Doubling Your Monetary Base and Surviving: Some International Experience

Richard G. Anderson , Charles S. Gascon, and Yang Liu

The authors examine the experience of selected central banks that have used large-scale balance- sheet expansion, frequently referred to as “quantitative easing,” as a monetary policy instrument. The case studies focus on central banks responding to the recent financial crisis and Nordic cen tral banks during the banking crises of the 1990s; others are provided for comparison purposes. The authors conclude that large-scale balance-sheet increases are a viable monetary policy tool pro vided the public believes the increase will be appropriately reversed. (JEL E40, E52, E58)

Federal Reserve Bank of St. Louis Review , November/December 2010, 92 (6), pp. 481-505.

he recent financial crisis has chal lenged rently sparse empirical evidence suggests that monetary policymakers around the quantitative easing actions likely must be large world on a scale that has not been seen because the private-sector’s substitution elastici - since the 1930s. In normal times, the ties among high-quality financial assets are small. Tmonetary policy for most central banks is imple - In this article, we examine the experience of mented by (i) targeting an overnight interest rate selected central banks that have used large-scale and (ii) holding as assets securities issued by balance-sheet expansion as a policy instrument. the country’s own national treasury. In some We conclude that such increases are a viable cases, a central bank’s assets also include for eign monetary policy tool for central banks with sig - exchange or other nations’ sovereign debt. When nificant independence and credibility, assuming large shocks occur and in response the policy the public believes the increase will be appropri - rate has already been reduced to (near) zero, ately reversed. some central banks have aggressively expanded To some analysts, large balance-sheet their balance sheet, a policy widely referred to increases raise the specter of higher inflation. as quantitative easing .1 In the United States, for Historically, an absence of fiscal discipline was example, the Federal Reserve’s mid-2010 bal ance the cause of large-scale increases in central bank sheet was approximately triple its size of two balance sheets. Sargent (1982), for example, years earlier. reviews cases of hyperinflation and Meltzer (2005) The essence of quantitative easing policies is reviews monetary policy in the United States dur - the purchase of assets from the private sector with newly created central bank deposits; such 2 exchanges promise to reduce both risk and term See Bernanke, Reinhart, and Sack (2004). Purchasing lower-qual ity assets raises discussion of the boundary between monetary and premia in longer-term interest rates. 2 The cur - fiscal policy. Recent academic papers include those by Jeanne and Svensson (2007), Cúrdia and Woodford (2010a,b), Gertler and Karadi (2009), Reis (2009), Borio and Disyatat (2009), and 1 See Bernanke and Reinhart (2004). Söderström and Westermark (2009).

Richard G. Anderson is an economist and vice president, Charles S. Gascon is a research support coordinator, and Yang Liu is a research associate at the Federal Reserve Bank of St. Louis. © 2010, The Federal Reserve Bank of St. Louis. The views expressed in this article are those of the author(s) and do not necessarily reflect the views of the Federal Reserve System, the Board of Governors, or the regional Federal Reserve Banks. Articles may be reprinted, reproduced, published, distributed, displayed, and transmitted in their entirety if copyright notice, author name(s), and full citation are included. Abstracts, synopses, and other derivative works may be made only with prior written permission of the Federal Reserve Bank of St. Louis.

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 481 Anderson, Gascon, Liu ing the late 1960s and 1970s. Recent actions in adjustment costs, although there may be interim the United States, United Kingdom, Switzerland, increases in economic activity. 5 Australia, and others have proactively used mas - A central bank’s promise to reverse a large- sive balance-sheet changes as a policy tool while scale balance-sheet increase in a timely fashion sustaining a commitment to avoid rapid infla tion. lacks credibility if the central bank is not suffi - ciently independent of the political process. Although earlier studies tended to be equivocal SOME MACROECONOMIC regarding a negative correlation between infla tion THEORY and central bank independence, more recent research using longer sample periods and broader Our principal lesson—that large, visible measures has found stronger correlations (Crowe money injections made in response to special and Meade, 2008). Central banks that have used events can increase near-term economic activity quantitative easing successfully rank high on without increasing inflation if policymakers credi - measures of independence, transparency, and bly commit to reverse the increase at a later date— accountability. Laurens, Arnone, and Segalotto arises in a variety of macro models. The key (2009), for example, ranked 98 central banks on element is that inflation expectations are little these characteristics—successful central banks affected by increases in central bank balance (except Australia) tended to rank at or above the sheets that are perceived as temporary. Goodfriend 15th percentile. The Sveriges Riksbank (Sweden) and King (1981) showed this result in the con text and the Swiss National Bank (SNB) are ranked of Barro’s (1976) rational expectations model by 4th and 5th, respectively. The Reserve Bank of introducing a central bank that credibly commits Australia (RBA), however, ranked 48th. to a long-run path for the money stock even while sharply increasing the near-term money supply. 3 Recently Berentsen and Waller (2009) showed the same result in a search-theoretic real business CASE STUDIES: cycle model. 4 In contrast, many early rational SUCCESSFUL LARGE-SCALE expectations macroeconomic models (during the BALANCE-SHEET INCREASES 1970s) specified that all changes in the money This section explores the practical use of supply were unanticipated and permanent—that large-scale balance-sheet increases as a policy is, the money stock followed a random walk. In such models, changes in the money stock, because instrument. Selected countries with recent large- they were anticipated to be permanent, caused scale central bank balance-sheet increases are 6 the price level to jump and real economic activ ity shown in Table 1. A subset of these countries is to remain unchanged. Similar results arise in the explored in greater detail. The countries loosely classical long-run equilibria of New Keynesian fall into three groups: (i) countries that responded models that contain incomplete information and in a temporary manner to the recent financial crisis, (ii) the Nordic countries during the bank ing

3 Specifically, Goodfriend and King (1981, p 382) outline a mecha - nism by which “for a given wealth, an individual who suffers an 5 See, for example, Woodford (2003) and Clarida, Galí, and Gertler anticipated temporary reduction in measured real balances might (1999). Among the differences in these papers noted by Berentsen shift expenditure from present to future periods, in order to take and Waller (2009, p. 2) is that New Keynesian models rely on “nom i- advantage of lower net costs of transactions in these periods.” nal rigidities, such as price or wage stickiness, that allows mone tary Presumably a sharp but temporary increase in money balances policy to have real effects” and that the models “are ‘cashless’ in provided by the central bank might induce individuals to shift the sense that there are no monetary trading frictions.” In their expenditure to present from future periods to take advantage of general equilibrium real business cycle model, all prices are flexi ble now-lower costs in the present period. but money overcomes trading frictions. Hence, in New Keynesian models, ad hoc stickiness may allow real effects of monetary shocks 4 The Berentsen-Waller model (2009) is based on the Lagos-Wright even under complete information. double coincidence of wants framework. Monetary policy is assumed to have short-run and long-run components, the former 6 The currency symbols used throughout the text are listed in Table 1. focused on stabilizing real activity (in the presence of shocks) and Unless otherwise indicated, monetary values are listed as U.S. the latter on the long-run inflation trend. dollars.

482 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu l t g , i f a n s e d e r s i o m t g r e e

r n e r i a g s o n k a u

r l a ) e a l n r d e e n

F c i a e e s )

w F r v t B

e M

h e , ; c S I n l t 9 9 9 9 1 7 3 7 4 7 7 5 5 0 8 6 0 0 2 0 2 1 8 ) h

(

h ...... , a e s

n m t t T r i a i 8 5 3 6 9 4 0 0 0 2 4 6 0 4 0 3 0 8 1 3 4 4 6

e d t i

u . i m s i 7 4 8 5 0 9 0 8 5 5 8 7 8 9 4 9 1 1 0 0 1 0 4 t s y i t n n m g

g s r r 7 1 4 3 1 2 1 1 5 1 2 2 2 2 1 l u e n , t o e d a u

F e s e 2 n C

u p R e

B j a c r

y u

e f r r r r b e o W

t h a o

e o u c .

t y f n

s o P r

k e y .

