In recent years, the IMF has released a growing number of reports and other documents covering economic and financial developments and trends in member countries. Each report, prepared by a staff team after discussions with government officials, is published at the option of the member country.

Abstract

In recent years, the IMF has released a growing number of reports and other documents covering economic and financial developments and trends in member countries. Each report, prepared by a staff team after discussions with government officials, is published at the option of the member country.

Monetary Policy Communication and Forward Guidance25

A. Introduction

57. Forward guidance has taken a more important role during the Global Financial Crisis (GFC) and its aftermath. Already prior to the crisis, some central banks used explicit forward looking language as a device to increase transparency and strengthen the effectiveness of monetary policy. However, with the onslaught of the GFC, a growing number of central banks, including the U.S. Federal Reserve (the Fed), began to use forward guidance as a way to provide greater clarity about their policy intentions and reaction function.26

58. The Fed has used forward guidance to add stimulus and reduce uncertainty about future policy.27 Whether forward guidance can achieve these objectives depends on whether it is perceived as credible and whether it can enhance policy predictability through systematic and clear communication.28 In normal times, both credibility and predictability are helped by a well established pattern of past policy behavior that has proven successful in achieving the central bank’s stated policy objectives. However, as argued by Woodford (2012), in unusual circumstances, when policymakers have to break from past behavior, forward guidance becomes particularly challenging and requires more explicit explanations to be effective. Indeed, since December 2008, the Fed has taken several steps to clarify its goals and policy strategy, and to provide information about the expected path for policy rates (see Box 1).

59. Assessing the effectiveness of forward guidance requires distinguishing between shifts in the behavior of monetary policy and in the economic outlook. A forward guidance announcement associated with a more protracted policy rate path can be interpreted as a signal of either (i) a weaker economic outlook and/or lower inflation, or (ii) a more accommodative policy stance given current and projected economic conditions. Both would lead to lower expected interest rates. Similarly, lower policy rate uncertainty could reflect clearer and more systematic Fed communication or reduced uncertainty about the economic outlook.

The Fed’s Communication since 2008

Communication channels. In March 2011, the FOMC introduced regular post-meeting press conferences by the chairman. The press conferences coincide with the committee’s publication of the Summary of Economic Projections (SEP) and are intended to “further enhance the clarity and timeliness of the Federal Reserve’s monetary policy communication.”

Policy objectives and strategy. In January 2012, the FOMC published a statement on its longer-run goals and monetary policy strategy, formally committing to a 2 percent inflation target and assessment of maximum employment based on a range of indicators. The statement also clarified that the FOMC follows a balanced approach in making tradeoffs between the two objectives when necessary.

Forward guidance. In January 2012, the FOMC included federal funds rate projections by FOMC participants in the SEP (the so-called “dot” graph). Since December 2008, the FOMC has gone through four different forward guidance regimes:

  • Qualitative FG I: In December 2008, together with establishing a target range for the federal funds rate of 0 to 1/4 percent, the FOMC introduced qualitative policy rate guidance by indicating that economic conditions are likely to “warrant exceptionally low levels of the federal funds rate for some time.” In March 2009 the language was changed to “for an extended period.”

  • Date based FG: In August 2011, the FOMC shifted to date-based guidance by declaring that economic conditions are likely to “warrant exceptionally low levels of the federal funds rate at least through mid-2013”. In January 2012, the date was changed to late 2014, and in September 2012, it was changed to mid-2015.

  • State based FG: In December 2012, the FOMC transitioned to state-based guidance by stating that the committee deemed it appropriate for the federal funds rate to be at its zero lower bound at “least as” long as unemployment remains above 6.5 percent, projected inflation is no more than 2.5 percent and longer term inflation expectations are well anchored. This announcement also coincided with an extension of the long-term asset purchase program. In December 2013 the language was changed from “at least” to “well past.”

