The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
EMBARGOED UNTIL FRIDAY, FEBRUARY 23, 2018 AT 10:15 A.M.; OR UPON DELIVERY Comments on “A Skeptical View of the Impact of the Fed’s Balance Sheet” Eric S. Rosengren President & CEO Federal Reserve Bank of Boston February 23, 2018 U.S. Monetary Policy Forum New York, New York bostonfed.org Overview The paper: ▶ Provides a nice overview of the literature on large scale asset purchases (LSAPs) ▶ Provides a careful historical description of the various LSAP programs ▶ Expands on the empirical literature ▶ Concludes that: ▶ Previous literature may overstate reliability and effectiveness of LSAPs ▶ LSAPs are not a full replacement for conventional policies 2 Why Should We Care, Now That the LSAP Program is Winding Down? ▶ How likely is it that we will need the LSAP Program in the future? ▶ My view is that it is quite likely that we will, because: ▶ U.S. is likely to have low real rates for some time, due to slow productivity and slow labor force growth ▶ The median long-run federal funds rate in the most recent SEP is only 2.8 percent ▶ Almost all recessions have resulted in the Fed lowering nominal rates by much more than 2.8 percentage points ▶ But if LSAPs are indeed not effective, then the Fed may need to take other measures ▶ Might entail altering the monetary policy framework, in a way that would be more likely to avoid short-term rates hitting zero 3 LSAP versus Conventional Policy Efficacy ▶ Central bank policymakers have a much better understanding of the impact of short-term interest rates, developed over 30 years ▶ Compared to LSAPs, policymakers have navigated many more episodes of short-rate tightening and easing ▶ Still, appropriate identification is non-trivial, and we still experience puzzles ▷ ▷ Proverbial “long and variable lags” Long rate “conundrums” – the expected response of long rates does not materialize ▶ For LSAP analysis, assessment is much more difficult – in part, due to very limited historical experience ▶ Basically just three observations – three LSAP programs during and post crisis ▶ Only one exit from LSAP program – and it has barely begun 4 The Paper’s Empirical Approach ▶ Identifies days where the 10-year Treasury moves more than one standard deviation and uses Reuters news reports to identify “Fed News” versus other events ▶ 1125 event days, 161 Fed event days, 348 economic data event days, 191 Europe event days ▶ Assumptions ▶ Fed events are accurately identified and non-Fed events are not due to Fed actions ▶ Events captured in one day – no lingering impact 5 Are the Paper’s Assumptions Actually Innocuous? ▶ Speeches by FOMC participants often referred to policy as “data dependent” ▶ Note: 348 economic data event days, 191 Europe days, but 161 Fed days – third party identification does not necessarily make it more accurate ▶ “Data dependence” implies meaningful economic data have implications for monetary policy ▶ Weak economic data in the study viewed as having no implications for monetary policy ▶ However, weak data likely to also imply more LSAPs and longer period before raising rates ▶ Similarly, Europe days have implications for both LSAPs in Europe and future strength in U.S., from international impact 6 10-Year Treasury Rate ▶ Potential confounding events? ▶ Fiscal policy ▷ Changes in government deficit expectations ▷ Government shutdowns, debt ceilings, Treasury debt management changes ▶ International: LSAP programs and interest rate changes in Japan and Europe likely influence U.S. Treasuries – global arbitrage in sovereigns ▶ Inflation surprises (because of focus on nominal rates) – what happened to real rates? 7 Alternative Focus on Enduring Impact ▶ I suggest more focus on term premia… potentially providing better measure of effects of LSAPs? ▶ Term premia in U.S. and countries that used LSAPs remain low by historical standards – could this, in part, be the enduring impact from LSAPs? ▶ Did LSAP programs lower the volatility of financial markets? ▶ Consider suggestive evidence – using other than event study techniques 8 Figure 1: Ten-Year Treasury Term Premium June 1961 - January 2018 6 Percent 4 2 0 LSAPs Begin and End 10-Year Term Premium -2 Jun-1961 Jun-1971 Jun-1981 Jun-1991 Jun-2001 Jun-2011 Recession Source: Federal Reserve Bank of New York, Adrian, Crump and Moench (ACM) Treasury Term Premia Estimates; NBER; Haver Analytics 9 Figure 2: Ten-Year Treasury Term Premium and Excess Reserves January 2000 - January 2018 3.5 Trillions of Dollars Percent 3.5 3.0 3.0 2.5 2.5 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0 -0.5 -1.0 Jan-2000 Excess Reserves 10-Year Term Premium Average 10-Year Term Premium Jan-2005 Jan-2010 -0.5 -1.0 Jan-2015 Recession Note: The average term premium is calculated for four different periods – when excess reserves are less than $500 billion, $500 billion to $1 trillion, $1-$2 trillion and $2 trillion or more. Source: Federal Reserve Bank of New York, Adrian, Crump and Moench (ACM) Treasury Term Premia Estimates; Federal Reserve Board; NBER; Haver Analytics 10 Figure 3: CBOE Market Volatility Index and Excess Reserves January 2000 - January 2018 2.8 Trillions of Dollars Index 60 Excess Reserves VIX Average VIX 2.1 45 1.4 30 0.7 15 0.0 Jan-2000 0 Jan-2005 Jan-2010 Jan-2015 Recession Note: The average VIX is calculated for four different periods – when excess reserves are less than $500 billion, $500 billion to $1 trillion, $1-$2 trillion and $2 trillion or more. Source: CBOE, WSJ, Federal Reserve Board, NBER, Haver Analytics 11 Overall Assessment ▶ It was a difficult empirical task to unravel the impact of limited programs with confounding effects ▶ With only 3 decisions, having so many events describing the decisions reduces the average effect, which is one reason why I have a more favorable assessment of LSAPs ▶ Might want to expand ways of determining enduring effects beyond event analysis ▶ Agree the evidence is consistent with some impact, of uncertain magnitude ▶ While not tested in this paper – agree that short-term rates are the better understood and tested way to conduct monetary policy 12 Implications for Next Recession, in My View ▶ Should avoid hitting effective lower bound with short-term rates ▶ Fiscal policy represents one alternative ▶ Difficult to depend on given political pressures and uncertainties ▶ Large deficits now may make future actions difficult ▶ Could alter the Fed’s monetary policy framework to reduce probability of hitting zero lower bound ▶ There are many possible alternative frameworks – I personally view inflation range with varying inflation target as promising 13