Chapter 3: Research Design and Methodology
3.2. Sample size
The 167 NZX listed equities were obtained in September 2017 from the main board of the New Zealand stock exchange. Companies that did not report operating leases were excluded from the study. In addition, adopting the approach of Xu et al. (2017) and the IASB, banks, financial
companies, funds and insurance firms were excluded from the sample because of the nature of their operations and industry regulations. The standard-setter in its Effects Analysis for IFRS 16 excluded banks and insurance companies because of the disproportionate size of their balance sheets compared with other firms. These exclusions result in 88 NZX listed companies that reported long- term operating leases in their financial statements for 2017. However, an additional 12 exclusions had to be made when the sampled firms did not report the following elements:
an appropriate interest rate that could be used as a discount rate for the future lease payments;
Interest expense; and
31 The latter two elements are required for calculating the discount rate, as suggested by Imhoff et al. (1997) and Xu et al. (2017), with the formula as follows:
(3.1) Discount rate = Interest expense/Interest-bearing debt
However, given that NZ IFRS 16 provides the firms an exemption for capitalization of short-term leases22, to use the above formula we need to obtain a non-current Interest-bearing debt. Our
analysis shows a number of New Zealand listed companies lack such an element on their balance sheets, hence the 12 exclusions included in our initial sample. Seven of these twelve companies did not report any non-current liabilities and the rest did not have the Interest-bearing elements required to enable us to use the formula (3.1). For example, Moa Group Limited (MOA), a brewing company, did not report any Interest expense for the period reviewed. Instead, it reported a net Finance income of $ 7 000, and a Gross profit of over $3 million. In addition, MOA’s balance sheet has only two elements under Non-current liabilities, Employee benefits and Deferred tax liabilities. Based on the financial report of this company we conclude that, in 2017, MOA financed its
operations and off-set its Loss for the year with equity financing instead of taking loans, because this entity issued new shares worth over $4 million. Similarly, Pacific Edge Limited (PEB), a cancer diagnostics company, did not report any Interest expense but showed an Interest income of $248 601. In addition, this entity did not report any Non-current liabilities for the year. Finally, PEB discloses that the company had no borrowing costs for the year ended 31 March 2017 and also for the prior reporting period. Like MOA, PEB used equity financing in the reporting period. A good illustration of the exclusions to our initial sample is Briscoe Group Limited (BGP), a well-known retailer. This company reported Interest expense for the review period, but the company disclosed in Note 5.1 in its annual report that it mainly uses borrowings that fall under the element of Current liabilities. For the financial year 2017, BGP reported only one element of balance sheet under Non- current liabilities, Trade and other payables. Although the company discloses that the interest payable for the used short-term borrowings it uses is based on a “BKBM rate”, it did not explained how this rate is obtained. In addition, BGP mentions an unknown margin added to the base “BKBM” rate. Such a lack of numeric information and the short-term character of the borrowings seems to be ineadequate for a reliable establishment of the discount rate required for a prudent capitalization procedure. Such a capitalization procedure is needed given that, according to BGP’s financial statements, its operating leases last beyond year 5. To complete exemplifying the exclusions to our initial sample, Hallenstein Glasson Holdings Ltd (HLG), another well-known retailer in New Zealand. For the 2017 financial year, this company did not report any Non-current liabilities. In addition, HLG
22 Short-term leases are defined as “a lease that, at the commencement date, has a lease term of 12 months or less” (NZ IFRS 16, Appendix A).
32 did not report any Interest expense, but showed Finance income for the year. Therefore, as can be seen from the above examples, there are New Zealand listed firms that did not disclose elements of financial statements that would allow us to use the formula (3.1) and calculate the required
capitalization rate. As our intention is to avoid unsubstantiated sample-wide assumptions, these companies were therefore excluded from the final sample in the study. Table 3-1 details the final sample.
Table 3-1 The composition of the final sample for the study (compiled by author)
Category Number of Equities on the Main
Board of NZX
Initial sample size 167
Multiple equities of the same firm -8
Annual report not available -1
Banks, insurance companies,
financial companies or funds. -65
Newly listed company (revealed and screened during additional data check in August 2018)
+1
No operating leases reported by the
companies -6
No interest rate, interest expenses and interest-bearing debts information reported
-12
Final sample size 76
The study includes a sample of 45% of the NZX listed companies. The de facto size of the sample relative to the total population is larger because the Main Board of NZX shows listings for equities, not companies. According to Roscoe (1975), a successful study with experimental controls, such as matched pairs, can be conducted with a sample of 10-20 subjects. However, Roscoe (1975)
recommends a sample size of 30 or more for ex-post facto research and for most experimental research. Gay & Diehl (1992) consider a sample of 10% of the population as the bare minimum or, if the population is small, 20% or more of the population is required for a descriptive study. A sample of at least 30 subjects is needed to establish a relationship in correlations study. For a causal- comparative and experimental study, the desired minimum is 30 subjects per group (Gay & Diehl, 1992). Quinlan (2011) suggests for a small population under 1000, a study would require a sample of about 30 %. Finally, a small sample is deemed appropriate for a small economy, when it is not
33 economically feasible to collect or analyse a large data sample (Isaac & Michael, 1995). From these perspectives, the sample of 76 listed firms from a population of 168 listed equities is considered sufficient for our study. In addition, compared with previous research on New Zealand, the sample size of this study is significantly larger than the 38 NZX listed firms used by Bennett & Bradbury (2003).