u r s a n n a b a e n t o t

i a o s c e e ( a y

a B y n l r

M d , b

a a o

c 0 l i n v v e y l a 2

r a s r m

o n

t e e a p o s s t p t e o a

h i a a J e n h t t y

t p b n

f a r Y

D K D P B e , W K D S R R P R R L e a o D n o N K K y s s Y e t K

y N r K Z B S Z R H r R E U R U L E E e u U U G U U t h P m e m k e S i

a i i t

C J K t A A E B C D E E I E E N P R S S C A G U r

n t n t t

e t

n g a e o I n n n n h n e n n n n n n n n n n n n n n n n n n n n w

n n B

a i T

u h m o o o e o o o o o o

o o o o o o o o o o o o o o , o r

t i i i d . i i i i l i i i i i i i i i i i i i i i i u o

h l l l l l l l l l l l l l l l l l l l l l l l u o k n t l l l e s c l l l l l l l l l l l l l l l l l l l l m t

q i i i i

i i i i i i i i i i i i i i i i i i i i a

d n

y r r r r

s r , a e ” b b b b t b b b b b b t t b b b b b b b b b b k e b s d s e

e b h

I i a h e e t

s s 7 6 3 9 4 3 9 5 4 9 5 6 7 5 8 3 8 3 1 0 7 9 9 t d

t f r i ...... s e a k l f e i i o a u o 1 3 6 4 4 5 2 1 2 3 2 6 1 3 2 9 0 8 9 7 9 8 0 e P

s h R b e r

2 7 6 5 1 4 9 1 6 7 3 1 7 1 1 3 5 0 1 1 7 0 5 r

k s r a s a c i 1 1 2 5 1 3 1 1 1 1 1 3 2 3 1 2 2 1 o i e l c n e , , l e n F

i h r a b n

5 2

t s i . m

u f u

s u

s y y l , o p e n l

a b s p t d

o a

e d a i t n n d n t o b a a a i M e

l u o

r c i d . t y u i e

l i t

8 8 e r f s . 1 8 9 9 5 6 9 8 ” t c i a e s 0 0 P s a I l s y 0 i 0 0 9 0 0 0 0

n t v “

0 0 0 n f u n

e 8 0 0 0 9 0 0 0 0 s g e e i 2 2 c 1 i l o 0 0 e

o n 0 2 2 2 1 2 2 2 2 e

n r

0 l t i g 1 1 t 7 9 h 8

o r r 0 o k r e r r r r r r r a a c 2 a n 0 0 T 9 0 0 e 2 t

e e l

i 9

n c e c m e e e e e e e

9 4 s . 2 2 m i 9 0 7 0 i

a n k u y a b b r 0

8 9

f s a i 0 9 b b b b b b b b i 1 2 0 2 e p

r c n d B 0 0 0 h y y

e e t l b 0 9 s v

y 0

a s m m r a s r r l 2 0 0 a r m t i m m m m m m m 2 1 h h h b n

2 k y l e a

i a b e e - a a u c 2 2 e

l e r e e e c e c e e e c

c t t o

a i e b r s e o ’ e e

t r r y r u u a v t c c c c c c c r o r r t y y s e c u e p p t b t a a a a n l l n n c

c r r n n e e o e e e e e

o b p p e e e P e e “ n d s

a u u u u a o

e i n , s J D S S D N D M J J J D M O D D M J D A M J F m

e n C h d ,

e o e t

r a , t r a t

i u

e d l d s a m e o i a

g t f o n e n

s t r b t e a S e

d a 7 r a a o h l i s F d 9 n t d

l e 9

e 0 a ( B e n 7 h 9 y s

s n t n 1 t

7 , r 0 i i e 9 1

9 a u n a s ; t 0

6 a 0 l y 2 e t t t n i 9 f 0 e t t t t b

o 2 r 0 y 0 0 p i r

r u 1 n n n o 0 U e

0 y r t e

n n n 2 0

2 c

i t t t o ,

a 2 r 0 n e

e e t t n r 6 m a h r 2 e e e e e e

L n r t t 2 s s s n n n

a y o i o

0 o e n n s s s s t

o u h d y . i f w r r e l

t e e a e e l e e e 0 n 0 t u t

i f a e e b e e e

) r e n n M r r r

e t s s s s r b 0 a S 2

e r r r s s u t e a n s e l d

d a l o

e e n J P P M P l l e e e b t 0 s

f r b l J t i P P P e e i i a l u n n s m i r r r u – – – – – o b 2 r r c s e o – e c – – –

a c a n r

e e

m m 8 e a 8 8 8 8 P m P P

r r o F P P 3 e 5 6 6 , i n n r i s o k e p c 0 c 0 0 9 0 – – – e o i s i O P s M – p – – i 9 0 0 0 c d o

e n t l

i 0 0 0 9 0 2 7 9 c n n e – – A e f 8 9 8 9 0 0 0 n r t a a – h p r 2 2 2 1 2 n 0 b 0 0 i o e 4 6 – u d t 0 0 0 1 a 2 2 2 i o

D P 5 i n B

i l p J

0 0 0

d 9 0 l 1

r r r r 0 r 0 0 r y r – – 0 D r r r r a y – e a e 2 2 2 r a 9 0 e w 0 d

e 2 2 2

e e e e e e i

r 3 7 0 –

c e f e e

t 3 a t v t 1 2 0 z c i h s o t y y y

b r b b b b r 9 r 0 2 6 r n

t ( 9 n e b i b b b 2

e i s e r

r r e I 9 0 ’ e e 0 y y T e

e e r i t k 9

p r a a a , s m m m m m d

n r r r k 1 2 w 0 s m l m m m 1 e s b h b b r

i e m

n o e e e e e a a n u u u S 2 l l n e e e c e u u s P

t o t t t t o o

i i a r r r R o d a , e i r u u

t c t t v

a c c c m m r r g y ,

l e p p p p p n

b n r b b b - a n l c c c f n n e e e o s s a p p d u s d e o e e e e e g B e e e t r o u a u a k

i e c d

d i F O S D D A J J A A S M D S J J S N O O F F S n e

t e i o a e e R h s a

d g m

d t t s v r

i w i e a u e l

l e e e F s S l a h n

e k b , t , r r e e

m e r u t

k n a g h e s h p n c a n

i n a e r

m t i a w s n

s - s e o w s C B I 8 )

a g

g u , إ n u

0 r

. n y o a i r d l R د 0 d a d c

l i g n e , 2 d e ,

o n

i r r r d n o j d a .

i e ) ) e p u D u a ) t t l F n s B a s a E a p a

a č m ) r a a i o l , a )

M a

s A e £ e ) a K I e N J b

$ M

( s d d

, r ,

i n ,

s € e u e i

m e Z m P e

)

s d

m t s c K l , n i h R h m s l e $ B

n a y N t n i n t o Z

f r

a a w a i s e , e i , a l a

y t G o r e S ) C i h g

, U ( m e

s c b D ( D r r s

r , s w n k

d c )

) E a e i i )

n ) S I )

d n k A ( Z ) i d a f n d $ y

d d ,

$ r e ) s n )

б f , a e e l / n U , s z

n N e i ) n r o a y o ) , d

k ( o у B s H d h e K ( € l s ) e ) , z r

r u

a (

k D c e s r i S , ) €

a e € р , r a a i d

£ K k ( r r C )

d u o

e , t

r t N r g ¥

n , R b K ( w ₩ , a , U u , n R a

e S o

l e l a b h e L D p n B y , l K o S E P , R a / l t g i R /

A B i y ( l r A

c ) c k n u R , c P U Y t

S i ( n r s ( S / s e B

n p o n ¥ (

a o I G

U o p

, ( n a E U U i a W c

e m c a l

d (

o N r u r

i ( N , t n E d i E L o m a l

d E ”

a R h t e / l m R h n ( o a ( e / e a n ( e r Y s t

r ( E e a

C s l y o

r e R c l

s f c o

t i s d b K

o ( a d l n n d v n r P ’ e / e b t h r

n d l o ( s B e y r o n e J e C ’ b e o e g u

n o o a a n ( n r t r T u t l p d ( r

F u n e o n

k k t t a / r g

d e

a n a p n o

l / e d r u

o m b n r a u r C d e S l i n u t a / C 1 d k

n f l d n e /

/ a e

k o k h / a e a , A n e u a m

: u a a e A K S e z o / o n l /

r t e e y h

/ i e i

r R a /

k i b f E n y e r h l w s t

m r /

r a y e T r p d e t e d w s / , d / t n

t / s d d d o / d / e a C e

a Z v : l e c

S u l k h a d t n n a a n

i i n u r

r i i e e e N n m s z n n E R h d . b c

t n e a t t t s a b n t e e e p a R R z k b c g a a T l i i i l u n i w i d e l l s s a

U i s r e s l e l g y l m a n e e p k e e e e n n n n o h z e e h a O r e u e r u o e i h w w a o u s g O i r c r a h h r I S “ t u t b n T S C N R w A A B B C C D F E I I J K T N P R S S U U U

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 483 Anderson, Gascon, Liu

Figure 1 Doubling of the Monetary Base in Selected Countries

United States United Kingdom 600 250 500 400 200 300 150 200 100 100 0 2005 2006 2007 2008 2009 2005 2006 2007 2008 2009

Switzerland Japan 200 350 180 250 160 140 150 120 50 100 2005 2006 2007 2008 2009 1999 2000 2001 2002 2003 2004 2005 2006 2007

Sweden (1990s) Sweden (2000s) 240 350 300 200 250 160 200 150 120 100 80 50 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2005 2006 2007 2008 2009

Finland Australia 220 200 180 180 160 140 140 120 100 100 60 80 1991 1992 1993 1994 1995 1996 1997 1998 1999 2005 2006 2007 2008 2009

Iceland New Zealand 500 450 400 350 300 250 200 100 150 80 50 2003 2004 2005 2006 2007 2008 2009 2005 2006 2007 2008 2009

NOTE: The figure displays 10 cases of extraordinary monetary base changes in nine countries. To illustrate clearly the magnitude of the change, in each panel the monetary base series is indexed (normalized) to 100 at the first observation. The horizontal (time) scale varies by country, reflecting primarily four different episodes. Changes in the United States, the United Kingdom, Switzerland, Sweden (2000s), Iceland, and Australia reflect the 2008 global financial crisis. Changes in Finland and Sweden during the 1990s reflect the Nordic banking crisis. Changes in Japan reflect its quantitative easing from 2001-06. Finally, New Zealand increased its monetary base permanently in 2006 to improve operation of its payment system. SOURCE: Federal Reserve Board, Bank of England, Swiss National Bank, Bank of Japan, Sveriges Riksbank, International Monetary Fund, Central Bank of Iceland, Reserve Bank of Australia, and Reserve Bank of New Zealand.

