Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Monetary Policy, Corporate Finance & Investment
James Cloyne (UC Davis, NBER & CEPR)Clodo Ferreira (Bank of Spain)Maren Froemel (LBS & BoE)Paolo Surico (LBS & BoE)
March 2019
The views expressed are those of the authors and do not necessarily reflect theviews of the Bank of Spain, the Euro-system, Bank of England, MPC, FPC or
PRA. This research has been funded by the European Research Council.
Monetary Policy and Firm Finance
I How does monetary policy affect firm investment? Which type offirms are most responsive?
I How do firms’ balance sheets respond?
I How important are financial frictions?
Our Approach
� Firm-level panel approach for the US and UKI Heterogeneity in the dynamic effects of policy across firms.I Micro data with macro identification of policy rate changes.
� Which firms change investment the most?I Which proxies for financial constraints should we focus on?I Age, size, growth, leverage, liquidity, dividend status, QI Multivariate heterogeneity analysis.
� What happens to these firms’ balance sheets?I Borrowing, equity, earnings/cash flows, share prices.
� Heterogeneity used to examine the transmission mechanism.
Our Approach
� Firm-level panel approach for the US and UKI Heterogeneity in the dynamic effects of policy across firms.I Micro data with macro identification of policy rate changes.
� Which firms change investment the most?I Which proxies for financial constraints should we focus on?I Age, size, growth, leverage, liquidity, dividend status, QI Multivariate heterogeneity analysis.
� What happens to these firms’ balance sheets?I Borrowing, equity, earnings/cash flows, share prices.
� Heterogeneity used to examine the transmission mechanism.
Our Approach
� Firm-level panel approach for the US and UKI Heterogeneity in the dynamic effects of policy across firms.I Micro data with macro identification of policy rate changes.
� Which firms change investment the most?I Which proxies for financial constraints should we focus on?I Age, size, growth, leverage, liquidity, dividend status, QI Multivariate heterogeneity analysis.
� What happens to these firms’ balance sheets?I Borrowing, equity, earnings/cash flows, share prices.
� Heterogeneity used to examine the transmission mechanism.
Our Approach
� Firm-level panel approach for the US and UKI Heterogeneity in the dynamic effects of policy across firms.I Micro data with macro identification of policy rate changes.
� Which firms change investment the most?I Which proxies for financial constraints should we focus on?I Age, size, growth, leverage, liquidity, dividend status, QI Multivariate heterogeneity analysis.
� What happens to these firms’ balance sheets?I Borrowing, equity, earnings/cash flows, share prices.
� Heterogeneity used to examine the transmission mechanism.
Main Findings
� Investment1. Age is a robust predictor: Younger firms respond the most.
Quantitatively important to account for the aggregate response.
2. Especially pronounced for firms not paying dividends.
3. Robust to controlling for more traditional characteristics.
� Firm Finance4. Younger firms: lower earnings, lower credit scores and leverage.
Less likely to pay dividends. Borrowing is more asset-based.
5. After a contractionary monetary policy, net worth falls for all firms.But borrowing falls the most for younger firms paying no dividends.
� Interpretation of the evidence/channel: higher interest rates–>–> lower asset values –> borrowing falls –> investment falls.
Main Findings
� Investment1. Age is a robust predictor: Younger firms respond the most.
Quantitatively important to account for the aggregate response.
2. Especially pronounced for firms not paying dividends.
3. Robust to controlling for more traditional characteristics.
� Firm Finance4. Younger firms: lower earnings, lower credit scores and leverage.
Less likely to pay dividends. Borrowing is more asset-based.
5. After a contractionary monetary policy, net worth falls for all firms.But borrowing falls the most for younger firms paying no dividends.
� Interpretation of the evidence/channel: higher interest rates–>–> lower asset values –> borrowing falls –> investment falls.
Main Findings
� Investment1. Age is a robust predictor: Younger firms respond the most.
Quantitatively important to account for the aggregate response.
2. Especially pronounced for firms not paying dividends.
3. Robust to controlling for more traditional characteristics.
� Firm Finance4. Younger firms: lower earnings, lower credit scores and leverage.
Less likely to pay dividends. Borrowing is more asset-based.
5. After a contractionary monetary policy, net worth falls for all firms.But borrowing falls the most for younger firms paying no dividends.
