2021-10-31-布鲁盖尔研究所-关于通胀_货币增长能告诉我们什么_（英）_21页_1mb

Summary of "Does Money Growth Tell Us Anything About Inflation?"

Core Content

This paper investigates the relationship between money growth and inflation, focusing on whether monetary aggregates are useful for inflation forecasting. It challenges the traditional quantity theory of money, which suggests that money growth directly affects inflation, by proposing a hypothesis that monetary aggregates only provide relevant information for inflation when monetary and inflationary conditions are unsettled.

Main Viewpoints

• Money's Decline in Relevance: Economists and central bankers, including the European Central Bank (ECB), have increasingly downplayed the role of monetary aggregates in inflation forecasting. The ECB's 2021 strategy review indicated a weakening link between money growth and inflation, and shifted focus to an integrated analytical framework.

• Public Interest vs. Expert Neglect: While the public (as reflected in Google search trends) continues to show interest in money and monetary aggregates, especially M1, economists and central bankers have moved away from relying on them for inflation analysis.

• Quantity Theory of Money: The theory, expressed as $MT = PQ$, assumes a constant velocity of money and exogenous real income, which makes money growth a determinant of inflation. However, this theory is not universally applicable and has been questioned for its limitations in modern economies.

• Pandemic Observations: During the pandemic, both the euro area and the US experienced spikes in money growth, but these were not sustained enough to indicate unsettled monetary or inflationary conditions. Thus, they did not lead to a stronger link between money and inflation.

• Empirical Testing: The paper tests the hypothesis that money matters only in unsettled conditions using out-of-sample forecasts for different time periods.

Key Findings

Settled Conditions (1999–2021)

• In the euro area and the US, where inflation has been relatively stable (around 1.9% for the ECB and 2% for the Fed), money growth does not significantly improve inflation forecasts.
• Forecasting Models:
  • Auto Regression and Random Walk models perform better than models incorporating money growth.
  • Fixed 1.9% inflation forecast is highly accurate, reflecting central bank credibility.
  • Adding M3 (euro area) or M2 (US) to inflation models increases forecast error, suggesting that money is not a reliable predictor in stable conditions.

Unsettled Conditions (Pre-Euro/Pre-Great Moderation Periods)

• In the pre-euro period (1975–1998) for the euro area and the pre-Great Moderation period (1975–1985) for the US, money growth did help in forecasting inflation.

• In the 1970s and 1980s in Italy, where monetary and inflationary conditions were highly volatile, money growth (M2) significantly improved inflation forecasts.

• Conclusion: Money growth is more relevant for inflation forecasting when monetary and inflationary conditions are volatile. It does not provide useful information in settled conditions, where inflation is stable and predictable.

Hypothesis and Testing

• The paper proposes that money only matters for inflation in unsettled conditions.
• This hypothesis is tested using out-of-sample forecasting models across different time periods.
• The results support the hypothesis: in periods of high volatility, money growth improves inflation forecasts, but in stable periods, it does not.

Implications

• Central Bank Credibility: The effectiveness of central banks in maintaining inflation targets (e.g., ECB's 1.9% and Fed's 2%) reduces the need for money-based forecasting.
• Pandemic Impact: The sharp increase in money growth during the pandemic was largely a result of government stimulus and liquidity support, not a sign of sustained inflationary pressure.
• Future Outlook: If the current pattern of high money growth is temporary, it may not lead to long-term inflation. However, if inflation becomes volatile again and money growth remains high, the link between money and inflation could re-emerge.

Conclusion

• The paper concludes that money growth is not a reliable indicator of inflation in stable monetary and inflationary environments.
• However, in periods of monetary and inflationary instability, money growth can provide useful information for forecasting inflation.
• The results align with an economic history approach, which considers path dependency rather than universal economic relationships.
• The current situation, while showing increased money growth, does not yet indicate unsettled conditions, and thus does not provide a strong signal for future inflation.