1. 1 Department of Economics, Stockholm School of Economics, P.O. Box 6501, SE-113 83 Stockholm, Sweden, www.sse.edu Lars E.O. Svensson Web: larseosvensson.se The RBNZ and IJCB conference “Reflections on 25 years of Inflation Targeting” Wellington, New Zealand, December 1-2, 2014 Forward guidance

2. 2 Outline  Forward guidance as special or normal policy?  The Swedish experience of publishing a policy-rate path: Success and dramatic failure  The broader picture of Riksbank monetary policy  The New Zealand experience of publishing an interest-rate path: Less dramatic, monetary policy better focused  Conclusions

3. 3 Forward guidance: Special or normal?  Some info about future policy settings Policy rate, other policy instruments (balance sheet)  Recently: Fed, BoC, ECB, BoE Special: Binding Effective Lower Bound Implementing more expansionary monetary policy

4. 4 Forward guidance: Special or normal?  For many years: RBNZ, Norges Bank, Sveriges Riksbank, Czech NB Normal: Published forecast of interest rate (policy rate or 90-day path) RBNZ 1997-, NB 2005-, SR 2007-, CNB 2008-  Integral part of flexible inflation targeting Stabilize inflation around announced inflation target and resource utilization around long-run sustainable rate Forecast targeting: Set policy rate and policy-rate path such that forecasts for inflation and resource utilization “look good”

5. 5 Forward guidance in form of published policy rate path: Reasons for normal? 1.Transparency: Coherent CB forecast of target variables requires CB forecast of instrument. Then publish this. 2.Effectiveness: Management of expectations (of the future policy rate) 3.Informativeness: Central bank should have more info about its future policy settings. Should be useful info for private sector and other authorities 4.Justification: Provides a coherent way of justifying policy choice by comparison with policy alternatives 5.Accountability: Simplifies external evaluation of policy by comparison with policy alternatives and assessments of tradeoffs between target variables

6. 6 Developments of forward guidance as something normal Forecasts of target variables conditional on: 1.Constant policy rate If inflation forecast above (below) target, increase (reduce) policy rate Inconsistencies, not credible (Leitemo, Woodford) 2.Market expectations of policy rate: “market policy-rate path” CB forecasts of target variables may not “look good” CB policy-rate path may differ from market path 3.Central-bank forecast of policy rate: “CB policy-rate path” Consistent (inconsistent) if market path equal (differ) from CB path

7. 7 The Swedish experience of publishing a policy-rate path  Previously discussed by Svensson (2009, 2010) and Woodford (2012, 2013)  Success and failures  Evaluate according to (1)Predictability: Market forward rates predict policy-rate path (2)Credibility: Market forward rates in line with (adjust towards) policy-rate path (successful management of expectations)  Distinguish “actual” and “intended” monetary policy  A “credible “ policy-rate path may not necessarily be “appropriate”  Success: February 2009  Failure: September 2011

8. 8 Riksbank policy rate and Riksbank policy-rate paths Feb 2007-Sep 2014

9. 9 Riksbank policy rate and market policy-rate paths Feb 2007 – Sep 2014 A brief history Feb 2007 – Sep 2014: GoGo

10. 10 Success!

11. 11 Failure!

12. 12 The Swedish experience  What was behind the discrepancy between Riksbank and market policy-rate paths in September 2011?  What was the broader picture of Riksbank policy?

13. 13 Behind the discrepancy: Another discrepancy for foreign policy rates  September 2011: Riksbank forecast of foreign policy rates much above market expectations of foreign policy rate

14. 14 Behind the discrepancy: Yield curves 1.7 pp

15. 15 Behind the discrepancy: Another discrepancy for foreign policy rates  A lower Riksbank forecast for foreign policy rates would have implied a lower policy-rate path With unchanged policy-rate path, a forecast of stronger krona, the inflation forecast would have shifted down and the unemployment forecast would have shifted up  High forecast of foreign policy rates served to shift inflation forecast up and unemployment forecast down, supporting high policy-rate forecast  Resulted in Riksbank inflation forecasts biased upwards