484 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu

Figure 2 Central Bank Policy Rates

United States United Kingdom 6

5 t t n n

e 4 e

3 c c r r e 2 e 1 P P –1 0 2005 2006 2007 2008 2009 2005 2006 2007 2008 2009 Switzerland Japan 3.5 Upper t

Lower t 0.5 n

2.5 n e e

c 0.3 c

r 1.5 r e

e 0.1 P 0.5 P –0.5 –0.1 2005 2006 2007 2008 2009 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Sweden (1990s) Sweden (2000s) 25 5 t 20 t 4 n n e 15 e 3 c c r 10 r 2 e e P 5 P 1 0 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2005 2006 2007 2008 2009

Finland Australia 12

t 7 t n n

e 8 e

c 5 c r r

e 4

e 3 P 0 P 1 1993 1994 1995 1996 1997 1998 1999 2005 2006 2007 2008 2009

Iceland New Zealand 20.0 t

t 8

n 15.0 n e e

c 6 c

r 10.0 r e

e 4

P 5.0 0.0 P 2 2003 2004 2005 2006 2007 2008 2009 2005 2006 2007 2008 2009

Euro Zone

t 4.0

n 3.0 e c

r 2.0 e

P 1.0 0.0 2005 2006 2007 2008 2009

NOTE: The figure displays central bank policy rates in nine countries and the euro zone. In some cases, dramatic decreases in policy target rates accompanied expansions of the monetary base. In others, changes were modest (e.g, Australia and New Zealand). In the 1990s, foreign exchange crises occasionally caused sharp increases in policy rates not accompanied closely by changes in the mone - tary base (e.g., Sweden and Finland). The Bank of Japan kept the uncollateralized overnight call rate as low as 0 percent during the quantitative easing period. Rates shown (daily data): United States, federal funds rate; United Kingdom, Bank Rate; Switzerland, target range of 3-month LIBOR rate; Japan, uncollateralized overnight call rate; Sweden, marginal rate (January 1990–May 1994), and repo rate (June 1994–present); Finland, tender rate (January 1993–December 1998), and minimum bid rate of the European Central Bank’s main financing operation (January 1999–present); Australia, interbank overnight cash rate; Iceland, nominal discount rate; New Zealand, Official Cash Rate; euro zone, minimum bid rate of the European Central Bank’s main financing operation. Sweden’s official interest rate topped 50 percent in September 1992 as the result of the Riksbank’s exchange rate defense. We omit observations of Sweden’s marginal rate that are higher than 24 percent. SOURCE: Federal Reserve Board, Central Bank of Iceland, Bank of England, Swiss National Bank, Bank of Finland, Bank of Japan, European Central Bank, Bank of Russia, Riksbank, Reserve Bank of Australia, Reserve Bank of New Zealand.

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 485 Anderson, Gascon, Liu

Figure 3 Actual and Expected Inflation

United States United Kingdom 5 4 t t 3 n n

e 2 e c c

1 r r e e 0 P

P –1 –3 –2 2005 2006 2007 2008 2009 2005 2006 2007 2008 2009

Switzerland Japan 2.5 2.0 t t 1.5 1.0 n n e e c c 0.5

r 0.0 r e e P

P –0.5 –1.0 –1.5 –2.0 2005 2006 2007 2008 2009 1999 2000 2001 2002 2003 2004 2005 2006 2007

Sweden (1990s) Sweden (2000s) 10 4 t t n n

6 e 2 e c c r r e e 2 0 P P –2 –2 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2005 2006 2007 2008 2009

Finland Australia 4.0 5.0 t

3.0 t

n 4.0 n e e c 2.0 c r 3.0 r e e

P 1.0 P 2.0 0.0 1.0 1992 1993 1994 1995 1996 1997 1998 1999 2005 2006 2007 2008 2009

Iceland New Zealand 16 4.5

t 12 t n n e 3.5 e

c 8 c r r e e P 4 2.5 P 0 1.5 2003 2004 2005 2006 2007 2008 2009 2005 2006 2007 2008 2009

Consumer Price Inflation: Year-over-year increase in consumer prices (monthly) Expected Inflation: Anticipated change in consumer prices during the next 12 months (monthly)

SOURCE: United States, Survey of Consumers conducted by Thomson Reuters and the University of Michigan; United Kingdom, Citigroup Inflation Tracker conducted by YouGov and Citigroup; Australia, Melbourne Institute Survey of Consumer Inflationary Expectations conducted by the Melbourne Institute of Applied Economic and Social Research; New Zealand, Survey of Expectations conducted by the Reserve Bank of New Zealand; Sweden, Consumer Tendency Survey conducted by the National Institute of Economic Research; Finland, Consumer Survey conducted by Statistics Finland; Japan, Consumer Confidence Survey conducted by the Economic and Social Research Institute. Values of expected inflation in Japan are calculated using the median of the anticipated change in con - sumer prices during the next 12 months. Switzerland, Consumer Confidence Survey by the State Secretariat for Economic Affairs. Switzerland’s Consumer Confidence Survey reports an index of expected inflation. Data are transformed into an index with a mean equal to the mean consumer price inflation series over the observed period.

486 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu

INFLATION EXPECTATIONS Well-anchored inflation expectations are crucial to the success of unconventional monetary policy actions, including large increases in central bank balance sheets. Figure 3 shows both con - sumer price inflation and expected inflation for selected countries that have experienced such increases. Actual inflation is measured as the year-over-year increase in consumer prices. Expected inflation is the anticipated change in consumer prices during the next 12 months as determined from household surveys. We selected household surveys based on their availability across coun - tries. Other surveys are available. A cross-country comparison of inflation expectations is rou tinely included in the IMF’s World Economic Outlook . The Reserve Bank of New Zealand’s Survey of Expectations is a market-based survey, whereas all other surveys are household based. A com pari - son of household surveys and market surveys can be found in Batchelor and Dua (1989). The onset of financial crisis and recession typically reduces both actual and expected infla - tion. At the same time, if expansionary monetary policies are anticipated to stimulate economic activity, households might expect that actual inflation will return to its long-run trend in the near future. This pattern is evident in most of the countries we surveyed. During 2008, for example, the United States, the United Kingdom, Switzerland, Sweden, and Australia experienced sharp decreases in actual and expected inflation. At the end of 2008, expected inflation generally stabilized (albeit at a lower rate than the recent trend) even as actual inflation continued to fall. As economic activity stabilized during 2009, inflation expectations increased (particularly in the second quarter), even while actual inflation continued to ease. Higher expectations were somewhat validated by higher inflation during the second half of 2009. It appears that both actual and expected inflation had returned to the long-run trends by the end of 2009. The 1990s Nordic banking crisis is another example. Inflation in Sweden and Finland was high in 1990 and 1991. As Swedish and Finnish central banks and governments pursued aggressive expansionary policy, consumer price inflation declined below these central banks’ inflation tar - gets. The National Institute of Economic Research in Sweden and Statistics Finland started to sur - vey inflation expectations in 1993 and 1995, respectively. In both countries, expected inflation was stable around the respective central bank’s inflation target.

crisis of the 1990s, and (iii) selected other coun - in 2006 when the Bank of Japan (BOJ) reversed tries, for comparison, with (apparently) perma nent policy. New Zealand, in a cooperative agreement increases. Figure 1 shows changes in the mone tary with its banks to improve operation of the pay - base of these countries during periods of quan ti - ment system, doubled its monetary base to a new, tative easing (each country’s series is normalized sustained level. to 100 at the date when major balance-sheet Large-scale balance-sheet increases in expansion began). 7 Sweden’s monetary base response to financial crises often occur after dur ing the Nordic banking crisis of the 1990s, pol icymakers have reduced their target interest for example, rapidly doubled and remained at rate to (near) zero; the paths of central bank pol icy that level for two years before slowly returning rates are shown in Figure 2. During December to its pre-expansion level. Japan’s monetary base 2008, for example, the BOJ reduced its target over - increased slowly starting in 2001 and fell rap idly night interest rate to 0.1 percent and the Federal Reserve reduced its target federal funds rate to a 7 Generally, the monetary base is defined as the sum of currency in range between zero and 0.25 percent. Exceptions circulation outside the central bank plus deposits of financial are Australia and Iceland, which are discussed institutions at the central bank. Variations for individual coun tries are noted in the case studies. later.

