after Lehman’s bankruptcy.
- Emerging markets (EM) dynamics:
  - EM equity markets peaked in November 2007 and were initially less affected than epicenter countries.
  - Persistency of the crisis, deterioration of advanced-economy fundamentals, and rising global risk aversion hit EM countries in late 2008 after Lehman.
  - Flows to EM equity and debt mutual funds turned negative: total foreign assets in EM equity mutual funds peaked in November 2007; assets in EM debt mutual funds began to fall rapidly beginning in September 2008.
  - EM corporate spreads (over Treasuries) escalated sharply after Lehman; similar escalation observed for high-yield corporate credit in the United States and Europe.
  - EM countries with large current account deficits and heavy reliance on foreign wholesale funding were most affected; initial financial spillovers morphed into real-sector problems (industrial production and GDP growth plunging).
  - Large fiscal stimulus plans were implemented beginning in late-2008 to counteract adverse real-sector impacts and provide liquidity and credit support to domestic banking systems.
- Earlier warning episodes:
  - February 2007: sharp sell-off in EM equity, fixed income, and currency markets (short-lived), demonstrating rapid, broad-based reappraisal of risk and flight to quality.
  - Late-February 2007: Shanghai stock market correction from unwinding of large long equity positions; ABX (BBB) index began to decline; carry trades in Brazil, South Africa, Turkey unwound; implied volatilities across fixed-income and equity markets rose; large declines in China, Malaysia, Philippines, Turkey.
  - Sovereign spreads across EM moved with market direction but were less affected than equity markets.
- Spring 2009 improvements:
  - Global market conditions improved since spring of 2009, bolstered by central bank interventions, fiscal stimulus packages, and a nascent pick-up in economic activity.
  - Volatility across asset classes subsided (but remained above pre-crisis levels); emerging markets benefited from improved global risk appetite.
- Analytical motivation:
  - The paper proceeds from this qualitative view to a more formal analysis examining regime changes in global market conditions to identify indicators useful for anticipating systemic events and for timing policy exits from crisis interventions.

### Methodology: Markov-regime switching analysis (Section IV introduction)
- Objective: use Markov-regime switching techniques to examine financial stress formally.
- Model choice rationale:
  - High-frequency financial data are intrinsically volatile during stress; the ARCH Markov-Switching model (SWARCH) by Hamilton and Susmel (1994) is chosen because it differentiates between volatility states (e.g., low, medium, high).
  - Univariate SWARCH models are adopted with variables in first differences to address non-stationarity.
- Model structure (preserve exact formulations and notation as presented):
  - Let t y be a vector of observed variables and t s denote an unobserved random variable with values 1, 2, ..., K governing the conditional distribution of t y.
  - Transition probabilities pij combine into a K×K transition matrix with Prob(s_t = j | s_{t-1} = i) = p_{ij} (as stated in Equation (1)).
  - Mean equation: AR(1) process (Equation (2)): t t t y y ε φ α + + = −1
  - Time-varying variance parameterization (Equation (3)) with state-dependent ARCH parameters and scaling factor t S g normalized to unity for the low volatility regime.
  - Error term υ_t ~ N(0,1) i.i.d.; state set {1,2,3}; dummy 1−d_{t−1} defined by sign of ε̃_{t−1}.
- Modeling note: an ARCH specification is estimated because GARCH(p,q) is not nested within the SWARCH framework due to its implicit infinite lag representation.

*Source: _wp09230 - Section III discusses the role of global market conditions and systemic risk from a qualitative*

### 1. SWARCH Model based on Euro-Dollar Forex Swap

### 1. SWARCH Model based on Euro-Dollar Forex Swap

### Euro-Dollar forex-swap regime dynamics and timeline
- The euro-U.S. Dollar Forex swap moves from a low to a medium volatility regime in the beginning of August 2007, then enters the high volatility state right after the Lehman collapse in September 2008, remaining there until the end of November 2008.
- The move into the high volatility state on September 15, 2008, coincides with a sharp increase in counterparty risk from Lehman’s failure and a sizeable dollar shortage with margins and haircuts increasing on most dollar-denominated assets.
- Many non-U.S. banks, especially European ones, faced a shortage of U.S. dollar funding for their conduits and SIVs beginning in the summer of 2007 and increasingly used FX swap and cross-currency swap markets (euro and sterling were used as funding currencies).