16. 16 Riksbank CPI inflation forecasts 2011 and 2012 and outcome

17. 17 Actual yield curve implied much easier policy than intended  Market did not believe Riksbank policy-rate path  Market possibly understood that it would lead to too strong a krona, which would have forced the Riksbank to back down (this confirmed by informal discussions with market participants)  “Actual” MP much easier than “intended”  “Intended” MP would have implied 5-year bond rate 1.7 pp higher  Good for the economy that the RB policy-rate path was not credible

18. 18 The broader picture: CPI inflation 1995-2014 on average below target

19. 19 Average inflation expectations close to target 1997-2011: Average inflation below average inflation expectations To New ZealandNew Zealand

20. 20 Average unemployment 1997-2011 about 0.8 pp higher than if inflation had been on target Svensson (2015), “The Possible Unemployment Cost of Average Inflation Below a Credible Target,” American Economic Journal: Macroeconomics 2015(1). 0.8 pp

21. 21 Why the tightening in June/July 2010?  Leaning against the wind because of concerns about household debt and housing prices  Not clearly explained June/July 2010:  “Another factor [in relation to the policy-rate increase] is that household indebtedness has increased significantly in recent years” (June/July 2010 press release)  “[A]n interest rate increase was also a signal to avoid new financial imbalances from building up and that household indebtedness ought not to rise too much” (Governor Ingves, in June/July 2010 minutes)  Sending a signal to households that mortgage rates would soon return to normal  Tradeoff w/ inflation and unemployment not admitted until op-ed by Governor Ingves in October 2012

22. 22  Riksbank and Fed forecasts quite similar  Policies very different Fed: Keep policy rate between 0 and 0.25%, forward guidance, prepare QE2 Riksbank: Start raising the policy rate from 0.25 to 2% in July 2011  Riksbank: Ex ante clearly a premature tightening Ex ante: Riksbank and Fed forecasts June 2010 Source: Svensson, Lars E.O. (2011), “Practical Monetary Policy: Examples from Sweden and the United States,” Brookings Papers on Economic Activity, Fall 2011, 289-332. UnemploymentInflation

23. 23 The leaning: Policy rates in Sweden, UK, and US; Eonia rate in euro area Oct 28

24. 24 The leaning: Inflation in Sweden, euro area, UK, and US

25. 25 The leaning: Real policy rate in Sweden, UK, and US, real Eonia rate in euro area

26. 26 Ex post: Policy-rate increases from summer of 2010 led to inflation below target and higher unemployment (and probably a higher debt ratio) Cont. Source: Svensson (2013), “Unemployment and monetary policy – update for the year 2013,” Svensson (2013), “Leaning against the wind increase (not reduces) the household debt-to-GDP ratio”, posts on larseosvensson.se. LTV cap

27. 27 Unemployment in Sweden and New Zealand

28. 28 Riksbank’s case for leaning against the wind  Higher debt could imply (1) a higher probability of a future crisis, or (2) a deeper future crisis if it occurs  Hence, a tradeoff between (a) tighter policy now with lower debt but worse macro outcome now and (b) easier policy now with more debt but worse expected future macro outcome  Worse outcome now is an insurance premium worth paying  Is that true?  The answer can be found in the Riksbank’s own boxes in MPR of July 2013 and February 2014, plus Schularick and Taylor (2012) and Flodén (2014)  This involves putting numbers on the cost and benefit of leaning

29. 29 Cost of 1 pp higher policy rate: 0.5 pp higher unemployment rate Source: MPR July 2013, chapt. 2; Svensson, post on larseosvensson.se, March 31, 2014.

30. 30 Benefit (1) of 1 pp higher policy rate: Lower probability of a crisis  1 pp higher policy rate leads to 0.25 % lower real debt in 5 years  Lowers probability of crises by 0.25*0.4/5 = 0.02 pp  Assume 5 pp higher unemployment in crisis (Riksbank crisis scenario, MPR July 2013, box)  Benefit (1): Expected lower future unemployment: 0.0002*5 = 0.001 pp  Cost: Higher unemployment rate now: 0.5 pp  Schularick & Taylor (2012): 5 % lower real debt in 5 yrs implies 0.4 pp lower probability of crisis (average probability of crises about 4 %)  Riksbank, MPR Feb 2014, box: Source: Svensson, post on larseosvensson.se, March 31, 2014.