  • Qualitative FG II: Finally, in March 2014, the FOMC went back to qualitative guidance by removing the reference to specific unemployment and inflation thresholds. Instead the committee stated that it would be “appropriate to maintain the current target range for the federal funds rate for a considerable time after the asset purchase program ends, especially if projected inflation continues to run below the Committee’s 2 percent longer-run goal.” and that it “anticipates that, even after employment and inflation are near mandate-consistent levels, economic conditions may, for some time, warrant keeping the target federal funds rate below levels the Committee views as normal in the longer run.”

60. This selected issues paper addresses two main questions. First, did forward guidance represent a shift in the reaction function of the Fed or simply signal a weaker economic outlook? Second, did forward guidance move policy rate expectations as intended and reduce policy uncertainty? To complement existing studies, we focus on the impact of the different forward guidance regimes on policy uncertainty.29

B. Did Forward Guidance Represent a Change in the Fed’s Reaction Function?

61. Monetary policy before the crisis is well proxied by an estimated reaction function over inflation and unemployment. Given the Fed’s dual mandate of price stability and maximum employment a simple pre-crisis reaction function was estimated by regressing the federal funds rate (ff) on core PCE inflation (π) and the unemployment gap (ug).30

fft=0.86+1.82Πt+1.58utgR2=0.77(0.34)(0.13)(0.15)

Although the overall fit of the regression is quite high, there are two past periods where the implied policy path persistently differs from the actual (see Figure 1). The later episode, which occurred in the mid 2000s and where the predicted path was higher than the realized path, is particularly noteworthy as it coincided with the Fed’s first experiment with explicit forward guidance (i.e., August 2003 to December 2005).31 Figure 1 also shows the implied policy rate path over the medium term based on the latest FOMC projection (June 2014). The large deviation between the implied path (dashed blue) and that of the committee’s median projection (dashed red) suggests a marked break from pre-crisis behavior.

Figure 1.
Figure 1.

Forward Guidance Compared to Pre-crisis Policy Behavior

Citation: IMF Staff Country Reports 2014, 222; 10.5089/9781498331104.002.A003

Sources: FOMC and Staff Estimations

62. Starting from 2012, the FOMC members’ projected policy path deviates significantly from pre-crisis behavior, suggesting a shift in the Fed’s reaction function. In early 2012, the implied path of the federal funds rate based on pre-crisis behavior and the median of FOMC member’s projected policy rate path were fairly well in line with each other. 32 However, while the implied policy path shifted up during 2012, due to a slight improvement in the FOMC’s economic outlook, the median projected path was lowered and the lift-off date pushed further out (see Figure 2 and Table 1). Indeed, some FOMC members explicitly stated that the shift to date-based guidance did not constitute a more pessimistic view on the US economic outlook.33

Figure 2.
Figure 2.

A Shift to an Expansionary Stance Despite an Improving Economic Outlook, 2012

(September, 2012)

Citation: IMF Staff Country Reports 2014, 222; 10.5089/9781498331104.002.A003

Table 1.

FOMC’s Economic Outlook and Policy Projections for 2014

article image
Source: Selected Economic Projections, FOMC. The numbers correspond to the mid point of the central tendancy for each variable.

63. Alternatively, the shift in the FOMC member’s projected path could potentially be explained by unusual headwinds from the crisis. The shift to a more protracted policy rate path could also reflect a focus on broader measures of slack or a lower neutral real rate.34

  • Greater labor market slack than indicated by headline unemployment (or concerns that inflation was likely to run below target). The crisis has been associated with a greater decline in labor force participation than can be explained by demographic factors, as well as a rising number of part-time employed.35

  • A lower neutral real policy rate. Shifts in savings preferences, lower expected future growth, as well as global conditions could have contributed to a decline in the neutral real interest rate. For example, estimates by Laubach and Williams (2001) suggest a current real neutral federal funds rate that is close to zero.