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 487 Anderson, Gascon, Liu

The inflation experience in these countries 2009 up to $1.25 trillion of agency MBS and up is shown in Figure 3, which displays both actual to $200 billion of agency debt, plus up to $300 inflation and survey-based measures of expected billion in longer-term Treasury securities during inflation. Consistent with the visibility of the the next six months. These purchases, later financial crisis and high credibility levels of these referred to as the Large-Scale Asset Purchase central banks, inflation expectations moved little, program, sustained the size of the Fed balance if at all, as balance sheets increased—indeed, the sheet even as various credit and lending programs time series for actual and anticipated inflation closed. As of the April 2010 FOMC meeting, total are nearly indistinguishable. assets were $2.34 trillion.

The United States8 United Kingdom Before September 2008, the size of the Federal Rapid expansion of the U.K.’s monetary base Reserve’s balance sheet had changed little during began in February 2009, eventually tripling to the financial crisis of 2007-08 because holdings of £208.04 billion in July 2009 from £68.69 billion Treasury securities decreased as lending through in January 2009. Motivating aggressive increases credit-market programs increased (Figure 4). In in the monetary base was a sharp slowing in eco- September, the Fed ceased shrinking its Treasury nomic activity: Real output during 2008:Q4 and portfolio and large-scale balance-sheet increases 2009:Q1 fell at 7 percent and 10 percent annual began. In turn, the federal funds rate slipped rates, respectively. The Bank of England (BOE)’s steadily; on December 16, 2008, the Federal Monetary Policy Committee (MPC) had reduced Open Market Committee (FOMC) set a target its policy rate (Bank Rate) from 5 percent in range for the federal funds rate of 0 to 0.025 per- October 2008 to 1 percent in February 2009.10 Yet, cent. On November 25, 2008, the Federal Reserve forecasts suggested an increased risk that inflation announced that it would purchase up to $100 billion of debt issued by the Federal Home Loan might undershoot the MPC’s 2 percent target. Banks, FNMA, and FHLMC, plus up to $500 bil- In March, the MPC decided to ease monetary lion of mortgage-backed securities (MBS) backed conditions in the United Kingdom by reducing by FNMA, FHLMC, and GNMA with the stated Bank Rate to 0.5 percent and to begin aggressive purpose to “reduce the cost and increase the avail- expansion of its balance sheet. The Bank’s first ability of credit for the purchase of houses.”9 purchase was £75 billion of government bonds Purchases began in January 2009. (gilts) during the first week of March. During As of late January 2009, the Federal Reserve’s March, the Bank purchased £982 million of com- total assets and liabilities were approximately mercial paper, £128 million of commercial bonds, $2 trillion versus $900 billion in late August and £12.9 billion of gilts. The monetary base rose 2008; purchases of housing-related debt and MBS to £90.12 billion by the end of March.11 By the accounted algebraically for about one-third of end of May, additional purchases pushed the the increase and a variety of credit and lending BOE’s total assets close to £300 billion and the programs accounted for the rest. monetary base to £156.14 billion. At its March 17-18, 2009, meeting the FOMC Figure 5 shows the impact of these programs announced its intent to purchase by year-end on the BOE’s balance sheet and the monetary base.

8 See Bernanke (2009) for a summary of the U.S. experience. 10 Before March 5, 2008, the U.K. Asset Purchase Facility (APF) purchased £986 million in commercial paper. Because these purchases 9 GNMA (Government National Mortgage Association, or “Ginnie were financed by the sale to the public of Treasury bills, and hence Mae”), part of the U.S. Department of Housing and Urban Develop - had little impact on the monetary base, we omit them from our ment, issues no debt but does issue MBS on which it guarantees analysis. For details of the APF’s operation, see BOE (2009a,b). payment of principal and interest. FNMA (Federal National Mortgage Association, or “Fannie Mae”) and FHLMC (Federal 11 In 2006 the BOE discontinued publication of its monetary base, Home Loan Mortgage Corporation, or “Freddie Mac”) issue both referred to as M0. Here, we calculate the monetary base as the debt and MBS on which they guarantee the timely payment of sum of BOE banknotes in circulation plus deposits of banks at the principal and interest. BOE (reserves).

488 NOVEMBER/DECEMBER 2010 FEDERALRESERVEBANKOFST. LOUIS REVIEW Anderson, Gascon, Liu

Figure 4 Composition of Federal Reserve Balance Sheet

$ Billions Assets 3,000 Short-Term Lending to Financial Firms and Markets Rescue Operations 2,500 Operations Focused on Longer-Term Credit Conditions Traditional Portfolio

2,000

1,500

1,000

500

0 Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr 07 07 07 07 08 08 08 08 09 09 09 09 10 10

$ Billions Liabilities 3,000 Source Base Treasury Financing Account 2,500 Traditional Liabilities and Capital Account

2,000

1,500

1,000

500

0 Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr 07 07 07 07 08 08 08 08 09 09 09 09 10 10

NOTE: See Federal Reserve Bank of St. Louis, U.S. Financial Data, for a description of chart categories. SOURCE: Federal Reserve Board, Federal Reserve Statistical Release H.4.1.

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 489 Anderson, Gascon, Liu

Figure 5 Composition of Bank of England Balance Sheet

£ Billions Assets 350 Other Assests Long-Term Reverse Repo 300 Bonds and Securities Purchased Advance to HM Government 250 Short-Term Market Operations

200

150

100

0 May Aug Nov Feb May Aug Nov Feb May Aug Nov Feb May Aug Nov Feb 06 06 06 07 07 07 07 08 08 08 08 09 09 09 09 10

£ Billions Liabilities 350 Other Liabilities 300 Short-Term Market Operations Reserves Notes in Circulation 250

200

150

100

0 May Aug Nov Feb May Aug Nov Feb May Aug Nov Feb May Aug Nov Feb 06 06 06 07 07 07 07 08 08 08 08 09 09 09 09 10

SOURCE: Bank of England ( www.bankofengland.co.uk/markets/balancesheet/ ).

490 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu

On the asset side of the balance sheet, the large Reserve in October 2008, eventually reaching increase in the “Long-Term Reverse Repo” cate - more than CHF 60 billion in March 2009. In gory reflects the assets acquired by expanding December 2008 the SNB created a loan stabiliza - the types of collateral that may be pledged on tion fund to “finance the acquisition of illiquid traditional lending facilities. The later increase assets from UBS, largely composed of assets in the “Other Assets” category reflects the BOE’s backed by US residential and commercial mort - increase in dollar lending and assets purchased gages.” Under terms of the loan, UBS, among under the Asset Purchase Facility. On the liabili - other financial institutions, will make partial ties side of the BOE’s balance sheet, the mone tary payments extending up to 12 years. (Assets in base is measured as the sum of “Notes in Circula - this program were CHF 22 billion as of September tion” plus “Reserves” at the BOE. Note that the 2009.) Between October 2008 and April 2009, BOE issues liabilities (BOE bills) that are not part the Swiss monetary base approximately tripled, of the monetary base as we have measured it in reaching CHF 117 billion in April 2009. On this analysis. During the fall of 2008, the BOE March 12, 2009, the SNB announced a policy issued 1-week maturity bills to finance expanded shift toward foreign exchange market interven - lending to banks. The BOE shifted away from tion, saying the appreciation of the Swiss franc using its own bills for financing during 2009, “represents an inappropriate tightening of mone - causing liabilities in the “Short-Term Market tary conditions.” Operations” category to decline and bank reserves The SNB’s unconventional policies sharply to increase. increased the size of its balance sheet and the The BOE reports that balance-sheet actions Swiss monetary base. Figure 6 shows the growth reduced yields on medium- and long-dated gov - and changing composition of the SNB assets and ernment bonds, as well as spreads of commercial liabilities. The largest increase among assets is in paper and commercial bonds over overnight index foreign currency swap transactions. On the lia bili - swaps. Yields on 10-year U.K. bonds fell when ties side, the largest increase is in bank deposits purchase programs were announced and subse - at the SNB (top category on the graph)—from an quently drifted upward as purchases occurred. average of CHF 6 billion in 2007 to more than Inflation expectations plummeted from a high of CHF 75 billion in March 2009. To temper the 4.5 percent in September 2008 to a low of 1 per - increase in the monetary base and “absorb liq uid - cent (see the dashed line in Figure 3) before later ity in the market” resulting from unconventional drifting and leveling off near 2 percent. monetary policies, the SNB began issuing its own debt in October 2008 (labeled as “SNB Debt Switzerlan d Certificates” in Figure 6). As of September 2009, The impact of the global financial crisis on the SNB had CHF 25 billion outstanding in SNB Switzerland has been modest, albeit sufficient to notes. These notes have a maximum maturity of result in recession. During 2008:Q4 and 2009:Q1, one month. real gross domestic product (GDP) declined at an Although we cannot yet assess the impact, if annual rate of 2.5 percent and 3.5 percent, respec - any, of the monetary base expansion on infla tion, tively. The Swiss National Bank’s (SNB) 2009 it seems reasonable to explore the SNB’s infla tion Financial Stability Report describes the steps they forecasts underlying its policy actions. In March took. 12 In October 2008 the SNB reduced its tar get 2009, the SNB forecast deflation for most of 2009 for the 3-month Swiss franc London Interbank and close to zero inflation in 2010 and 2011. The offering rate (LIBOR) and began expanding its forecast six months later projects deflation only balance sheet. The SNB also began participating in early 2009 and 2 percent inflation by the end in foreign currency swaps with the Federal of 2011. Perhaps tripling the monetary base has forestalled further undesired decreases in infla -

