Summary Data Comments Model Comments Bottom Line
Firm Financing over the Business Cycle
Juliane Begenau and Juliana Salomao
discussion by Toni Whited
NBER Summer Institute
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
In the data, large and small firms finance differently over
the business cycle.
I
I
Small:
I
Raise both debt and equity more in booms than recessions.
I
Do not pay out much to shareholders ever.
Large:
I
Raise more debt relative to equity in booms than in recessions.
I
Payout is procyclical.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
The paper builds a model of ex-ante homogeneous firms to
understand these facts.
I
The model can be described as Hennessy and Whited (2007) meets
Hopenhayn (1992).
I
Dynamic industry equilibrium model with both immature and
mature firms because of endogenous entry and exit.
I
Firms use capital to produce output and finance with risky net debt,
costly external equity, and internal funds.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
The intuition can be seen with a stylized policy function
debt
net payout
capital
0
0
size
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
The intuition can be seen with a stylized policy function
debt
net payout
capital
high shock net payout
high shock capital
high shock debt
0
0
size
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
The presence of both mature and immature firms is
important!
I
Small firms are below an optimal scale and always want to invest
more in good times.
I
They exhaust debt capacity and use some equity finance.
I
Large firms hover around an optimal size.
I
Optimal investments are smaller, so they largely finance with debt in
good times.
I
Small firms rarely distribute to shareholders, and large firms pay out
more in good times.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
Outline
I
Challenge the facts.
I
Explain how the model can reconcile these new facts.
I
Extra comments depending on how fast I talk.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
Compustat equity issuance data are terrible!
I
The numbers include what we want to model: seasoned equity
offerings and private placements.
I
The numbers are dominated by what we do not want to model:
option exercise.
I
This problem matters more for large than for small firms.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
SEOs are close to Compustat “equity issuance” for small
firms.
0.020
Compustat Issuance
SEOs
0.018
0.016
0.014
0.012
0.010
0.008
0.006
0.004
0.002
0
1990
1995
2000
Discussion
2005
Firm Financing over the Business Cycle
2010
Summary Data Comments Model Comments Bottom Line
SEOs are nowhere near Compustat “equity issuance” for
large firms.
0.010
Compustat Issuance
SEOs
0.009
0.008
0.007
0.006
0.005
0.004
0.003
0.002
0.001
0
1990
1995
2000
Discussion
2005
Firm Financing over the Business Cycle
2010
Summary Data Comments Model Comments Bottom Line
SEO Proceeds: No cyclicality for large and slight
cyclicality for small firms.
0.020
Large
Small
0.018
0.016
SEO Proceeds
0.014
0.012
0.010
0.008
0.006
0.004
0.002
0
1990
1995
Discussion
2000
2005
Firm Financing over the Business Cycle
2010
Summary Data Comments Model Comments Bottom Line
SEO Incidence: Strong cyclicality for both large and
small firms
Large
Small
0.025
SEO Incidence
0.020
0.015
0.010
0.005
0
1990
1995
Discussion
2000
2005
Firm Financing over the Business Cycle
2010
Summary Data Comments Model Comments Bottom Line
The SEO facts and the Compustat facts are roughly the
same for small firms.
I
Phwew!
I
No such luck for large firms.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
Much of the movement in net payout is from the payout
side.
Total Payout
0.030
Large
Small
0.025
0.020
0.015
0.010
0.005
0
1990
1995
2000
Discussion
2005
Firm Financing over the Business Cycle
2010
Summary Data Comments Model Comments Bottom Line
How can a model of ex-ante homogeneous firms give us
acyclical proceeds but cyclical incidence for large firms?
I
We need more frequent, smaller investment projects in booms than
in recessions.
I
Why? Financing is driven by the budget constraint!
I
One possibility is countercyclical irreversibility.
I
If investment is harder to reverse in recessions, optimal projects will
be larger when they happen, but they will happen less frequently.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
The paper claims that large firms issue debt to finance
payout in booms
I
In the data, this type of activity is rare.
I
I suspect it also is in the model.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
0.20
0.016
0.15
0.012
0.10
0.008
0.05
0.004
0
0
1985
1990
Large payout
1995
2000
Large debt issuance
Discussion
2005
Refinancing
Firm Financing over the Business Cycle
2010
Large payout and refinancing
Large debt issuance
Danis, Rettl and Whited (2014) show that refinancing
activity is rare.
Summary Data Comments Model Comments Bottom Line
Refinancing is nearly nonexistent in dynamic
investment-finance models.
I
I solved and simulated the Hennessy and Whited (2007) model and
found no observations in which a firm did a large payout and a large
concurrent debt issuance.
I
Why?
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
Refinancing is nearly nonexistent in dynamic
investment-finance models.
I
Firms optimally want to preserve debt capacity.
I
The value of this free debt capacity is more valuable to shareholders
in nearly all states of the world than the value of taxable dividends.
I
We can observe concurrent debt issuance and dividend increases, but
no refinancings, per se.
I
A positive shock increases cash flow, optimal investment
expenditures, and optimal payout. Distress costs fall, and debt rises,
but not to finance payouts, per se.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
Why .55?
I
Returns to scale are very decreasing.
I
This number comes from a model with a convex cost of dividends
and no adjustment costs (Hennessy and Whited 2007).
I
Leverage in the model is way too high (.38)
I
Investment policy is way too conservative.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
Other stuff
I
“This generates an endogenous debt limit that becomes binding
when a firm’s funding needs exceed its debt funding cost.”
This makes no sense. A comparison of prices and quantities. In
these endogenous default models, funding costs are basically the
risk-free rate up to a limit, and then they skyrocket. So there is
something “like” a debt limit but not really.
I
Get rid of all of all mention of collateral. There is no collateral
constraint in the model. Firms can borrow more than their capital
stocks.
I
“Large firms on the other hand find equity financing too costly.”
No. The difference in financing policies comes from the differences
in investment policies.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
This is a hard paper to discuss.
I
It is asking an interesting and important question.
I
It is answering the question in careful and sensible way.
I
I think that moving the focus to active equity issuances and away
from “accounting” net payout would make a great deal of sense.
Discussion
Firm Financing over the Business Cycle
Summary Data Comments Model Comments Bottom Line
Danis, A., Rettl, D., Whited, T.M., 2014. Refinancing, profitability and
capital structure. Forthcoming, Journal of Financial Economics.
Hennessy, C.A., Whited, T.M., 2007. How costly is external financing?
Evidence from a structural estimation. Journal of Finance 62,
1705–1745.
Hopenhayn, H.A., 1992. Entry, exit, and rm dynamics in long run
equilibrium. Econometrica 60, 1127–1150.
Discussion
Firm Financing over the Business Cycle