64. Empirical evidence suggests that markets interpreted date-based forward guidance as a shift in the reaction function, and not as a deterioration in the economic outlook. Femia et al. (2013) use primary dealer survey data to show that date based forward guidance coincided with a perceived shift to a more accommodative monetary stance. Similarly, Raskin (2013) provides evidence of a shift in the Fed’s reaction function by showing that interest rate expectations became significantly less sensitive to macroeconomic surprises after the introduction of date-based forward guidance.36

65. Event studies have also found that forward guidance moved expectations in the intended direction. By defining a narrow window around FOMC announcements, event studies are designed to exclude other factors that might influence expectations such as news about the economic outlook. Campbell et al. (2012) and Femia et al. (2013) find significant announcement effects that lowered expected short term rates, as well as rates on longer term Treasuries and corporate bonds. Woodford (2012) shows that announcements had significant intra-day effects lowering interest rates and contributing to a flatter yield curve.

66. However, market responses also appear to have differed across forward guidance regimes. First, while the initial introduction of qualitative guidance in December 2008 and the date based guidance in August 2011 had sizable negative impacts on policy rate expectations, the shift to state based guidance in December 2012 had virtually no impact on policy rate expectations (see Figure 3). The latter result may reflect the fact that market participants had anticipated the shift, but it is also in line with the FOMC statement that made it clear that the FOMC viewed the new forward guidance as consistent with its earlier date-based guidance. Second, the shift back to qualitative guidance in March 2014 coincided with an increase in policy rate expectations at announcement. However, this shift may also reflect an upward revision in FOMC member’s projections for policy interest rates that were released at the same time.37

Figure 3.
Figure 3.

Interest Rate Impact on Forward Guidance Announcement Days 1/

(Daily change in the expected rate, percent)

Citation: IMF Staff Country Reports 2014, 222; 10.5089/9781498331104.002.A003

1/ Daily change in the interest rate of Eurodollar futures.Sources: Bloomberg and staff estimates.

C. Did Forward Guidance Reduce Policy Uncertainty?

67. Forward guidance has been associated with a decline in policy uncertainty. While most of the recent literature has focused market expectations about interest rates, less attention has been paid to the impact of forward guidance on policy rate uncertainty. Bauer (2012) finds that early announcements of forward guidance reduced uncertainty about future interest rates. Filardo and Hoffman (2014) show descriptive evidence that forward guidance has coincided with lower interest rate uncertainty at shorter horizons. Indeed, implied volatility of swaptions (a measure of interest rate uncertainty) has declined during the forward guidance period, with some increase in volatility at longer horizons from mid-2013 onwards (see Figure 4). The period of date based forward guidance in particular coincides with a sizeable decline in uncertainty. At the same time, the introduction of date based forward guidance resulted in a significant increase in market expectations about the time to lift-off from the ZLB (see Figure 5). While time to lift-off is also impacted by changes in the economic outlook, the upward shift at the announcement of date based guidance suggests that a significant part of the dampening impact of forward guidance on rate uncertainty worked through this channel. The period with state based forward guidance coincides with some increase in volatility at longer horizons and, as the recovery proceeded, a decline in the expected distance to lift-off. Finally, the return to qualitative forward guidance has been associated with an increase in policy uncertainty.

Figure 4.
Figure 4.

Implied Volatilities of Swaptions

Citation: IMF Staff Country Reports 2014, 222; 10.5089/9781498331104.002.A003

Sources: Bloomberg and staff estimates.
Figure 5.
Figure 5.

Time to Liftoff

(Years)

Citation: IMF Staff Country Reports 2014, 222; 10.5089/9781498331104.002.A003

Source: Bloomberg and staff estimates.

68. Uncertainty about the economic outlook has also fallen. Lower policy uncertainty could reflect less uncertainty about the near-term economic outlook. Forecaster disagreement about both inflation and the unemployment rate (as measured by the dispersion in private sector forecasts four quarters ahead) have also declined over time (see Figure 6). This is true in particular for the period after date based forward guidance was announced.

Figure 6.
Figure 6.

CPI and Unemployment Rate 1/

(Percent)

Citation: IMF Staff Country Reports 2014, 222; 10.5089/9781498331104.002.A003

1/ Measurement of dispersion, or the difference in forecasts between the 75th and 25th percentiles.Source: Philidelphia Federal Reserve.