� Interpretation of the evidence/channel: higher interest rates–>–> lower asset values –> borrowing falls –> investment falls.
Literature
Empirics...I Age & employment (Haltiwanger et al., 2013, Bahaj et al., 2018)
I Age, size, leverage & business cycles(Dinlersoz et al., 2018, Crouzet & Mehrotra 2018)
I Firm Finance & business cycles(Covas & den Haan, 2011, Begenau & Salomao, 2018)
I Investment & financial frictions (Fazzari et al. 1988, Gertler &Gilchrist 1994, Ottonello & Winberry 2018, Jeenas, 2018)
I Firm borrowing constraints (Lian & Ma, 2018, Drechsel, 2018)
Financial frictions...I Age & growth prospects (Cooley-Quadrini, 2001, Cooper et al. 2006)
I Leverage, asset prices/collateral values & monetary policy(Bernanke, Gertler & Gilchrist, 1999, Kiyotaki & Moore, 1997, etc.)
Outline
Data & Approach
Age as a Proxy for Financial Constraints
Heterogeneity in the Response of Investment
Firm Finance and Balance Sheet Response
Concluding remarks
Firm Data: Panel of Public Firms
I Compustat quarterly panel (US). Worldscope annual panel (UK).Sample period: 1986-2016.
I Also make use of corporate bonds and asset price data(CRSP and Thomson Reuters)
I Key variables of interest:I Investment: capital expenditure/net PPE.I Age: Worldscope years since incorporation.I Other variables: assets, debt, leverage (debt/assets), liquidity,
Tobin’s Q, equity, share prices, earnings/sales, dividends paid,interest payments.
7 / 30
Monetary Policy: Identification
I Gertler-Karadi approach: High frequency surprises in short ratefutures around policy announcements.
I Instrument available since 2001 for the U.K. (Gerko-Rey) and1991 for the U.S. (Gertler-Karadi).
I Gertler-Karadi (2015)/Mertens-Ravn (2013): surprises asproxies for structural shocks in the Vector Autoregression.
I Identifies a series of monetary policy shocks for the full sample.SHOCK SERIES
8 / 30
Empirical Specification
∆hXi,t+h = γhi +
G∑g=1
βhg · I
[Zi,t−1 ∈ g
]· Rt +
G∑g=1
αhg ·I[Zi,t−1 ∈ g
]+εi,t+h
I Baseline Xi,t+h: capex/net PPE at horizon h. Also look at equity,borrowing, earnings, share prices etc.
I Zi,t−1: variable defining a group: age, size, growth, leverage,beta, paying dividends in previous year. Could be multivariate.
I Rt : interest rate in GK/GR instrumented with structural shocks.
9 / 30
Sense Check: The Average Effect
United States United Kingdom
-1-.5
0.5
Percent
1 4 8 12 16 20Quarters
-.6-.4
-.20
.2.4
Perc
ent
0 1 2 3 4Year
Response of the investment ratio to a 25 basis point increase in interest rates. Confidence bands 90%. Firm-time clustering.
Consistent with MACRO EVIDENCE using data from national statistics.IRFs even more similar when reporting at the same ANNUAL FREQUENCY
10 / 30
Outline
Data & Approach
Age as a Proxy for Financial Constraints
Heterogeneity in the Response of Investment
Firm Finance and Balance Sheet Response
Concluding remarks
Size, Growth and Earnings by AGE
Younger firms are smaller, have lower cash-flows but grow faster
Size Asset growth Earnings
Uni
ted
Sta
tes
4.5
55.
56
6.5
log(
asse
ts)
5 15 25 35 45 55Age
24
68
%
5 15 25 35 45 55Age
24
68
1012
%
5 15 25 35 45 55Age
Uni
ted
Kin
gdom
3.5
44.
55
log(
asse
ts)
5 15 25 35 45 55Age
56
78
910
%
5 15 25 35 45 55Age
89
1011
12%
5 15 25 35 45 55Age
Regressions of the variable of interest on age, squared age, sectorsXtime fixed effects (and size).
12 / 30
Financial Characteristics by AGEYounger firms: lower credit scores/less likely to pay dividends.