12 tion (or even deflation): Inflation expectations For additional details, see Swiss National Bank (2009); all quota - tions in this section are from this report. plummeted in mid-2008, reaching close to zero

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 491 Anderson, Gascon, Liu

Figure 6 Composition of the Swiss National Bank Balance Sheet

CNF Billions Assets 300 Swap Transactions Loan Stabilization Fund 250 Dollar Repo Claims Foreign Currency Investment 200 Franc Repo Claims Other Assets

150

100

0 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan 05 05 06 06 07 07 08 08 09 09 10

CNF Billions Liabilities 300 Deposits at SNB SNB Debt Certificates 250 Other Term Liabilities Other Liabilities Banknotes in Circulation 200

150

100

0 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan 05 05 06 06 07 07 08 08 09 09 10

SOURCE: Swiss National Bank.

492 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu per cent by March 2009 (see Figure 3). Expected ance sheet, reaching roughly ¥40 trillion in 2006 infla tion has been positive and slowly trending (Figure 7). The BOJ began a two-tier exit strategy upward since the SNB began quantitative easing. in 2006 when the CPI displayed signs of steady increase, allowing its holdings of bank bills and Japan long-term government securities (JGBs) to run Japan’s economic growth has slowed sharply off. Empirical studies have concluded that the since the bursting of the asset price bubble in the BOJ’s policy actions reduced longer-term rates, early 1990s—from a 5.1 percent annual rate dur - thereby flattening the yield curve, and had a ing the latter half of the 1980s, to a 1.5 percent positive, but small, effect on economic growth annual rate during the 1990s, to a less than 0.5 (Ugai, 2007). percent annual rate since 2000. 13 Here, we review In response to the current financial crisis, three episodes of BOJ quantitative easing efforts— the BOJ reduced its policy target rate to 0.3 per - the zero interest rate policy of the 1990s, a quan - cent from 0.5 percent on October 31, 2008, and titative easing policy from 2001 to 2006, and its to 0.1 percent on December 19, 2008, and has actions in response to the most recent financial initiated or expanded several programs to pro vide crisis. funds to the market. 15 Commercial paper was During the 1990s, the BOJ adopted the “zero purchased outright during 2009:Q1 (¥1.6 tril lion) interest rate policy” regime in which the policy but a rapid runoff occurred during the second target rate (overnight call rate) was set at 0.1 per - quarter, and corporate bond purchases were made cent. The BOJ maintained its balance sheet at a during the third quarter (¥400 billion). In addition, level just sufficient to sustain the overnight call approximately ¥4 trillion of commercial paper rate (near) zero. Nevertheless, real GDP growth was purchased during the first quarter under repo, during the decade averaged only 1.5 percent per and approximately ¥6 trillion in “special funds” year and the economy was stagnant at the decade’s was provided to banks as advances against cor - end. Worse, the threat of deflation had not eased— porate debt as collateral. These combined actions year-over-year consumer price index (CPI) infla - were modest relative to the size of the BOJ’s bal - tion was negative (see Figure 3). ance sheet (see Figure 7). In March 2001, with a policy rate at zero, the In November 2009, the BOJ declared that the BOJ initiated a quantitative easing policy in which economy had officially entered a period of defla - it would maintain the call rate at zero until the tion with a negative year-over-year change in the year-over-year increase in the CPI “became posi - CPI. On December 1, 2009, the BOJ announced 14 tive on a sustained basis.” The expansion of a new liquidity supply initiative to fight weak the balance sheet was regulated by a targeted level economic activity, deflation, and a rising yen of current account balances held by banks at the exchange rate. This program calls for the BOJ to BOJ. To achieve its targets, the BOJ purchased furnish up to ¥10 trillion in 3-month loans to government securities and bank bills backed by banks at the 0.1 percent target level of the policy eligible collateral (i.e., corporate bonds or com - rate (overnight unsecured call rate) against a mercial paper). The BOJ more or less smoothly variety of collateral, including JGBs, corporate increased its holdings of long-term Japanese gov - bonds, and commercial paper. The BOJ said the ernment bonds (JGBs); its holdings doubled by program should “further enhance easy monetary 2006 to roughly ¥90 trillion. In contrast, the BOJ conditions” and “encourage a further decline in purchased bank bills to quickly increase its bal - longer term interest rates.” Deflation continues in Japan, however, with a pace of –1 percent per 13 Shirakawa (2010) argues that quantitative easing policy actions successfully boosted Japanese economic activity and the story of year as of April 2010 (see Nishimura, 2010). “The Lost Decade” is myth.

14 For details, see Oda and Ueda (2007) or Maeda et al. (2005). 15 The BOJ’s actions as of April 2010 are well summarized in “The Humpage and Shenk (2008) provide a very readable summary. Bank of Japan’s Policy Measures in the Current Financial Crisis” Shiratsuka (2009) summarizes recent thought. (www.boj.or.jp/en/type/exp/seisaku_cfc/index.htm ).

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 493 Anderson, Gascon, Liu

Figure 7 Composition of Bank of Japan Balance Sheet

Yen Trillions Assets 180 Bills Purchased Loans and Discounts 160 Receivables under Resale Agreement Government Securities 140 Foreign Currency Assets 120 Other

100

0 Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan 00 01 02 03 04 05 06 07 08 09 10

Yen Trillions Liabilities 180 Deposits of the Government 160 Payables under Repurchase Agreement Deposits 140 Bank Notes 120 Other

100

0 Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan 00 01 02 03 04 05 06 07 08 09 10

NOTE: The monetary base of Japan is approximately the sum of banknotes in circulation outside the BOJ plus deposits held by banks at the BOJ. SOURCE: Bank of Japan.

494 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu

Scandinavia Policy actions by the Riksbank during the recent financial crisis resemble those during the Sweden and Finland provide a unique set of 1990s. The growth and composition of the cases because they have each undergone two Riksbank’s balance sheet are shown in Figure 8. episodes of quantitative easing during the past In September 2008, the Riksbank created a loan two decades: once during the banking crisis of facility that provided access to U.S. dollars, funded the 1990s and more recently during the 2007-09 by currency swap agreements with the Federal financial crisis. They are excellent case studies Reserve. 17 This program increased the balance- of how to do it right . sheet asset item “Claims on Residents inside Sweden. Sweden was affected by two severe Sweden Denominated in Foreign Currency” and financial crises during the past two decades: the the liability item “Liabilities outside Sweden 1990s Nordic banking crisis and the 2007-08 Denominated in Kronor.” To further assist banks, global financial crisis. Each time, the Riksbank in October 2008 the Riksbank created an addi - used large-scale increases in its balance sheet as tional loan facility designed to accept collateral a policy tool; the monetary base more than dou - with maturities longer than those accepted at its bled during the Nordic banking crisis and has traditional lending facilities. tripled during the most recent global financial To fund some lending programs, the Riksbank crisis. issued debt— Riksbank certificates —with a matu - The Nordic banking crisis of the early 1990s rity of one week. The Riksbank, unlike other affected all Scandinavian countries. 16 In each central banks, includes these certificates in its country, central bank support to the banking sys - measure of the monetary base (see Figure 8). 18 tem sharply increased the nation’s monetary base. Based on this measure, Sweden’s monetary base Starting in 1992, the Riksbank used its foreign increased almost fivefold from SEK 105.92 bil lion currency reserves to provide liquidity support to in September 2008 to a peak in November 2009. banks. Because the government had guaranteed Alternatively, measured using the commonplace all bank debt, the Riksbank allowed banks to bor - definition of the monetary base as the sum of row freely through normal liquidity facilities. currency in circulation plus the deposits held The monetary base more than doubled within 10 by financial institutions at the central bank, months—from SEK 83.73 billion in August 1993 Sweden’s monetary base tripled to peak of SEK to SEK 208.26 billion in June 1994. When condi - 319 billion in December 2008. tions stabilized, the monetary base decreased Finland. Finland likely was the country rapidly to SEK 81.11 billion by February 1997. most severely affected by the Nordic banking Although inflation did not increase significantly crisis, recording decreases in real GDP growth during or after the monetary base expansion for three consecutive years (1990:Q3–1993:Q3). 19 period, CPI inflation during 1994 was more than From 1992 to 1997, strong government inter ven - double the Riksbank’s 2 percent inflation target tion included equity investments in the nation’s and inflation expectations increased modestly as the Riksbank expanded its balance sheet. As a 17 For details, see Öberg (2009). result, even with the monetary base at elevated 18 For comparison purposes, the data in Figure 1 report a measure of levels, the Riksbank gradually increased its pol icy the monetary base without Riksbank certificates. target rate during 1994 and 1995. The inflation 19 The Bank of Finland’s takeover of the shaky commercial bank rate returned to a 2 percent pace by late 1995, Skopbank in the fall of 1991 perhaps was the climax of the coun - try’s financial crisis. Skopbank was one of the pillars of Finland’s slipping negatively the following year. Inflation commercial banking industry, widely referred to as the “central expectations retreated below the Riksbank’s 2 bank” for the country’s savings banks. The takeover is reported to have eventually cost FIM 15 billion, the equivalent of 3 percent of percent inflation target to near 1.5 percent. GDP in 1991 (Sandal, 2004). The eventual resolution of the crisis and unwinding of government support programs cost Finland an amount equal to approximately 6 percent of one year’s GDP; for 16 Honkapohja (2009) and Anderson (2009) discuss causes and the Sweden and Norway, the eventual cost was near zero. For addi - policy response; see also the references cited therein. tional details, see Anderson (2009).