69. Regression evidence. To futher analyze the impact of forward guidance on uncertainty, we regress measures of uncertainty on future interest rates on indicator variables for the forward guidance periods, controlling for economic uncertainty and other factors (including broader market uncertainty and risk aversion as proxied by the VIX index) (see the Appendix for a more detailed description of the data). Specifically, we estimate OLS regressions:

VOLt=α+βECONt+i=13λiFGi+Xt+ut

Where VOL is a measure of uncertainty about future interest rates (implied volatility of swaptions at different time horizons); ECON a measure of economic uncertainty (forecaster disagreement on unemployment and inflation) and FG are step dummy variables for the three forward guidance regimes (with the first qualitative forward guidance period as the omitted category). We also include control variables X (VIX, time to lift off from the ZLB and the OIS-Libor spread). The regressions were estimated using daily data over the sample period from 12/16/2008 through 06/16/2014.

70. Regressions evidence confirms that forward guidance was associated with lower policy uncertainty, with some differences across regimes. Several interesting results emerge:

  • Economic uncertainty increases uncertainty about future policy interest rates. As expected, msot coefficients for forecaster disagreement about the unemployment rate and inflation are positive and statistically significant (see Table 2). Broader market uncertainty (as measured by the VIX) is also associated with higher uncertainty about future policy interest rates.

  • Date-based forward guidance is associated with lower policy rate uncertainty on average. This effect is primarily driven by date-based forward guidance successfully pushing the lift-off date further out, and therefore providing greater clarity about the period when policy rates were expected to remain close to zero (see Figure 7).

  • State-based forward guidance may have reduced policy uncertainty above and beyond time to-lift-off. In particular, there is some evidence that once time to lift-off is controlled for, state-based forward guidance was associated with lower policy uncertainty. This may reflect the fact that at shorter horizons, state based forward guidance provided a more systematic approach to evaluating progress towards liftoff from ZLB. Indeed, this seems to be consistent with Bernanke’s (2013) argument that an important limitation to date based guidance was that it did not explain how future policy would be affected by changes in the economic outlook.

  • While based on a short sample, there is some evidence that the return to qualitative forward guidance was associated with lower uncertainty. Despite the increase in uncertainty at announcement, the regression evidence suggests that qualitatitive forward guidance has been associated with lower uncertainty. This may reflect additional forward guidance about the path of policy rates post lift-off.

Table 2.

Regressions Results

article image
Figure 7.
Figure 7.

Cumulative Effects of Foward Guidance on Interest Rate Volatility, Swaption 1y 1y

Citation: IMF Staff Country Reports 2014, 222; 10.5089/9781498331104.002.A003

1/ Interest rate volatility is measured as the change in volatility of Swaptions with a 1 year expiry and 1 year tenor.Sources: Bloomberg and staff estimates.

71. These regression results are subject to a number of caveats. The descriptive regressions do not identify causal effects and estimated impacts could reflect omitted factors. Furthermore, the analysis of volatility is complicated by the ZLB which tends to shrink volatility. However, the qualitative results for the forward guidance periods are robust across a number of alternative specifications, such as using alternative measures of uncertainty (e.g. realized volatility of eurodollar futures or the MOVE index of implied volatility in Treasuries), other variables to capture economic uncertainty (e.g. the Citi surprise index), and controlling for other factors (e.g. the impact of euro area sovereign stress using the spread between euro area core versus perphery bonds). To minimize complication from the ZLB the analysis refrains from comparisons with non-ZLB periods.

D. Conclusions

72. Uncertainties remain whether the deviation from pre-crisis policy is due to a change in the Fed’s reaction function or lingering headwinds from the crisis. Empirical evidence suggests that market participants have interpreted the protracted policy path as a shift in the Fed’s reaction function and not as response to a deteriorating outlook. While the Fed has mentioned a number of reasons for the protracted path, there have been calls for more clarity from both within the FOMC and from market participants. 38 Further communication about the FOMCs consensus view on the rationale for the protracted policy path could facilitate a smoother and more predictable normalization process.