Credit scores Dividends & BondsU
nite
dS
tate
s
0.2
.4.6
.81
prob
abili
ty
5 15 25 35 45 55Age
low rating high rating
0.2
.4.6
.81
prob
abili
ty
5 15 25 35 45 55Age
Prob. issuing bondsProb. paying dividends
Prob. buy-backs
Uni
ted
Kin
gdom
6065
7075
80Li
near
Pre
dict
ion
5 15 25 35 45 55Age
Credit Score and Age
0.2
.4.6
.81
Prob
abili
ty
5 15 25 35 45 55Age
Prob. issuing bondsProb. paying dividends
Prob. buyback
Based on regressions of the variable of interest on age, squared age, sectorsXtime fixed effects (and size).
13 / 30
Leverage and Liquidity by AGE
Younger firms are less leveraged/hold more liquid assets
Leverage LiquidityU
nite
dS
tate
s
2527
2931
3335
%
5 15 25 35 45 55Age
1416
1820
%
5 15 25 35 45 55Age
Uni
ted
Kin
gdom
1012
.515
17.5
20%
5 15 25 35 45 55Age
1012
1416
%
5 15 25 35 45 55Age
Based on regressions of the variable of interest on age, squared age, sectorsXtime fixed effects (and size).
14 / 30
Summary: What Does Age Capture?
Younger firms tend to:
I be smaller
I have lower earnings
I have lowerI credit scoresI probability of paying dividends
But younger firms also have:
I lower leverage and higher liquid assets
I faster growth and higher (average) Tobin’s Q
Outline
Data & Approach
Age as a Proxy for Financial Constraints
Heterogeneity in the Response of Investment
Firm Finance and Balance Sheet Response
Concluding remarks
Response of Investment by AGE
Younger Middle-aged Older
Uni
ted
Sta
tes
-1.5
-.5-1
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-.5-1
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-.5-1
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-1.5
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
-1.5
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
-1.5
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
17 / 30
Investment Response by AGE & DIVIDENDS: U.S.
Younger OlderN
Odi
vide
nds
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
Div
iden
ds
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
18 / 30
Investment Response by AGE & DIVIDENDS: U.K.
Younger OlderN
Odi
vide
nds
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
Div
iden
ds
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
19 / 30
YOUNGER Firms Drive the Average Effect
Younger OlderNo Div Paid Div No Div Paid Div
U.S. 75.5% 6.7% 13.0% 4.8%[ 68.1 , 82.8 ] [ -1.6 , 15.5 ] [ 11.7 , 14.5 ] [ 1.9 , 7.4 ]
U.K. 83.6% 13.1% 2.9% 0.4%[ 70.4 , 96.8 ] [ 2.9 , 23.2 ] [ -2.2 , 8.1 ] [ -5.9 , 6.9 ]
Notes: 95% CI in square brackets
20 / 30
Multidimensional Heterogeneity Analysis
Age is correlated with a range of other factors. Do our IRFs simplycapture one of these other factors? No.
Results are robust to conditioning on:
1. Size charts
2. Leverage charts
3. Liquidity charts
4. Firm growth charts and Tobin’s Q charts
5. Risk see section 7.2 in the paper
21 / 30
Outline
Data & Approach
Age as a Proxy for Financial Constraints
Heterogeneity in the Response of Investment
Firm Finance and Balance Sheet Response
Concluding remarks
BORROWING responds most for Younger/No Div.
Younger & NO dividends Older & Paying dividendsU
nite
dS
tate
s
-4-2
02
Percent
1 4 7 10 13 16 19Quarters
-4-2
02
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-5-2
02
Perc
ent
1 2 3 4 5Year end (Q4)
-5-2
02
Perc
ent
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.Muted, more homogeneous and less persistent response of interest payments
23 / 30
Borrowing: Asset vs. Earning-Based
∆Bi,t =G∑
g=1
β1,g ·I [Zi,t−1 ∈ g]·COLLi,t−1+G∑
g=1
β2,g ·I [Zi,t−1 ∈ g]·EBITDAi,t−1+X′i,tγ+εi,t
U.K. U.S.Young / No Div Old / Div Young / No Div Old / Div
COLLATERAL
0.025*** 0.012 0.063*** 0.038**(0.009) (0.009) (0.013) (0.014)
EBITDA
-0.013 0.069*** 0.007 0.048**(0.011) (0.019) (0.016) (0.018)
Dependent variable: ∆ long-term debt
Note: regressions include time-sector, group and firm fixed effects, plus a range of other lagged firms’ characteristics ascontrols. Standard errors are clustered by time and firm.
