Effect of Merger on the Credit Rating and Performance of Taiwan Security Firms

Abstract

The effect of a merger on credit rating is investigated by testing the significance of change in a firm’s rank based on comprehensive performance score and synergistic gains. We extract principle component factors from a set of financial ratios. Percentage of variability explained and factor loadings are adjusted to get a modified average weight for each financial ratio. This weight is multiplied by the standardized Z value of the variable, and summed a set of variables to get a firm’s performance score. Performance scores are used to rank the firm. Statistical significance of difference in pre- and post-merger rank is tested using the Wilcoxon sign rank (double end).

We studied the merger of financial firms after the enactment of Taiwan’s Merger Law for Financial Institution in November 2000 to examine synergies produced by merger. Synergistic gains affect corporate credit ratings. After taking into account the large Taiwan market decline from 1999 to 2000, test results show there is no significant operating, market, and financial synergy produced by the merger firms. Most likely explanations for the insignificant rank changes are short observation period and the lack of an adequate sample in this investigation.

We identify and define variables for merger synergy analysis followed by principal component factor analysis, variability percentage adjustment, and performance score calculation. Finally, Wilcoxon sign rank test is used for hypothesis testing. Reader is well referred to the appendix for details.

Appendices

Appendix 1: Variables for Merger Synergy Analysis

Post-merger credit rating of the firm depends on the extent of synergy produced by the merger. Components of merger synergy are operating synergy, market synergy, and financial synergy.

Each synergy component is determined by firm characteristics: financial structure, solvency, asset utilization, profitability, cash flow, growth, scale, and industry-specific ratio. In this study a number of financial ratios are used to assess operating performance of the firm before and after the merger. The following section outlines firm characteristics and variables to measure operating performance.

21.1.1 Merger Synergy

Operating synergy – it refers to the improvement of operating efficiency achieved via scale economy, transaction cost economy, and differential efficiency caused by merger. Market synergy – increase in market share due to enhanced negotiating power and dominant pricing strategy.

Financial synergy – diversification of financial risk and cost of capital reduction.

Operating Synergy

• A. – Financial structure

  • X_i1 – Debt ratio

  • X_i2 – Ratio of long-term capital to fixed assets

• B. Solvency

  • X_i3– Current ratio

  • X_i4– Quick ratio

• C. Asset utilization

  • X_i5 – Operating return on assets

  • X_i6 – Net worth turnover ratio

• D. Profitability

  • X_i7 – Return on assets

  • X_i8 – Return on equity

  • X_i9 – Profit margin

  • X_i10 – Earnings per share

• E. Cash flow

  • X_i11 – Cash flow to short-term liability

  • X_i12 – 5-year cash flow to debt obligations

  • X_i13 – Retention ratio

• F. Growth

  • X_i14 – Growth in revenue

  • X_i15 – Growth in earnings

• G. Size

  • X_i16 – Total assets

  • X_i17 – Net worth

• H. Industry-specific ratios

  • X_i18 – Consignment to current assets

  • X_i19 – Long-term financing to net worth

Market Synergy

• X_i5 – Operating return on assets turnover

• X_i20 – Market share variability

Financial Synergy

• X_i21 – Operating risk variability

Appendix 2

21.2.1 Principal Component Factor Analysis of Merger Synergies

Principal component factor analysis is used to examine merger synergies. It is a multivariable statistic method focusing on the relationship between groups of variables. Its purpose is to express the original data structure with fewer factors while keeping most of the information provided by the original data structure.

Factor analysis is composed of two parts: one is common factor, and the other is specific factor. Factor analysis intends to group the variables with same common factors. In other words, it discusses how to break down every variable X_i of P variables X₁∼X_P into the linear combination of q common factors fj(q, and q≤p), j = 1, 2 …q and specific factor ε_i. The model is as follows:

$$ \begin{array}{c}{X}_1={\mu}_1+{L}_{11}{F}_1+{L}_{12}{F}_2+\cdots +{L}_{1q}{F}_q+{\varepsilon}_1\\ {}{X}_2={\mu}_2+{L}_{21}{F}_1+{L}_{22}{F}_2+\cdots +{L}_{2q}{F}_q+{\varepsilon}_2\\ {}\vdots \\ {}{X}_p={\mu}_p+{L}_{p1}{F}_1+{L}_{p2}{F}_2+\cdots +{L}_{pq}{F}_q+{\varepsilon}_p\end{array} $$

(21.1)

where F₁ …, F_q are common factors, ε_i is specific factor for variable X_i, and L_ij is the factor loading of variable X_i and common factor F_j. Factors extracted are independent and also analysis preserves the information in the original variables. There is no overlap of information among principle components. The model should be parsimonious, in the sense that a few principle components should be able to replace the original set of variables.