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 495 Anderson, Gascon, Liu

Figure 8 Composition of the Riksbank’s Balance Sheet

SEK Billions Assets 800 Claims on Residents Inside Sweden Denominated in Foreign Currency 700 Lending to Monetary Policy Counterparties Claims on Residents Outside Sweden 600 Denominated in Foreign Currency Gold and Other Assets 500

400

300

200

100

0 Mar Sep Mar Sep Mar Sep Mar Sep Mar 06 06 07 07 08 08 09 09 10

SEK Billions Liabilities 800 Other Liabilites and Capital 700 Liabilities Outside Sweden in SEK Riksbank Certificates 600 Bank Deposits Currency in Circulation 500

400

300

200

100

0 Mar Sep Mar Sep Mar Sep Mar Sep Mar 06 06 07 07 08 08 09 09 10

NOTE: The Riksbank defines Sweden’s monetary base as “Currency in Circulation” plus “Bank Deposits” and “Riksbank Certificates.” SOURCE: The Riksbank.

496 NOVEMBER/DECEMBER 2010 FEDERALRESERVEBANKOFST. LOUIS REVIEW Anderson, Gascon, Liu

Figure 9 Composition of Bank of Finland’s Balance Sheet

FIM Billions Assets 100 Other Assets Gold and Foreign Currency Claims 90 Claims on Financial Institutions Loans for Stabilizing the Money Market 80 Capitalized Expenditures Due to 70 Stabilizing the Money Market

0 Jan Jan Jan Jan Jan Jan Jan Jan 91 92 93 94 95 96 97 98

FIM Billions Liabilities 100 Other Liabilities and Capital 90 CDs 80 Liabilities to Financial Institutions Notes and Coin in Circulation 70

0 Jan Jan Jan Jan Jan Jan Jan Jan 91 92 93 94 95 96 97 98

SOURCE: Bank of Finland.

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 497 Anderson, Gascon, Liu banks and in a government-funded bank guar an - reduced by 300 basis points during the last four tee fund. By the end of 1997, the government had months of 2008 (including a 100-basis-point cut extended FIM 43 billion in direct bank support. 20 on October 8, 2008, coordinated with 50-basis- The Finnish monetary base expanded from FIM point or larger reductions by other G-10 central 34.5 billion in January 1992 to FIM 68.7 billion banks). Private net capital flows were negative in in March 1997 (Figure 9). Banking institutions the third and fourth quarters of 2008, in part started repaying the government in October because Australian banks found Federal Reserve 1997; thereafter, the monetary base decreased foreign currency swap lines a lower-cost source of rapidly to FIM 36.2 billion in December 1997. funds than alternatives, particularly in 2008:Q4. 21 Inflation in Finland was high before the bank - By mid-2009, global short-term credit markets ing crisis started: The monthly year-over-year CPI were normalizing and Australian banks were rate topped 7 percent in spring 1990 and stayed obtaining funds in the market below the cost of above 5 percent during 1990. Inflation subsided the Fed’s swap facility. The policy target rate as the crisis deepened; the year-over-year rate reached its low of 3 percent on April 8, 2009, fell below 2 percent (the Bank of Finland’s infla - and on October 7, 2009, the Research Bank of tion target) in mid-1993. Inflation remained low Australia (RBA) started the process of increasing for the remainder of the 1990s and inflation expec - its target rate toward “a more normal setting” tations remained anchored between 2 and 2.5 with a 25-basis-point increase. percent. The RBA’s policy actions do not merit the Finland adopted the euro in 1999, and hence label “quantitative easing” because the policy during the financial crisis of 2007-09 exercised target rate never reached zero during mid-2009. no independent monetary policy. Its response Yet, the RBA balance sheet expanded sharply has been limited to banking supervision and between August and December 2008, resulting in support (Liikanen, 2009). Perhaps the largest a 54 percent increase in the monetary base (mea - country-specific impact was the spinoff of the sured as the sum of “Reserves and Notes” plus Finnish operations of the Icelandic bank Glitnir “Term Deposits”). Figure 10 shows the growth into a new Finnish corporation in October 2008. and changing composition of the RBA balance sheet. A number of domestic programs affected the balance sheet, including (i) broadening the pool of eligible collateral accepted by the RBA, OTHER CASES OF LARGE-SCALE (ii) conducting open market operations at longer BALANCE-SHEET INCREASES maturities to increase the impact on longer-term In this section, we briefly examine the expe - yields, and (iii) offering term deposits at the RBA. 22 riences of three other countries with recent large- Deposits obtained by the RBA under the Federal scale central bank balance-sheet increases. Reserve’s swap lines are included in the asset “Gold and Foreign Exchange,” whereas the Australia Australian dollars held as collateral by the Federal Reserve are included in the liability item “Deposits Australia experienced a sharp but mild reces - of Overseas Institutions.” Interestingly, the RBA sion during late 2008 and early 2009, caused in unwound its balance-sheet expansion in the first part by reduced export demand and weaker con - half of 2009 without rapid increases in its policy sumer confidence. Stevens (2009) notes that the target rate: By May 2009, the level of the mone - Reserve Bank of Australia had started easing tary base had returned approximately to its trend pol icy in early September due to moderating with the policy target rate at 4.25 percent. demand, reducing its target overnight rate from 7.25 percent to 7 percent. Easing accelerated 21 D’Arcy and Ossolinski (2009) report that some Australian banks, after the Lehman Brothers failure; the target was successful bidders for U.S. dollars at RBA auctions, apparently reloaned the dollar funds to foreign parents.

20 For details, see Mutikainen (1998). 22 For additional details, see Stevens (2010) and Debelle (2008).

498 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu

Figure 10 Composition of Reserve Bank of Australia Balance Sheet

A$ Billions Assets 180 Gold and Foreign Exchange 160 Australian Dollar Securities Other Assets 140 Clear Items

120

100

0 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan 05 05 06 06 07 07 08 08 09 09 10

A$ Billions Liabilities 180 Other 160 Deposits of Government Deposits of Overseas Institutions 140 Term Deposits Reserves and Notes 120

100

0 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan 05 05 06 06 07 07 08 08 09 09 10

NOTE: The monetary base of Australia is approximately equal to “Reserves and Notes” plus “Term Deposits.” SOURCE: Reserve Bank of Australia.

