Communicating Monetary Policy Intentions - the Case of Norges Bank

1
Communicating Monetary Policy Intentions - the
Case of Norges Bank

• Amund Holmsen
• 16 May 2009
• (with Jan F. Qvigstad, Øistein Røisland
• and Kristin Solberg-Johansen)

2
Overview

1. What do we communicate?
2. Interest rate forecasts four years on a review
   of pros and cons
3. Challenges

3
Conclusion

• It seems to work well in Norway.

4
Talking about the future

• strong vigilance is therefore of the
  essence... (Trichet, August 2007)
• ...and anticipates that economic conditions are
  likely to warrant exceptionally low levels of the
  federal funds rate for an extended period."
  (FED, April 2009)
• ...the target overnight rate can be expected to
  remain at its current level until the end of the
  second quarter of 2010... (BoC, April 2009)

5
Changes in Norges Banks interest rate assumption

• 2001 - 2002 Constant interest rate
• 2003 - 2005 Markets interest rate expectations
• with comments
  
• 2005 Our own interest rate forecast

6
Ingredient 1 Baseline scenario (MPR 2/08)
Output gap
Key policy rate
CPI
CPI excl taxes and energy
7
Ingredient 2 Shift scenarios
Key policy rate
Output gap
Higher inflation
Higher inflation
CPI x taxes/energy
Higher inflation
8
Projected interest rate path
MPR 2/08
MPR 1/08
90 70 50 30
9
Ingredient 3 Delta accounting of the interest
rate path
10
Pros and cons revisited
11
Reviewing some counter-arguments

• Is conditionality misunderstood?
• Are policy makers constrained?
• Is it possible to decide on a whole path?

12
The yield curve moves on economic news
July 2008
Key policy rate
Norges Bank forecast
Norges Bank forecast in the previous report
Implied forward rates day before report
Implied forward rates after previous report
13
Reviewing some counter-arguments

• Is conditionality misunderstood?
• Are policy makers constrained?
• Is it possible to decide on a whole path?

14
Interest rate forecasts with fan chart from MPR
3/2008 Percent
15
Reviewing some counter-arguments

• Is conditionality misunderstood?
• Are policy makers constrained?
• Is it possible to decide on a whole path?

16
Reviewing some pro-arguments

• Is the reaction function better anchored?
• Test 1 Are market rates reasonably aligned with
  our forecast?
• Test 2 Are there smaller jumps in market rates
  around policy announcements?

17
b) June 2006
a) November 2005
Baseline scenario
Forward rates
Forward rates
Baseline scenario
c) June 2007
d) March 2008
Baseline scenario
Baseline scenario
Forward rates
Forward rates
18
Market rates as exogenous assumptions
Interest rate forecasts
Change in 12-month LIBOR krone rate from the day
of a policy announcement to the following day,
and averages for the two periods. Basis points.
19
Reviewing some more pro-arguments

• Fewer misunderstandings
• Easier to talk about the future
• Exit strategy as integrated part of the
  communication
• Credible interest rate forecast vs quantitative
  easing

20
Challenges

• Modelling optimal monetary policy
• Consistency
• Over time and accross states of the economy

21
Alternative approaches

• Simple interest rate rule
• rt art-1 (1-a)b1(Etptk -p)b2yt b3Dyt
• Optimal policy Minimizing a loss function
• L (p - p)2 ?y2 d(r - r-1)2

22
Simple rule

• rt art-1 (1-a)b1(Etptk -p)b2yt b3Dyt
• Two approaches
• Coefficients optimized over unconditional loss
• Coefficients optimized over conditional loss
• We faced the rules vs discretion issue and had
  to take a stand!

23
Forward looking Taylor rule vs Timeless
In 2006 we were able to reproduce our
forward-looking Taylor-rule forecast with optimal
policy under timeless perspective with the loss
function
Interest rate
Timeless
MPR 2/2006
Output gap
Inflation
Timeless
Timeless
MPR 2/2006
MPR 2/2006
24
Ramsey and Timeless (baseline scenario)
Key policy rate
Ramsey
Timeless
Output gap
Inflation
Timeless
Timeless
Ramsey
Ramsey
25
Timeless with different ?s
Key policy rate
?0.40
Timeless and Ramsey - ?0.30 - Weight change in
interest rate0.2
?0.20
Baseline scenario
Inflation
Output gap
?0.20
?0.20
Baseline scenario
?0.40
?0.40
Baseline scenario
26
Alternative approaches to commitment

• Current approach Forecasts, alternative
  scenarios and interest rate account
• Publish loss function (Svensson)
• Interest rate rule
• Target criterion (WoodfordGiannoni)
• (p - p) ?(y-y-1)0

27
Conclusions

• The experience is good
• The conditionality and the uncertainty in the
  forecast seem well understood
• Monetary policy appears to have become more
  predictable
• The policy discussion is brought closer to the
  research frontier
• Still early. If nothing else better economists

28
Communicating Monetary Policy Intentions - the
Case of Norges Bank

• Amund Holmsen
• 16 May 2009
• (with Jan F. Qvigstad, Øistein Røisland
• and Kristin Solberg-Johansen)