31. 31 Benefit (2) of 1 pp higher policy rate: Smaller increase in unemployment if crisis  1 pp higher policy rate leads to 0.44 pp lower debt ratio in 5 yrs  Smaller increase in unemployment in crisis: 0.44*0.02 = 0.009 pp  With probability of crisis as high as 10 %, divide by 10 (Shularick & Taylor: 4 %)  Benefit (2): Expected lower future unemployment: 0.0009 pp  Cost: Higher unemployment now: 0.5 pp  Flodén (2014): 1 pp lower debt ratio may imply 0.02 pp smaller increase in unemployment rate in crisis  Riksbank MPR Feb 2014, box: Source: Svensson, post on larseosvensson.se, March 31, 2014.

32. 32 Summarize cost and benefit of 1 pp higher policy rate  Riksbank’s case does not stand up to scrutiny Should have been > 1!

33. 33 The New Zealand experience of publishing an interest-rate path  Longest experience, starting in June 1997  Previously examined by Archer (2005), Moessner and Nelson (2008), Anderson and Hofman (2009), Detmers and Nautz (2012), Bergstrom and Karagedikli (2013)  Less dramatic than the Swedish experience  RBNZ monetary policy better focused

34. 34 RBNZ policy rate, 90-day rate and RBNZ 90-day paths Mar 1999 – Sep 2014

35. 35 RBNZ policy rate and market policy-rate paths Mar 2004 – Sep 2014

36. 36 RBNZ monetary policy: The broader picture

37. 37 RBNZ monetary policy: The broader picture Inflation deviation from target midpoint

38. 38 Average inflation expectations equal average inflation: No increase in average unemployment To SwedenSweden

39. 39 Unemployment in Sweden and New Zealand New Zealand: A brief history 2004Q1-2014Q3: GoGo

40. 40 Success: March 2005

41. 41 Success: March 2009

42. 42 Actual policy tighter than intended: Sep 2009

43. 43 Dec 2011: Actual policy easier than intended

44. 44 March 2012: Actual policy follows market

45. 45 Conclusions  Swedish experience much more dramatic, and special  Leaning against the wind: A high policy-rate path got priority over inflation and resource utilization  Case of “wag the dog”  New Zealand experience much less dramatic  Policy better focused on traditional goals  A few cases of actual policy tighter/easier than intended  Market sometimes ahead of RBNZ  Sometimes problems with transparency, effectiveness, informativeness, justification, and accountability  Keep publishing policy-rate paths

46. 46

47. 47 History: Sweden

48. 48

49. 49

50. 50

51. 51

52. 52

53. 53

54. 54

55. 55

56. 56

57. 57

58. 58

59. 59

60. 60

61. 61

62. 62

63. 63

64. 64

65. 65

66. 66

67. 67

68. 68

69. 69

70. 70

71. 71

72. 72

73. 73

74. 74

75. 75

76. 76

77. 77

78. 78

79. 79

80. 80

81. 81

82. 82

83. 83

84. 84

85. 85

86. 86

87. 87

88. 88

89. 89

90. 90

91. 91 Back

92. 92

93. 93 History: New Zealand 2004Q1- 2014Q3

94. 94

95. 95

96. 96

97. 97

98. 98

99. 99

100. 100

101. 101

102. 102

103. 103

104. 104

105. 105

106. 106

107. 107

108. 108

109. 109

110. 110

111. 111

112. 112

113. 113

114. 114

115. 115

116. 116

117. 117

118. 118

119. 119

120. 120

121. 121

122. 122

123. 123

124. 124

125. 125

126. 126

127. 127

128. 128

129. 129

130. 130

131. 131

132. 132

133. 133

134. 134

135. 135 Back