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 499 Anderson, Gascon, Liu

There is no evidence that the RBA’s actions reducing the krona requirement to 30 percent. affected inflation. Before the crisis, the Australian Bank runs began the first week of October: economy had experienced year-over-year growth Currency in circulation increased 53 percent in consumer prices averaging just under 3 per cent, that week (see Central Bank of Iceland, 2009b, within the RBA’s target range of 2 to 3 percent. pp. 24-25). The CBI used its reserves—even old Inflation expectations, however, reached as high banknotes no longer intended for circulation— as 6 percent as Australia’s business cycle peaked to meet public demand. The combination of in May 2008 (see Figure 3), but the late 2008 dis - increased loans to banks 24 and currency in cir - turbances in financial markets caused inflation culation caused the CBI’s balance sheet to increase expectations to decline precipitously. Unlike in rapidly during the fall of 2008. At its peak, the United States, United Kingdom, and Japan, Iceland’s monetary base had increased by 70 inflation expectations in Australia never raised per cent with highly volatile swings from month the specter of deflation. A prompt unwinding of to month. its balance sheet appears to have protected infla - The CBI sought to distinguish its traditional tion stability. approach to monetary policy (targeting the dis - count rate) from its balance-sheet actions. In October 2008, for example, the CBI increased its Iceland policy rate (the nominal discount rate) from 12 We mention Iceland primarily because of its percent to 18 percent. At this time, year-over- year widely reported role in the banking crises of sev - inflation had been steadily increasing since 2007 eral European nations. Iceland is unusual, in this and was close to 15 percent. In pursuing its infla - analysis, because its policy target rate and cen tral tion target of 2.5 percent, the CBI had held its bank balance sheet increased simultaneously. policy rate above 10 percent since 2005. 25 The Iceland’s financial system was seriously CBI changed course in March 2009 after infla tion harmed by the recent financial crisis. Three of began to subside (albeit still at high levels). The the country’s largest commercial banks failed in nominal discount rate dropped to 11 percent in October 2008; the government assumed the role November 2009. Because inflation was well above as insurer of their deposits and replaced each the target at the beginning of the crisis, it is diffi - board of directors. 23 These banks, however, had cult to assess whether the balance-sheet expan - customer liabilities with other countries in addi - sion affected either actual or expected inflation. tion to Iceland, equal to roughly 10 times Iceland’s However, the CBI forecasts that inflation will annual gross national product (see Central Bank reach its target of 2.5 percent sometime in early of Iceland [CBI], 2009b, p. 13). As a result of the 2011 (Central Bank of Iceland, 2009b), suggest ing bank failures, Iceland’s exchange rate fell sharply that the monetary base increases are not antici - in October 2008; Iceland obtained a $2.1 billion pated to increase inflation pressures. loan from the International Monetary Fund (IMF) to stabilize its financial system. New Zealan d 26 Before receiving the IMF loan to stabilize the New Zealand’s monetary base increased 138 country’s economy, in January 2008 the CBI had percent between July and December 2006. No expanded the list of eligible collateral at its regu - adverse “shock” to the economy caused the lar lending facilities to include bonds issued in Reserve Bank of New Zealand (RBNZ) to increase dollars, euros, or British pounds but continued its monetary base. Rather, the increase was the to require that at least 50 percent of collateral be denominated in kronor. By August 2008, the CBI 24 Iceland’s treasury purchased a 75 percent share in one bank (Glitnir) went a step further, expanding the list of eligible for €600 million and a week later the CBI loaned €500 million to collateral to include asset-backed securities and another bank (Kaupthing) for four days. 25 The CBI adopted an inflation target in March 2001.

23 For additional details, see Central Bank of Iceland (2009a). 26 For additional details, see Nield (2008).

500 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu

Figure 11 Composition of Reserve Bank of New Zealand Balance Sheet

NZD Billions Assets 40 Marketable Securities in Foreign Currency 35 Advances (in NZD) Other Assets in Foreign Currency 30 Government Securities (in NZD) Other Assets (in NZD) 25

0 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan 05 05 06 06 07 07 08 08 09 09 10

NZD Billions Liabilities 40 Other Liabilities 35 Reserve Bank Bills Deposits 30 Currency

0 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan 05 05 06 06 07 07 08 08 09 09 10

NOTE: The Reserve Bank of New Zealand does not publish a series on the monetary base. It does publish data on settlement institu tion deposits at the RBNZ and currency in circulation, the sum of which is a measure of the monetary base. SOURCE: Reserve Bank of New Zealand.

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 501 Anderson, Gascon, Liu culmination of a collaborative project between CONCLUSION the Reserve Bank and the settlement banks to improve the payment system to “reduce risk and During the past two decades, large increases— enhance certainty in the financial system” (Nield, and decreases—in central bank balance sheets 2008, p. 10). The RBNZ-operated payment sys tem have become a viable monetary policy tool. His - does not permit daylight overdrafts (i.e., pay ments torically, doubling or tripling a country’s mone - on behalf of a bank that exceed the bank’s avail - tary base was a recipe for certain higher inflation. able account balance at the RBNZ). Increased Often such increases occurred only as part of a settlement balances at the RBNZ significantly failed fiscal policy or, perhaps, as part of a pol icy reduced delays in the payment system. The RBNZ to defend the exchange rate. Both economic also began paying interest on its settlement bal - models and central bank experience during the ances to increase the acceptance of the system past two decades suggest that such changes are and to discourage banks from using these reserve useful policy tools if the public understands the balances to fund new lending. increase is temporary and if the central bank has The new settlement system was imple mented some credibility with respect to desiring a low, between August and October 2006. The RBNZ stable rate of inflation. We find little increased expected the monetary base would increase to inflation impact from such expansions. between NZ$7 billion and NZ$10 billion (Nield, For monetary policy, our study suggests sev - 2008) as banks gradually unwound Treasury bill eral findings: holdings and used foreign exchange swaps to (i) A large increase in a nation’s balance sheet purchase New Zealand dollars. The RBNZ deter - over a short time can be stimulative. mined that since the cash was purchased at rates (ii) The reasons for the action should be com - consistent with the policy rate, there would be municated. Inflation expectations do not no inflationary pressures. In fact, year-over-year move if households and firms understand growth in consumer prices fell from 3.4 percent the reason(s) for policy actions so long as in 2006:Q3 to 1.8 percent in 2007:Q3. It did, how - the central bank can credibly commit to ever, increase to 5 percent in 2008:Q3 only to unwinding the expansion when appropriate. again drop under 2 percent in 2009:Q2. Over the same period, inflation expectations remained (iii) The type of assets purchased matters less well anchored around 3 percent. than the balance-sheet expansion. The financial crisis hit New Zealand in mid- (iv) When the crisis has passed, the balance 2008. Between July 2008 and April 2009, as infla - sheet should be unwound promptly. tion expectations plummeted, the RBNZ reduced its target for the official cash rate to 2.5 percent from 8.25 percent. Although the RBNZ has not emphasized increases in its balance sheet as part of its policy, Figure 11 shows clearly that large balance-sheet increases did occur. More recently, the RBNZ balance sheet has contracted some what as the crisis has eased, although a weak eco nomic outlook has caused the RBNZ to sustain a low policy target rate.

502 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu

REFERENCES Anderson, Richard G. “Resolving a Banking Crisis, the Nordic Way.” Federal Reserve Bank of St. Louis Economic Synopses , 2009, No. 10, February 18, 2009; http://research.stlouisfed.org/publications/es/09/ES0910.pdf . Bank of England. “Asset Purchase Facility.” 2009a; www.bankofengland.co.uk/markets/apf/index.htm . Bank of England. “Quantitative Easing Explained.” 2009b; www.bankofengland.co.uk/monetarypolicy/assetpurchases.htm . Bank of Finland. Bank of Finland Annual Report 2008 . March 2009; www.bof.fi/NR/rdonlyres/19D57A56-427D-4A21-B3CB-0D3B32217A59/0/VK08E.pdf . Barro, Robert J. “Rational Expectations and the Role of Monetary Policy.” Journal of Monetary Economics , January 1976, 2(1), pp. 1-32. Batchelor, Roy A. and Dua, Pami. “Household versus Economist Forecasts of Inflation: A Reassessment: Note.” Journal of Money, Credit, and Banking , May 1989, 21 (2), pp. 252-257. Berentsen, Aleksander and Waller, Christopher. “Price Level Targeting and Stabilization Policy.” Working Paper No. 2009-033B, Federal Reserve Bank of St. Louis, revised October 2009; http://research.stlouisfed.org/wp/2009/2009-033.pdf (forthcoming in Journal of Money, Credit, and Banking ). Bernanke, Ben S. “The Federal Reserve’s Balance Sheet: An Update.” Speech at the Federal Reserve Board Conference on Key Developments in Monetary Policy, Washington, DC, October 8, 2009; www.federalreserve.gov/newsevents/speech/bernanke20091008a.htm . Bernanke, Ben S. and Reinhart, Vincent R. “Conducting Monetary Policy at Very Low Short-Term Interest Rates.” American Economic Review , May 2004, 94 (2), pp. 85-90. Bernanke, Ben S.; Reinhart, Vincent R. and Sack, Brian P. “Monetary Policy Alternatives at the Zero Bound: An Empirical Assessment.” Brookings Papers on Economic Activity , 2004, 2, pp. 1-78. Borio, Claudio and Disyatat, Piti. “Unconventional Monetary Policies: An Appraisal.” BIS Working Papers No. 292, Bank for International Settlements, November 2009; www.bis.org/publ/work292.pdf . Central Bank of Iceland. Annual Report 2008 . 2009a; www.sedlabanki.is/lisalib/getfile.aspx?itemid=7076 . Central Bank of Iceland. Financial Stability Report 2009 . October 6, 2009b; www.sedlabanki.is/?PageID=1061 . Crowe, Christopher and Meade, Ellen E. “Central Bank Independence and Transparency: Evolution and Effectiveness.” European Journal of Political Economy , December 2008, 24 (4), pp. 763-77. Clarida, Richard; Galí, Jordi and Gertler, Mark. “The Science of Monetary Policy: A New Keynesian Perspective.” Journal of Economic Literature , December 1999, 37 (4), pp. 1661-707. Cúrdia, Vasco and Woodford, Michael. “The Central-Bank Balance Sheet as an Instrument of Monetary Policy.” Prepared for Carnegie-Rochester Conference on Public Policy, “The Future of Central Banking,” April 16-17, 2010a; www.carnegie-rochester.rochester.edu/april10-pdfs/Curdia%20Woodford.pdf . Cúrdia, Vasco and Woodford, Michael. “Conventional and Unconventional Monetary Policy.” Federal Reserve Bank of St. Louis Review , July/August 92 (4), 2010b; http://research.stlouisfed.org/publications/review/10/07/Curdia.pdf . D’Arcy, Patrick and Ossolinski, Crystal. “Australian Capital Flows and the Financial Crisis.” Reserve Bank of Australia Bulletin , November 2009, pp. 1-6; www.rba.gov.au/publications/bulletin/2009/nov/pdf/bu-1109-1.pdf . Debelle, Guy. “Market Operations in the Past Year.” Presented at the 2008 FTA Congress, Melbourne, Australia, October 31, 2008, Reserve Bank of Australia; www.rba.gov.au/speeches/2008/sp-ag-311008.html .