24 / 30
Borrowing: Asset vs. Earning-Based
∆Bi,t =G∑
g=1
β1,g ·I [Zi,t−1 ∈ g]·COLLi,t−1+G∑
g=1
β2,g ·I [Zi,t−1 ∈ g]·EBITDAi,t−1+X′i,tγ+εi,t
U.K. U.S.Young / No Div Old / Div Young / No Div Old / Div
COLLATERAL 0.025*** 0.012
0.063*** 0.038**
(0.009) (0.009)
(0.013) (0.014)
EBITDA -0.013 0.069***
0.007 0.048**
(0.011) (0.019)
(0.016) (0.018)
Dependent variable: ∆ long-term debt
Note: regressions include time-sector, group and firm fixed effects, plus a range of other lagged firms’ characteristics ascontrols. Standard errors are clustered by time and firm.
24 / 30
Borrowing: Asset vs. Earning-Based
∆Bi,t =G∑
g=1
β1,g ·I [Zi,t−1 ∈ g]·COLLi,t−1+G∑
g=1
β2,g ·I [Zi,t−1 ∈ g]·EBITDAi,t−1+X′i,tγ+εi,t
U.K. U.S.Young / No Div Old / Div Young / No Div Old / Div
COLLATERAL 0.025*** 0.012 0.063*** 0.038**(0.009) (0.009) (0.013) (0.014)
EBITDA -0.013 0.069*** 0.007 0.048**(0.011) (0.019) (0.016) (0.018)
Dependent variable: ∆ long-term debt
Note: regressions include time-sector, group and firm fixed effects, plus a range of other lagged firms’ characteristics ascontrols. Standard errors are clustered by time and firm.
24 / 30
EQUITY (MKT. VALUE) falls
Younger & NO dividends Older & Paying dividendsU
nite
dS
tate
s
-8-6
-4-2
02
Percent
1 4 7 10 13 16 19Quarters
-8-6
-4-2
02
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-8-4
-20
2Pe
rcen
t
1 2 3 4 5Year end (Q4)
-8-4
-20
2Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%. Share Price
25 / 30
Response of EARNINGS
Younger & NO dividends Older & Paying dividendsU
nite
dS
tate
s
-1.5
-1-.5
0.5
1Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
1Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-1.5
-1-.5
0.5
11.
5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-1.5
-1-.5
0.5
11.
5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%. EBITDA
26 / 30
Transmission Mechanism
To recap:I Net worth falls for all groups.
I Borrowing of younger-no dividend firms is more correlated withasset values than with earnings.
I Borrowing only significantly falls for these firms.
Other channels?
1. Demand2. Growth and profitability3. Liquidity4. Risk
27 / 30
Sensitivity analysis
Results are robust to
I survival bias
I information effect
I sectoral heterogeneity
I ending the sample in 2007
28 / 30
Our contribution: FIVE NEW FINDINGS...
1. Younger firms respond more than any other group and drive theaggregate response of investment to interest rate changes
2. Results are more pronounced for young firms paying nodividends and robust to controlling for other firm characteristics
3. Younger firms’ borrowing is more asset-based (thanearning-based)
4. Net worth and share prices move for all firms
5. Borrowing responds most for younger firms.
29 / 30
Our contribution: FIVE NEW FINDINGS...
1. Younger firms respond more than any other group and drive theaggregate response of investment to interest rate changes
2. Results are more pronounced for young firms paying nodividends and robust to controlling for other firm characteristics
3. Younger firms’ borrowing is more asset-based (thanearning-based)
4. Net worth and share prices move for all firms
5. Borrowing responds most for younger firms.
29 / 30
...and AN INTERPRETATION
I Younger firms tend to borrow against the value of their assets tofund capital expenditure.
I Rate increases push down asset prices and collateral values.
I Borrowing constraints tighten: borrowing and investment falls.
I Younger firms account for a sizable part of the aggregateresponse of investment.
Young firms face financial frictions. Fluctuations in collateraland asset values can play a key role in the MTM.
