2016年-CEPS欧洲政策研究中心_Did_quantitative_easing_affect_interest_rates_outside_the_US_New_evidence_based_on_interest_rate_differentials_33页_2mb

Summary of the Document: Did Quantitative Easing Affect Interest Rates Outside the US?

Core Content

This document investigates whether the Federal Reserve's quantitative easing (QE) programs, particularly QE1, had a distinct impact on interest rates outside the United States. The study uses a cointegrated vector autoregressive (CVAR) model and recursive estimation techniques to analyze the effects of QE on the transatlantic interest rate relationship between the US and the euro area, while considering the global downward trend in interest rates.

Main Points and Findings

1. Context of QE and Interest Rates

• The financial crisis of 2008 caused significant economic and financial market turmoil, leading central banks to adopt non-standard monetary policies.
• The Fed was the most active in implementing QE, with multiple rounds (QE1, QE2, QE3) and additional programs like Operation Twist.
• QE was intended to lower long-term yields, stimulate economic activity, and prevent deflation by signaling future monetary policy support and altering portfolio balances.

2. Global Interest Rate Trends

• Long-term interest rates in advanced economies have followed a global downward trend both before and during the financial crisis.
• The euro area, despite not implementing QE until recently, also experienced significant rate declines, suggesting that QE may not have had a unique or independent impact on interest rates in the US compared to other regions.

3. QE Transmission Channels

• Two main transmission channels are identified in the literature: the signalling channel and the portfolio-balance channel.
• The signalling channel suggests that QE signals future monetary easing, influencing market expectations.
• The portfolio-balance channel implies that QE alters the composition of financial portfolios, affecting interest rates.

4. Empirical Analysis Approach

• The authors use a cointegrated VAR (CVAR) model to estimate the long-run relationship between US and euro area interest rates.
• They incorporate nominal exchange rate data and test for structural breaks in the relationship around the time of QE implementation.
• The data spans from 2002 to 2014, and recursive estimation methods are applied to assess parameter constancy over time.

5. Key Results

• There is limited evidence that QE1 caused a break-up or destabilized the transatlantic interest rate relationship.
• The global downward trend in interest rates appears to have a more significant influence than QE itself.
• The impact of QE on interest rate differentials (USD vs. euro) is not significant, suggesting that QE may not have uniquely affected US interest rates.
• The exchange rate response to QE was also limited, which is inconsistent with the assumption that QE would have caused large depreciation of the USD.
• The persistence of monetary policy shocks is not well captured by event studies, which tend to focus on short-term effects.

Key Information

Quantitative Easing Programs in the US

• QE1 (2008–2010): Purchased $600 billion in government-sponsored enterprise (GSE) debt and mortgage-backed securities (MBS), and later added $300 billion in long-term treasury securities.
• QE2 (2010–2011): Purchased $600 billion in US treasuries.
• Operation Twist (2012): A maturity extension program involving $400 billion in long-term bonds and $200 billion in short-term bonds.
• QE3 (2012–2014): Targeted monthly purchases of $85 billion in MBS and long-term treasuries, with the continuation tied to labor market improvements.

Global Financial Market Integration

• Interest rates across advanced economies are highly correlated, both in downward trends and during cyclical fluctuations.
• The absence of a clear impact of QE on interest rates outside the US suggests that global factors play a more dominant role than national monetary policy.

Methodology and Data

• Data Period: 2002–2014
• Variables: US and euro area interest rates, exchange rates, and central bank balance sheets
• Models Used: Cointegrated VAR (CVAR), recursive estimation methods
• Key Tests: Unit root tests, Zivot-Andrews tests for structural breaks, residual analysis

Conclusion

• The study concludes that QE did not have a significant, independent impact on US interest rates relative to the euro area.
• The global downward trend in interest rates is more important than the specific QE programs in explaining the observed rate movements.
• The portfolio-balance and signalling channels may have played a role, but their effects were not distinct or identifiable in the transatlantic interest rate relationship.
• The long-term effects of QE on interest rates and exchange rates remain uncertain, and further research is needed to better understand the mechanisms of unconventional monetary policy.

Key Figures and Tables

• Figure 1: Federal Reserve Balance Sheet (2002–2014)

• Figure 2: CBOE Volatility Index (VIX) showing market uncertainty

• Figure 3: US: Central bank balance sheet, exchange rate, and inflation

• Figure 4: Euro area: Central bank balance sheet, exchange rate, and inflation

• Figure 5: Long-term interest rates in major currency areas since 1990

• Figure 6: Transatlantic long-term interest rate differential from 2007

• Figure 7: Nominal exchange rate

• Figure 8: Treasury bond yields - US and Germany

• Figure 9–14: Diagnostic tests and recursive estimates of the CVAR model

• Table 1: Impact of QE on long-term interest rates

• Table 2: Impact of QE on exchange rates

• Table 3: Counterfactual impact of QE on long-term interest rates

• Table 4–9: Unit root tests, residual analysis, and cointegration results

Outlook for Further Research

• The paper calls for more research on the international transmission of QE shocks.
• It highlights the need to account for global trends and structural changes in econometric models.
• The long-term effects of QE on economic variables such as interest rates, exchange rates, and inflation remain uncertain and require more detailed analysis.