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 503 Anderson, Gascon, Liu

Gertler, Mark and Karadi, Peter. “A Model of Unconventional Monetary Policy.” Unpublished manuscript, New York University, April 2009; www.econ.nyu.edu/user/gertlerm/gertlerkaradiapril2010conference.pdf . Goodfriend, Marvin S. and King, Robert G. “A Note on the Neutrality of Temporary Monetary Disturbances.” Journal of Monetary Economics , 1981, 7(3), pp. 371-85. Humpage, Owen F. and Shenk, Michael. “Japan’s Quantitative Easing Policy.” Federal Reserve Bank of Cleveland Economic Trends , December 10, 2008; www.clevelandfed.org/research/trends/2008/1208/01intmar.cfm . Honkapohja, Seppo. “The 1990’s Financial Crises in Nordic Countries.” Bank of Finland Research Discussion Papers No. 5, 2009. Jeanne, Olivier and Svensson, Lars E.O. “Credible Commitment to Optimal Escape from a Liquidity Trap: The Role of the Balance Sheet of an Independent Central Bank.” American Economic Review , March 2007, 97 (1), pp. 474-90. Laurens, Bernard J.; Arnone, Marco and Seglaotto, Jean-Francois. Central Bank Independence, Accountability, and Transparency . Hampshire, UK: Palgrave Macmillan, 2009. Liikanen, Erkki. “Central Banking in a Global Financial Crisis.” Presented at the Finnish Economic Association meeting, January 29, 2009; www.bof.fi/en/suomen_pankki/ajankohtaista/puheet/2009/EL_puhe29012009.htm . Maeda, Eiji; Fujiwara, Bunya; Mineshima, Aiko and Taniguchi, Ken. “Japan’s Open Market Operations under the Quantitative Easing Policy.” Bank of Japan Working Paper Series No. 05-E-3, April 8, 2005; www.boj.or.jp/en/type/ronbun/ron/wps/data/wp05e03.pdf . Meltzer, Allan. “The Origins of the Great Inflation.” Federal Reserve Bank of St. Louis Review , March/April 2005, 87 (2, Part 2), pp. 145-75; http://research.stlouisfed.org/publications/review/05/03/part2/Meltzer.pdf . Mutikainen, Tapio. “Recession, Economic Policy and Banking Crisis Management in Finland in the 1990’s.” Ministry of Finance discussion paper, June 23, 1998; www.vm.fi/vm/fi/04_julkaisut_ja_asiakirjat/03_muut_asiakirjat/discussion_paper.pdf . Nield, Ian. “Evolution of the Reserve Bank’s Liquidity Facilities.” Reserve Bank of New Zealand Bulletin , December 2008, 71 (4), pp. 5-17; www.rbnz.govt.nz/research/bulletin/2007_2011/2008dec71_4nield.pdf . Nishimura, Kiyohiko G. “Japan’s Economy and Monetary Policy.” Speech at a Meeting with Business Leaders in Miyagi, April 21, 2010; www.boj.or.jp/en/type/press/koen07/data/ko1004d.pdf . Öberg, Svante. “Sweden and the Financial Crisis.” Speech at Carlson Investment Management, Stockholm, January 20, 2009; www.riksbank.com/templates/Page.aspx?id=30276 . Oda, Nobuyuki and Ueda, Kazuo. “The Effects of the Bank of Japan’s Zero Interest Rate Commitment and Quantitative Monetary Easing on the Yield Curve: A Macro-Finance Approach.” Japanese Economic Review , 2007, 58 (3), pp. 303-28. Reis, Ricardo. “An Interpretation of the Unconventional U.S. Monetary Policy Response to the 2007-09 Crisis.” Presented at the September 2009 Brookings Conference on Economic Activity. (Forthcoming in Brookings Papers on Economic Activity ; August 2009 draft version available at www.brookings.edu/~/media/Files/Programs/ES/BPEA/2009_fall_bpea_papers/2009_fall_bpea_reis.pdf ). Sandal, Knut. “The Nordic Banking Crises in the Early 1990s—Resolution Methods and Fiscal Costs,” in Thorvald G. Moe, Jon A. Solheim, and Bent Vale, eds., The Norwegian Banking Crisis (Norges Banks Skriftserie/Occasional Papers No. 33, 2004). Chap. 3. Oslo, Norway: Norges Bank, 2004, pp. 77-115; www.norges-bank.no/upload/import/publikasjoner/skriftserie/33/chapter3.pdf . Sargent, Thomas U. “The Ends of Four Big Inflations,” in Robert Hall, ed. Inflation: Causes and Effects . Chap. 2. Chicago: University of Chicago Press, 1982, pp. 41-92; www.nber.org/chapters/c11452.pdf .

504 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW Anderson, Gascon, Liu

Shirakawa, Maasaki. “Uniqueness or Similarity? Japan’s Post-Bubble Experience in Monetary Policy Studies.” Keynote address at Second IJCB Fall Conference hosted by the Institute for Monetary and Economic Studies, the Bank of Japan. September 16, 2010; www.boj.or.jp/en/type/press/koen07/data/ko1009c.pdf . Shiratsuka, Shigenori. “Size and Composition of the Central Bank Balance Sheet: Revisiting Japan’s Experience of the Quantitative Easing Policy.” Bank of Japan, IMES Discussion Paper No. 2009-E-25, November 2009; www.imes.boj.or.jp/english/publication/edps/2009/09-E-25.pdf . Söderström, Ulf and Westermark, Andreas. “Monetary Policy When the Interest Rate Is Zero.” Sveriges Riksbank Economic Review , 2, pp. 5-30; www.riksbank.se/upload/Dokument_riksbank/Kat_publicerat/Rapporter/2009/2009_2_pv_eng.pdf . Stevens, Glenn. “The Conduct of Monetary Policy in Crisis and Recovery.” Address to The John Curtin Institute of Public Policy and Financial Services Institute of Australasia Public Policy Breakfast Forum, Perth, Australia, October 15, 2009; www.rba.gov.au/speeches/2009/sp-gov-151009.html . Stevens, Glenn. “Recent Financial Developments.” Address to ACI 2010 49th World Congress, Sydney, Australia, March 26, 2010; www.bis.org/review/r100330c.pdf . Swiss National Bank. 2009 Financial Stability Report . June 2009; www.snb.ch/en/mmr/reference/stabrep_2009/source/stabrep_2009.en.pdf . Ugai, Hiroshi. “Effects of the Quantitative Easing Policy: A Survey of Empirical Analyses.” Monetary and Economic Studies , March 2007, 25 (1), pp. 1-48. Woodford, Michael. “Inflation Targeting and Optimal Monetary Policy.” Federal Reserve Bank of St. Louis Review , July/August 2004, 86 (4), pp. 15-41; http://research.stlouisfed.org/publications/review/04/07/Woodford.pdf .

FEDERAL RESERVE BANK OF ST . LOUIS REVIEW NOVEMBER /DECEMBER 2010 505 506 NOVEMBER /DECEMBER 2010 FEDERAL RESERVE BANK OF ST . LOUIS REVIEW