2016年-FCA英国金融行为监管局_occasional_paper_14_26页_1mb

Summary of Liquidity in the UK corporate bond market: evidence from trade data

Core Content

This paper presents an analysis of liquidity in the UK corporate bond market from 2008 to 2014, using transaction-level data. The study aims to contribute to the ongoing debate on liquidity in financial markets, particularly in the context of regulatory changes post-financial crisis. The authors argue that, despite a decline in dealer inventories, the overall liquidity of the market has remained stable or even improved in recent years.

Main Findings

• Liquidity has remained stable or improved:

  • The decline in dealer inventories does not correspond to a reduction in market liquidity.
  • Liquidity measures (such as Amihud, BPW, and IRC) show a consistent pattern: liquidity levels were low at the start of the sample, increased during the financial crisis, and then declined to very low levels by the end of 2011, remaining stable since.
  • There is no evidence that liquidity has become more volatile or "flighty" since the financial crisis, even though mild to moderate shocks (e.g., the "taper tantrum" in 2013 and the US treasury flash crash in 2014) occurred.

• Liquidity risk has not increased:

  • Measures of liquidity risk, such as the standard deviation of liquidity indicators, remained flat since 2011.
  • The market has not shown increased sensitivity to liquidity shocks, despite the economic environment being more uncertain.

• Liquidity component of bond spreads is stable or declining:

  • For investment-grade bonds, the liquidity component of yield spreads averages 0.01 basis points, which is not statistically different from zero.
  • For speculative-grade bonds, the liquidity component is 13.5 basis points, or 3.4% of the total spread.
  • The liquidity component has either declined or remained stable over the period from 2011 to 2014, suggesting that liquidity is not a growing factor in bond spreads.

• Liquidity could still decline under extreme stress:

  • While liquidity has generally improved, there is evidence that it can deteriorate significantly during periods of severe stress, such as the financial crisis.
  • The authors caution that the findings do not imply that liquidity is always available; it remains vulnerable under extreme conditions.

Key Points

• Data and Methodology:

  • The FCA has access to detailed transaction-level data for the UK corporate bond market.
  • The study uses a composite liquidity measure, which is a weighted average of several liquidity indicators (Amihud, BPW, IRC, and Amihud risk).
  • The authors also examine the impact of liquidity on bond spreads and find that the liquidity premium has not increased significantly over time.

• Market Structure:

  • The UK corporate bond market has remained largely unchanged in structure over the last eight years.
  • Around 90% of trades occur off-exchange, and electronic trading platforms have not significantly increased their share.
  • Asset managers and dealers are the main participants, with dealers acting as net sellers and asset managers as net buyers.

• Liquidity Measures:

  • Amihud Index: Measures the average price impact per trade. Lower values indicate higher liquidity.
  • Imputed Roundtrip Cost (IRC): A proxy for transaction costs, used to estimate the cost of liquidity.
  • BPW Measure: Captures the magnitude of price reversals, providing insight into market depth and resilience.
  • Turnover Ratio and Zero Trading Days: These measures are less correlated with other liquidity measures and are not major drivers of the composite liquidity component.

• Limitations and Considerations:

  • The study is based on FCA data, which may not capture all aspects of liquidity, such as the cost of immediacy.
  • The switch from Sabre II to Zen in 2011 may have influenced the results, particularly in terms of data coverage and completeness.
  • The findings are specific to the UK market and the period studied, and cannot be generalized to other markets or timeframes.

Conclusion

The study concludes that, despite a reduction in dealer inventories, liquidity in the UK corporate bond market has not declined in a meaningful way. The market appears to have become more liquid in recent years, and the liquidity premium has not increased. However, the authors emphasize that liquidity can still deteriorate under severe stress, and thus regulatory vigilance is still necessary. The paper contributes to the debate on liquidity in financial markets by providing empirical evidence based on detailed transaction data.

Key Figures and Tables

• Figure 1: Shows that most trades occur off-exchange, with electronic trading platforms gaining only a small share.
• Figure 2: Displays trade sizes, showing a trend towards larger trades, which is contrary to what would be expected if liquidity were declining.
• Figure 3: Illustrates the decline in dealer bond inventories from £400bn in mid-2008 to £250bn by the end of 2014.
• Figure 4 and 5: Present quarterly liquidity measures and the evolution of turnover and zero trading days.
• Table 1: Summarizes the composition of the sample in Sabre II and Zen datasets.
• Table 2: Shows the correlation between liquidity measures, highlighting that turnover and zero trading days are not closely aligned with other liquidity indicators.
• Table 3: Provides the composite liquidity measure by bond characteristics (age, maturity, rating), showing that younger and shorter-maturity bonds are more liquid, and that higher-rated bonds are more liquid than speculative ones.

Authors and Context

• Authors: Matteo Aquilina and Felix Suntheim, from the Chief Economist's Department of the FCA.
• Purpose: To contribute to the discussion on liquidity in financial markets using granular, transaction-level data.
• Research Context:
  • Corporate bond markets are vital for both companies and households.
  • Post-crisis regulatory changes, particularly in bank capital requirements, have been suggested as a cause of reduced liquidity.
  • The study aims to provide factual evidence rather than assess socially optimal liquidity levels.

Final Note

The authors stress that their findings are specific to the UK corporate bond market and the period 2008–2014. They also note that the FCA does not claim that liquidity has not been affected by regulatory interventions, but rather that it has not become less available in normal times or more volatile in response to mild shocks.