AcF606 Lent Term 2020


Workshop 5: Valuation by the multiple method: Large sample
analysis

Due: Case Workshop 5

Submission of preparatory work: One mark per submission (up to the maximum 4
marks), provided that you (i) submit the work on Moodle before the start of the earlier
workshop of the week (even if you are assigned to a later one); and (ii) have made
genuine effort (regardless whether you get correct answers or not).

• If you have submitted the four cases already, you have reached the maximum of 4
marks, and this submission would not earn you extra marks.
• If you do not feel proficient in programming STATA at this moment, you may describe
what each set of codes is meant to achieve.

For this workshop, please use the STATA codes and datasets posted on Moodle.

Requirements:

1. Selectively read the paper “Who is My Peer? A Valuation-Based Approach to the
Selection of Comparable Firms” by Bhojraj and Lee (2002). The link (also on the
module reading list) is: https://doi-org.ezproxy.lancs.ac.uk/10.1111/1475-
679X.00054
Using your own, nontechnical words, briefly answer:
(i) What is the paper’s research question? Why is it important?
(ii) What is the main idea for the proposed “warranted multiple” method?
(iii) What are the main steps of the research design?

2. The STATA do-file “CW5_Q2.do” contains the programming codes. Run them with
the dataset “CW5_Q2.dta”. Briefly discuss:
(i) What are the main challenges in operationalizing the research design steps
outlined in Question 1?
(ii) How do you interpret the results?

3. Use one of the discounted models from Lectures 8-9 to derive the theoretical
determinants of the multiple EV/EBITDA1 (i.e., one-year-ahead forecasted
EBITDA1).

Hint: For this question, you may treat EBITDA as if it were free cash flow or operating
income.

4. Describe the research design to test the effectiveness of selecting comparable firms
based on the “warranted EV/EBITDA1” (“WEVE1”). The benchmark is the method
of selecting comparables from the same industry.

5. The dataset “CW5_Q5.dta” contains the following variables:

AcF606 Lent Term 2020


Variable Explanation (same as those in the original paper, except EVE1 and IndEVE1)
GVKEY Firm ID
FYEAR Fiscal year, time ID
SIC2 Industry ID
EVE1 The multiple of EV/EBITDA1, for individual firms
IndEVE1 Harmonic mean of EVE1 in an industry (similar to IndEVS in the original
paper)
IndPB Industry harmonic mean of PB
AdjPM Difference between the firm’s and the industry’s operating profit margins
LossPM AdjPM times an indicator variable, where the indicator variable is 1 if profit
margin≤0 and 0 otherwise
AdjGro Difference between analyst consensus forecast of the firm’s long-term growth
and the industry average
Lev Total long-term debt scaled by book value of ordinary equity
RNOA Operating profit scaled by net operating assets
RD The firm’s R&D expressed as a percentage of net sales.

Use this dataset to execute the research design in the preceding question.

Hint: The dataset has four years, so you will implement the warranted multiple method
three times (each time utilizes two year’s data). Make necessary changes to the
STATA codes from Question 2, which in its current form doesn’t allow for
multiple-year implementations.