### Findings (post-Lehman and 2009)
- U.S. dollar funding pressures declined dramatically since end-2008, largely due to facilities and swap arrangements between the Fed and several central banks.
- Since early-July 2009:
  - The probability of a low-volatility regime has increased to over 0.4.
  - The probability of a medium-volatility state, though declining, still remains at around 0.6.
- Last observation in the estimation of the model is July 23, 2009.

---

### 2. SWARCH Model based on the VIX

### Historical episodes of high volatility
- The VIX SWARCH model shows highest probability of being in the high volatility state during:
  - the Russian Crisis and LTCM default in 1998,
  - the period surrounding the WorldCom scandal and Brazil’s election in 2002,
  - the beginning of the subprime crisis in the fall of 2007,
  - the period following the Lehman collapse.
- The model briefly enters high volatility at the time of:
  - the Shanghai stock market crash and the first abrupt ABX (BBB) price decline in late-February 2007.
- During the Bear Stearns rescue, the VIX was more likely to be in the high rather than medium volatility state.

### Findings (post-Lehman and 2009)
- The most recent peak probability of being in a high-volatility state was reached at end-September 2008 and lasted until mid-March 2009, at close to 1.
- Thereafter the probability of being in a high-volatility state declined rapidly to reach less than 0.05 by end-April 2009 and has remained at around the same low level since then.
- Despite the decline in high-volatility probability:
  - Since early-May 2009 the VIX has remained at a high probability (nearly 1) of a medium-volatility state.
  - The probability of a low-volatility regime is nearly zero.

---

### 3. SWARCH Model based on TED Spreads

### Historical episodes and signals
- The three-month TED spread moved decidedly into a high volatility regime in the beginning of August 2007 and remained in it until the Bear Stearns rescue; the Lehman collapse again triggered a high volatility regime.
- The TED spread SWARCH framework also picks up the Russia and LTCM default in 1998 and the September 11 shock.
- The Shanghai stock market correction and the first round of ABX (BBB) price declines in late February 2007 appear as potential warning signals.

### Findings (post-Lehman and 2009)
- After peaking late in 2008, the probability of being in a high-volatility state based on TED spreads declined dramatically since early-April 2009 and has remained low since then.
- Since late May 2009:
  - The probability of being in a low-volatility state has been higher than 0.9 and has remained there.
  - The probability of being in a medium-volatility regime is quite low (less than 0.1).
- A temporary spike in the probability of a high-volatility state for the TED spread was observed in late-March 2009, likely related to the Federal Reserve announcement to start directly buying U.S. Treasury bonds.

---

### 4. Synthesis: systemic risk signals and policy implications

### Aggregate conclusions from models
- The Lehman failure was a key watershed event, but high volatility periods were also present before Lehman’s failure (e.g., Shanghai crash, ABX (BBB) decline in February 2007, beginning of subprime in August 2007, Bear Stearns rescue in March 2008).
- The bankruptcy of Lehman Brothers aggravated an already observable crisis characterized by persistent signs of a high volatility state.
- High volatility states can be viewed as a potential manifestation of systemic risk.
- The global market indicators examined sometimes do not stay in the high-volatility state for long, with exceptions such as the TED spread or the VIX.

### Policy-relevant findings
- Massive government interventions in late 2008 and early 2009 had a significant impact on easing stress in the interbank market, though the effect was smaller the more prolonged the crisis became.
- By mid-2009:
  - TED spread strongly signals a high probability of being in a low-volatility state, suggesting interbank market pressures had subsided dramatically since spring 2009.
  - The VIX and the forex swap market both signal a high probability of a medium-volatility regime, indicating continued moderate uncertainty in financial markets.
- These mixed signals imply caution for policymakers considering exit strategies from massive financial-system support.

### Recommended use of the approach
- Regime-switching (SWARCH) models can help policymakers evaluate:
  - when market conditions make even a relatively small shock capable of triggering systemic events,
  - when financial institutions and markets can become distressed due to unstable market conditions,
  - when conditions have improved sufficiently to begin withdrawing government support.
- The models are most useful when combined with other tools to detect systemic crises and their abatement.

### Caveats and future research directions
- Methodology caveats:
  - The approach was applied univariately to global market conditions; multivariate SWARCH models that combine various factors coherently and forward-lookingly are recommended for future research.
  - States are determined by the existing dataset; once high volatility periods exit the dataset, a “high volatility” state would signify less actual volatility compared to the crisis period.

*Source: IMF staff estimates and analysis in "SWARCH Model based on Euro-Dollar Forex Swap", figures and text as provided in the source content.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09230.pdf_