Effects of Managerial Incentive structure on the Pricing of Syndicated

Discussion of “Do Creditors Prefer Smooth Earnings?
Evidence from European Private Firms”
(Gassen and Fülbier)
2014 JIAR Conference – HK Poly
Ole-Kristian Hope
Overall Assessment
 Interesting study
 Nice setting
 Some room for further improvement
Why Care about Private Firms?
 In aggregate, 4 × employees; 3 × revenues; 2 ×
assets than public firms
 Drivers of economic growth
 > 99% of limited liability companies are not
listed
 Many interesting issues but much less researched
(various developments in private-firm accounting around
the world)
My Own Private-Firm Studies

“Financial Reporting Quality in U.S. Private and Public Firms.”
2013. The Accounting Review

“Agency Conflicts and Auditing in Private Firms.” 2012.
Accounting, Organizations, and Society

“Financial Credibility, Ownership, and Financing Constraints in
Private Firms.” 2011. Journal of International Business Studies

“Financial Reporting Quality and Investment Efficiency of Private
Firms in Emerging Markets.” 2011. The Accounting Review

“Auditor Independence in a Private Firm and Low Litigation
Setting.” 2010. The Accounting Review

New WP: “Stakeholder Demand for Accounting Quality and
Economic Usefulness of Accounting in U.S. Private Firms.” With
Thomas and Vyas
Potential to Strengthen Motivation
 Private-firm setting has potential to outcompete public-firm
setting when it comes to examining debt-contracting role of
smoothness
 Usefulness of accounting includes valuation and contracting
(etc.). Valuation role likely reduced for private firms
 Contracting role of accounting mainly studied for (1) executive
compensation; (2) political costs; and (3) debt covenants (WZ
1990)
 Small, often owner-managed firms  Political costs likely less
important. Compensation possibly less important
 Relatively clean setting to test debt-contracting role
Role of Income Smoothing
 Unclear in literature. Accrual accounting by def.
induces negative correlation between CF and
accruals
 In most prior research, smoothing viewed as
“bad” (EM)
 Here: Authors are agnostic
 “Informational” vs. “garbling” role of smoothing
(Dou, Hope, and Thomas 2013; Tucker and Zarowin 2006)
Endogeneity Issues
 Use residuals from model in Appendix to deal with
endogeneity of creditor financing / capital structure
 Explain better source of endogeneity. For example, firms with
smoother earnings have lower operating risk and therefore
higher optimal financial leverage. If A-2 does not fully control
for operating risk, endogeneity concern remains. But do
control for CV volatility
 Why rely on Michaelas, Chittenden, and Poutziouris (1999;
Small Business Economics) when well-established models exist
(e.g., Harris and Raviv 1991)?
 Consider changes model? Currently throw away a lot of info.
Consider defining SMOOTH as std dev of NI/CF last four
years?
 Alternatively look for IV or exogenous shock to leverage
Semantics: “Exogenous Smoothness”?
 Clearly, authors do not mean “exogenous,” and this
language has to go
 Use residuals from primary regression

[Alternative: Include controls in primary regression (or look for
IV)]
Strange Sample?
 Sample is highly unbalanced across 24 countries
 52.7% of firms from France and Spain!
 Basically no German or UK firms?
 Belgium 16% vs. Netherlands 5.6%?
 Why rely solely on Amadeus if want representative European
data? FAME for the UK?
 Effects for tests; inferences? (Need strong caveats in paper)

State: “most of our tests are based on country-industry level
variables”
Sample (continued)
 Many tiny firms (or should I say ?). Are these the firms
you want in your sample?
tiny
 Extreme variation in firm size  Need much stronger
size controls! (See my AOS paper?)
 Several statistics strongly dependent on including a
huge number of “mom & pop stores” (e.g., # obs. with
“dispersed ownership”)
 Make use of such variation in tests? (Wouldn’t expect
same relation for all?)
Analyses - H1
 Why is INT_COVERAGE used to test H1? Not
explained. Say negative coefficient is “consistent with
predictions” but such predictions are not developed
 Do you want an earnings-free measure of
INT_COVERAGE?
 “Trade credit stronger relation than other sources of
finance.” Based on? Consider standardized coefficients
if want to make such statements?
Analyses – H2
 Table 5 confusing. H2 suggests interested in interaction
of LEVERAGE and country-level variables
 To test H2, estimate (3) independently for each country
and leverage quintile. Why? Not explained
 Then calculate difference between highest and lowest
leverage quintile. Again, why? I presume no monotonic
relation?
 “DEBT_EFFICIENCY” only tested without controls. Why?

“To address the prominent problem of governance variables.” What
does this refer to?
Some Other Suggestions/Comments
 Rearrange tables so test variables first. More
important, discuss primary findings in more detail!
 Is there no intercept?
 Consider:

Toning down statements and adding more caveats due to pure
XS design

“net earnings primarily serving as a basis for distributable
income. This is an assertion only by Ball, Kothari, and Robin
(2000)

“like”  “such as”
 “where”  “in which” (or “for which”)
 Writing out author names first time you cite a paper
Concluding Remarks
Interesting study; nice setting
Some need for improvement (as is usually the case)
Best of luck with paper!