73. Forward guidance has generally been effective in moving policy expectations in the intended direction and in reducing policy uncertainty. This is consistent with the Fed’s stated objectives of forward guidance policy. However, recent events have also resulted in challenges, in particular as the economic recovery has been moving the Fed closer towards a turning point in monetary policy.

74. Across regimes, there is some evidence that more recent forward guidance was been more effective in guiding expectations and reducing uncertainty than initial qualitative forward guidance. This experience suggests a potential trade off between systematic communication and policy flexibility. More recently, there is some evidence that the recent shift back to qualitative forward guidance was associated with lower uncertainty, perhaps reflecting additional guidance about the path of policy rates post lift-off. Looking forward, while qualitative forward guidance will provide the FOMC with more policy flexibility, it also suggests a greater premium on clear and systematic communcation to avoid an increase in policy uncertainty as lift-off approaches.

References

  • Balakrishnan, R., Dao, M., Sole J. and Zook, J (2014): “Recent U.S. Labor Force Participation Dynamics: Reversible or Not?”, Selected Issues Paper.

    • Search Google Scholar
    • Export Citation
  • Bauer, M. (2012): “Monetary Policy and Interest Rate Uncertainty”, Federal Reserve Bank of San Francisco Economic Letter.

  • Bernanke, B. (2013): “Communication and Monetary Policy,” speech at the National Economists Club Annual Dinner.

  • Bullard, J. (2013)U.S. Monetary Policy: Easier than You Think It Is”, remarks before Center for Global Economy and Business at New York University’s Stern School of Business

    • Search Google Scholar
    • Export Citation
  • Campbell, J., Evans, C., Fisher, J. and Justiniano, A. (2012): “Macroeconomic Effects of FOMC Forward Guidance”, Brookings Papers on Economic Activity 35(2).

    • Search Google Scholar
    • Export Citation
  • Dudley, W., (2012). “Regional and National Economic Conditions”, remarks before the Morris County Chamber of Commerce, Florham Park, NJ, September 18.

    • Search Google Scholar
    • Export Citation
  • Dudley, W., (2012). “Conducting Monetary Policy: Rules, Learning and Risk Management,” Remarks at the C. Peter McColough Series on International Economics, Council on Foreign Relations, New York City.

    • Search Google Scholar
    • Export Citation
  • Femia, K., Friedman, S. and Sack, B. (2013): “The Effects of Policy Guidance on Perceptions of the Fed’s Reaction Function”, Federal Reserve Bank of New York Staff Reports.

    • Search Google Scholar
    • Export Citation
  • Filardo, A. and Hoffman, B. (2014): “Forward Guidance at the Zero Lower Bound”, BIS Quarterly Review.

  • Laubach, T. and Williams, J. (2001): “Measuring the Natural Real Rate of Interest”, The Review of Economics and Statistics 85(4).

  • Levin, A. (2014): “The Design and Communication of Systematic Monetary Policy Strategies”.

  • Raskin, M. (2013): “The Effects of the Federal Reserve’s Date Based Forward Guidance”, Finance and Economics Discussion Series, Federal Reserve Board.

    • Search Google Scholar
    • Export Citation
  • Swanson, E. and Williams, J. (2013): “Measuring the Effect of the Zero Lower Bound on Medium and Longer Term Interest Rates”, Federal Reserve Bank of San Francisco working paper.

    • Search Google Scholar
    • Export Citation
  • Yellen (2014): Transcript press conference, March.

  • Woodford, M. (2012): “Methods of Policy Accommodation at the Interest-Rate Lower Bound”, Jackson Hole (September 16).