30 / 30
Extra Slides
31 / 30
Monetary Policy Surprises and Shocks
High-frequency Surprises Policy ShocksU
nite
dS
tate
s
-.3-.2
-.10
.1%
1990m1 1993m1 1996m1 1999m1 2002m1 2005m1 2008m1 2011m1 2014m1Date
-3-2
-10
12
3%
1987m1 1991m1 1995m1 1999m1 2003m1 2007m1 2011m1 2015m1Date
Uni
ted
Kin
gdom
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Date
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
1985 1990 1995 2000 2005 2010 2015
Date
-4
-3
-2
-1
0
1
2
3
Back
32 / 30
Investment: National Statistics vs Micro dataLevels Growth rates
Uni
ted
Sta
tes
66.
57
7.5
8R
eal I
nves
tmen
t (lo
g)
1985 1990 1995 2000 2005 2010 2015Date
National Statistics Aggregated Micro Data
-30
-20
-10
010
20%
-20
-10
010
%
1985 1990 1995 2000 2005 2010 2015Date
National Statistics Aggregated Micro Data
Uni
ted
Kin
gdom
8.5
99.
510
10.5
Rea
l Inv
estm
ent (
log)
1985q1 1990q1 1995q1 2000q1 2005q1 2010q1 2015q1Date
National Statistics Aggregated from Micro-Data
-20
020
40%
-20
-10
010
2030
%
1985q1 1990q1 1995q1 2000q1 2005q1 2010q1 2015q1Date
National Statistics Aggregated from Micro-Data
Back
33 / 30
The response of aggregate investment
United Kingdom United States
4 8 12 16
Quarters
-1.5
-1.25
-1
-0.75
-0.5
-0.25
0
0.25
0.5
Per
cent
dev
iatio
n
4 8 12 16
Quarters
-1.5
-1.25
-1
-0.75
-0.5
-0.25
0
0.25
0.5
Per
cent
dev
iatio
n
Monetary Policy shock: 25 basis point increase. Bootstrapped Standard errors.Back to average effect
34 / 30
The response of aggregate investment
United Kingdom United States
Inve
stm
ent
4 8 12 16
Quarters
-1.5
-1.25
-1
-0.75
-0.5
-0.25
0
0.25
0.5
Per
cent
dev
iatio
n
4 8 12 16
Quarters
-1.5
-1.25
-1
-0.75
-0.5
-0.25
0
0.25
0.5
Per
cent
dev
iatio
n
Inte
rest
rate
4 8 12 16
Quarters
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
Per
cent
dev
iatio
n
4 8 12 16
Quarters
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
Per
cent
dev
iatio
n
Monetary Policy shock: 25 basis point increase. Bootstrapped Standard errors.Back to average effect
35 / 30
The response of selected macro variables
Employment Credit Spread IP
Uni
ted
Sta
tes
4 8 12 16
Quarters
-0.5
-0.25
0
0.25
Per
cent
dev
iatio
n
12 24 36 48
Months
-0.2
-0.1
0
0.1
0.2
0.3
0.4
Per
cent
dev
iatio
n
12 24 36 48
Months
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
Per
cent
dev
iatio
n
Uni
ted
Kin
gdom
12 24 36 48
Months
-0.5
-0.25
0
0.25
Per
cent
dev
iatio
n
12 24 36 48
Months
-0.2
-0.1
0
0.1
0.2
0.3
0.4P
erce
nt d
evia
tion
12 24 36 48
Months
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
Per
cent
dev
iatio
n
Monetary Policy shock: 25 basis point increase. Bootstrapped Standard errors.Back to average effect
36 / 30
The U.S. average effect reported at annual frequency
United States United Kingdom
-1-.5
0.5
Perc
ent
1 2 3 4 5End year (Q4)
-.6-.4
-.20
.2.4
Perc
ent
0 1 2 3 4Year
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.Back to average effect
37 / 30
Investment responses by PAYING DIVIDENDS
NO dividends paid Dividends paidU
nite
dS
tate
s
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-2-1
-.50
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-2-1
-.50
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
38 / 30
Investment response by SIZE
Smaller Medium Larger
Uni
ted
Sta
tes
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
39 / 30
‘Controlling’ for (SMALLER) size
NO dividends & Younger PAY dividends & OlderU
nite
dS
tate
s
-2-1
01
23
Percent
1 4 7 10 13 16 19Quarters
-2-1
01
23
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.Back to robustness summary
40 / 30
Investment response by ASSET GROWTHFaster-growing Slower-growing
Uni
ted
Sta
tes
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
41 / 30
