18
The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University of Manitoba Janet Morrill, University of Manitoba May 2005

The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Embed Size (px)

Citation preview

Page 1: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

The Incremental Value Relevance of Geographic

Segment Disclosures: Canadian Evidence

Roger Graham, Oregon State University

Cameron Morrill, University of Manitoba

Janet Morrill, University of Manitoba

May 2005

Page 2: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Presentation Agenda

1. Introduction

2. Literatureo Segmented Disclosureso Quebec environment and Political Risk

3. Research model

4. Data and Results

5. Conclusions

Page 3: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Segmented Disclosures

• Objective of segmented disclosures is to help users:– Better understand the enterprise’s performance– Better assess future cash flows– Make more informed judgments about the

enterprise as a whole

• “reportable segments” can be either geographic or by line of business

Page 4: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Changes to Canadian Section 1701(similar to SFAS 131)

• Reporting consistent with internal reporting

• Less information by geographic segment

• Geographic segment required for country of domicile and foreign countries combined, and on a country-by- country basis if material

• Is information by geographic segment important?

• Can there be important information at a level other than by country?

Page 5: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Usefulness of geographic segment disclosure (old rules)

• Weak evidence of market reaction to unexpected profits by geographic segmentBoatsman et al, 1993

• Over long windows, link between security returns and geographic segment earnings differs by segmentThomas, 2000

• No evidence of geographic usefulness under new rules

Page 6: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Usefulness of LOB disclosure

• Survey respondents perceive segmented information usefulBaldwin, 1984; Berger et al, 2003; Lobo et al 1998

• Link between market values and financial information is stronger when using segmented information instead of aggregated informationTse, 1989; Givoly et al, 1999

• Incremental content of l.o.b. is small but significant. Significance increases with differential profit and growth opportunitiesChen and Zhang, 2003

Page 7: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Our approach

To test whether information on extent of operations in Quebec (from a non-financial-statement source) is value relevant

Implications:1. Market is efficient with respect to other public

information2. Geographic segment information is value

relevant3. Information at a sub-national level can be

relevant.

Page 8: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Quebec’s Business Environment

• Political Uncertainty– Possibility of secession– Possible negative consequences

• Bill 101

• Quebec Civil Code

• Restrictions on Corporate Takeovers

Page 9: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Potential economic consequences of Quebec sovereignty: The market’s assessment

• Firms moving out of Quebec earn positive abnormal returns (Tirtiroglu, Bhabra and Lel, 2004)

• Quebec-headquartered firms trade at a discount (Graham, Morrill and Morrill, 2005)

• Quebec provincial bond yields increase as support for Quebec sovereignty increases (Johnson and McIlwraith, 1998)

Page 10: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Market – book model (GMM 2005)Basic model: MVEit+2 = ß1BVEit + ß2Eit + eit

MVEit+2 = t + t*Qit + ß1BVEit + ß2Q*BVE

+ ß3Eit + ß4Q*Eit + eit

MVE = market value of equity

Q = 1 for Quebec firm; 0 for non-Quebec

BVE = book value of equity

E = earnings from continuing operations = controls for size, growth, leverage,year

Results: ß1, ß3 > 0; ß2, ß4< 0

Page 11: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Our market – book modelMVEit+2 = t + t*Qit + ß1BVEit + ß2Q*BVE + ß3Eit

+ ß4Q*Eit + ß5PCTit + ß6Q*PCTit

+ ß7PCTit*BVEit + ß8PCTit*Eit + eit

MVE = market value of equity

Q = 1 for Quebec firm; 0 for non-Quebec

BVE = book value of equity

E = earnings from continuing operations

PCT = extent of firm operations in Quebec

= controls for size, growth, leverage,year

Page 12: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Data• Firms with accounting and market data

available in Compustat TSE data file

• Firms with Quebec segment information (employees, sales, assets) in Les Affaires

• Deleted firms with– Insufficient data– Negative book values– Extreme values (top and bottom 2%) of

earnings and sales growth

Page 13: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Descriptive statistics(* = QC vs. non-QC different, p<.05)

QC firms (82) Non-QC firms (51)

N Mean N Mean

QC employees* 590 54.5% 418 27.2%

QC sales* 211 58.3% 156 35.2%

QC assets* 250 67.4% 135 28.0%

EPS 590 $0.80 418 $0.77

BV per share 590 $9.17 418 $10.27

MV per share* 588 $11.64 417 $16.73

Page 14: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Market Value Regression ResultsBVE = book value of equity; E = earnings; Q = Quebec

dummy; PCT = % employees in QC.*(**) significant at p < 0.05 (p<0.01)

Ind. variable Exp. Model 1 Model 2 Model 3

BVE + 1.44** 1.30** 1.50**

E + 2.02** 1.75** 1.66**

Q*BVE - -0.20* 0.07

Q*E - -0.06 -1.05**

PCT*BVE - -0.36* -0.99**

PCT*E - 0.95* 3.16**

Adj. R2 0.85 0.85 0.86

Page 15: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Tests of incremental explanatory power

Effect of adding: F(df numerator, df denominator)

[probability]

PCT variables to location-only model

F(6,907) = 4.58 [ p<0.001 ]

Location variables to PCT-only model

F(5,907) = 7.77 [ p<0.001 ]

Page 16: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Conclusion• Firms located in Quebec have lower market

multiples on book value an earnings leading to lower firm values (we knew that already)

• Segment disclosure, even voluntarily disclosed and unaudited, seems to be value-relevant in addition to location of firm headquarters

Page 17: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Disclosure ImplicationsExpanded geographical disclosures could

provide relevant information• Bringing geographic information into realm

of financial statements allows it to be subjected to audit

• Re-visit decision to provide geographic information by country only– Allowing discretion or specifying conditions

could increase relevance– Danger is threat to reliability from “strategic

aggregation or disaggregation”

Page 18: The Incremental Value Relevance of Geographic Segment Disclosures: Canadian Evidence Roger Graham, Oregon State University Cameron Morrill, University

Paper is available at:http://www.umanitoba.ca/asper/faculty/cam.morrill/