Appendix 1: Data Description

Implied volatility of swaptions. A swaption is an option that gives one party the right, but not the obligation, to swap a fixed rate for a floating rate (based on the 3 month Libor). The expiry of the option denotes the amount of time the party has to exercise the option, while the option tenor is the duration of the contract once it’s exercised. The expiry can be thought of as an approximation of the interest rate horizon. For instance, an option with a 1 year expiry, 1 year tenor, represents market sentiment about the short-term rate over a year, one year from now.39

The dispersion of forecasts. The dispersion of the Philadelphia Federal Reserves’ Survey of Professional Forecasters (SPF) forecasts is the difference between the 75th and 25th percentiles of private market participants four quarters ahead forecast of the CPI and the unemployment rate.

Time to lift-off. We construct a 36-month ahead curve from generic monthly Federal Funds futures contracts for each day and count the number of months until the interest rate exceeds 50 basis points.

VIX. Measure of the implied volatility of the S&P 500 index options over the next 30 days.

OIS-Libor Spread. Spread between the overnight indexed swap and the 3-month LIBOR rates, a commonly used measure of credit and liquidity risk.

25

Prepared By Tim Mahedy, Jarkko Turunen and Niklas Westelius.

26

The Fed also had some experience with forward guidance prior to the crisis. For instance in August 2003, the FOMC stated that it believed that policy accommodation could be maintained for a “considerable period”.

27

See Bernanke (2013).

28

Credibility requires consistency between forward guidance and the central bank’s policy objectives (e.g., price stability and maximum employment) as well as with its economic outlook. Predictability requires that the public understands how the central bank adjust its projected policy path as underlying economic data changes.

29

This paper does not tackle other important communication issues, such as how to communicate about financial stability considerations, the future of the Fed balance sheet or operational framework (see also S. Gray and D. King “The Operational Framework for Monetary Policy”, Selected Issues Paper, 2014).

30

The OLS regression was estimated using quarterly data over the sample period 1985:Q1 and 2008:Q3, and the unemployment gap is defined as the difference between the unemployment rate and CBO’s estimate of the NAIRU. The numbers in parentheses are the standard errors of the estimated coefficient, indicating that they are all statistically significant at the 5 percent level. Moreover, the estimated reaction function suggests that the Fed raise the real interest rate in response to inflationary pressure (consistent with the Taylor principle) and that the long-run federal funds rate is approximately 4.5 percent, given an inflation target of 2 percent.

31

The first period in the 1990s has been attributed by some observers to a temporary hike in the neutral rate due to strong productivity growth as well as by the Fed’s policy of opportunistic disinflation. The explicit forward guidance in August 2005 was in response to concerns that the policy rate would hit the ZLB.

32

The implied path is derived using the midpoint of unemployment and inflation from the SEP.

34

See also Levin (2014).

36

Swanson and Williams (2013) also find that the sensitivity of short rates to macro news declined during the qualitative and date based forward guidance periods, suggesting that they were effective in anchoring rate expectations.

37

More generally, interpretation of specific events is complicated by announcements that occurred at the same time. For example, the introduction of date based forward guidance coincided with an extension of the LSAP program.

38

See Bernanke (2012) for a list of reasons for the protracted path and Plosser (2014) and the NY Federal Reserve Survey of Primary Dealers for calls for more clarity.

39

For robustness, we also run regressions using realized volatility of eurodollar futures, as well as the MOVE index of implied volatility in Treasuries as the dependent variable and other control variables. Realized volatility of Eurodollar futures contracts: Eurodollar futures contracts are derivatives based on the interest rate paid on dollar denominated short-term deposits outside of the United States. The realized volatility is computed as the 10-day standard deviation of the day-over-day rate change based on the end of day option price. Merrill Lynch Option Volatility Expectations Index (MOVE Index): the index is a weighted index of implied volatilities on 1-month Treasury options at different maturities. The weights are: 20% on 2yr, 20% on 5yr, 40% on 10yr, and 20% on 30yr. Realized volatility and the MOVE index show similar trends. We also contolled for euro area sovereign stress using the spread between euro area core versus periphery long term (10 year) bonds.

United States: Selected Issues
Author: International Monetary Fund. Western Hemisphere Dept.