‘Controlling’ for (FASTER) asset growth
NO dividends & Younger PAY dividends & OlderU
nite
dS
tate
s
-2-1
0.5
Percent
1 4 7 10 13 16 19Quarters
-2-1
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.Back to robustness summary Back to mechanism
42 / 30
Investment response by LEVERAGE
Lower Medium Higher
Uni
ted
Sta
tes
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-4-2
-10
2Pe
rcen
t
1 2 3 4 5Year end (Q4)
-4-2
-10
2Pe
rcen
t
1 2 3 4 5Year end (Q4)
-4-2
-10
2Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
43 / 30
‘Controlling’ for (LOWER) leverage
NO dividends & Younger PAY dividends & OlderU
nite
dS
tate
s
-2-1
0.5
Percent
1 4 7 10 13 16 19Quarters
-2-1
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.Back to robustness summary
44 / 30
Investment response by LIQUIDITY
Lower Medium Higher
Uni
ted
Sta
tes
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-.50
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-.50
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-.50
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
45 / 30
‘Controlling’ for (HIGHER) liquidity
NO dividends & Younger PAY dividends & OlderU
nite
dS
tate
s
-2-1
0.5
Percent
1 4 7 10 13 16 19Quarters
-2-1
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.Back to robustness summary Back to mechanism
46 / 30
Investment response by TOBIN’S QHigher Lower
Uni
ted
Sta
tes
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
-1-.5
0.5
Perc
ent
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
47 / 30
‘Controlling’ for (HIGHER) Tobin’s Q
NO dividends & Younger PAY dividends & OlderU
nite
dS
tate
s
-2-1
0.5
Percent
1 4 7 10 13 16 19Quarters
-2-1
0.5
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-4-2
-10
.5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.Back to robustness summary Back to mechanism
48 / 30
US Investment Response by BETA and ALPHA
Low Medium High
Bet
a
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
Alp
ha
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
49 / 30
‘Controlling’ for (HIGH) Alpha/Beta (US)
NO dividends & Younger PAY dividends & Older
Bet
aA
lpha
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.Back to robustness summary Back to mechanism
50 / 30
Firms Who Grow Old
Young & NO dividends Old & Paying dividends
Uni
ted
Sta
tes
-1.5
-1-.5
0.5
Percent
1 4 7 10 13 16 19Quarters
-1-.5
0.5
11.5
Percent
1 4 7 10 13 16 19Quarters
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.Back to robustness summary Back to mechanism
51 / 30
More homogeneous INTEREST PAYMENTS response
Younger & NO dividends Older & Paying dividendsU
nite
dS
tate
s
-4-2
02
4Percent
1 4 7 10 13 16 19Quarters
-4-2
02
4Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-10
-5-2
02
5Pe
rcen
t
1 2 3 4 5Year end (Q4)
-10
-5-2
02
5Pe
rcen
t
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%.
Back to Borrowing
52 / 30
SHARE PRICE falls
Younger & NO dividends Older & Paying dividendsU
nite
dS
tate
s
-8-6
-4-2
02
Percent
1 4 7 10 13 16 19Quarters
-8-6
-4-2
02
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-8-4
-20
1P
erce
nt
1 2 3 4 5Years
-8-4
-20
1P
erce
nt
1 2 3 4 5Years
Monetary Policy shock: 25 basis point increase. Standard errors clustering: by firms and time. Confidence band: 90%.Back
53 / 30
Response of EBITDA
Younger & NO dividends Older & Paying dividendsU
nite
dS
tate
s
-3-2
-10
12
Percent
1 4 7 10 13 16 19Quarters
-3-2
-10
12
Percent
1 4 7 10 13 16 19Quarters
Uni
ted
Kin
gdom
-4-2
-1-.5
02
Perc
ent
1 2 3 4 5Year end (Q4)
-4-2
-1-.5
02
Perc
ent
1 2 3 4 5Year end (Q4)
25 basis point increase in interest rates. Standard errors clustering by firm and time. Confidence band: 90%. Back
54 / 30