21.2.2 Variability Percentage Adjustment

All principle components with factor loading greater than 1 are selected. Then weights are assigned to different variables based on percentage of variability explained. The variability percentage adjustment is adjusted as follows:

$$ {V}_j^{adj}={\mathrm{V}}_{\mathrm{j}}/{\varSigma}_j{V}_j $$

(21.2)

where V_j (j = 1,2…q) is the percentage of variability explained by factor F_j.

21.2.3 Factor Loading Adjustment

The adjusted variability percentage V ^adj _j is multiplied by factor loading L_ij and summed over all variables (j = 1,2…q) to get the total loading fo variable X_i. Then adjust L_i such that total factor loadings add to 100.

$$ \begin{array}{l}{L}_i=\sum {V}_j^{adj}{L}_{ij}\\ {}{L}_i^{adj}={100}^{\ast}\left({\mathrm{L}}_{\mathrm{i}}/\sum {\mathrm{L}}_{\mathrm{i}}\right)\end{array} $$

(21.3)

Every variable has a weight, L ^adj _i before and after the estimation period; we average it to obtain weight W_i. For some variables, X_i, the greater the value of the variable, the better it is for the operating, financial, or merger synergy. Those variables are classified as positive variables. If opposite is true, then those variables are classified as negative variable. Hence adjusted variable weights are redesigned to correctly reflect operating synergy score: W_i ^* = {W_i or –W_i} for positive and negative variables, respectively.

21.2.4 Performance Scores

All variables are not measured in the same unit, so they are standardized as X_i ^∗ = {X_i–Ave(X_i)}/σ, where Ave(X_i) is the average and σ is the standard deviation of variable X_i (i = 1,2…p). The adjusted variable weight W_i ^* multiplied by standardized variable Z_i gives the performance score of variable X_i. Appropriate standardized performance score of the variable is used to rank the firm for its operating or financial or market synergy. The sum of the standardized performance scores over the set of variables gives the comprehensive performance score:

$$ S=\sum {W_i}^{*.}{Z}_i $$

(21.4)

where the summation is over the first 19 variables listed in Appendix 1. Then firms are ranked in terms of their respective comprehensive performance score. The greater the total score is, the better is the comprehensive performance rating. On the other hand, the smaller the total score, the worse is the performance rating and lower is its rank.

Appendix 3

21.3.1 Wilcoxon Sign Rank Test

Usual “t-test” method is unsuitable to examine the significance of change in operating, financial, or market performance before and after merger as we deal with ranks. We use the Wilcoxon sign rank test which checks whether two sets of ranks, pre- and post-merger, come from the same sample or two samples. Let D be the difference between observed value from the sample and reference value. Then we delete those observations whose value of D is zero and rank incrementally the rest of observations in terms of the absolute value of D. If two or more absolute values are the same, we give each value an appropriate rank, then average those ranks, and use the averaged rank as the rank of the same absolute values.

The statistic analysis method of test is as follows:

The differences between the post-merger synergy ranks (X_i) and corresponding premerger synergy ranks (Y_i), D_i is ranked in descending order. Let R_i be the serial number of D_i (if they have the same rank, then take their average value).

$$ \begin{array}{l}\begin{array}{cc}\hfill {\boldsymbol{D}}_{\boldsymbol{i}}={\boldsymbol{X}}_{\boldsymbol{i}}-{\boldsymbol{Y}}_{\boldsymbol{i}}\hfill & \hfill \boldsymbol{i}=\mathbf{1},\mathbf{2},\cdots \boldsymbol{n}\hfill \\ {}\hfill {\boldsymbol{R}}_{\boldsymbol{i}}=\boldsymbol{rank}\left(\left|{\boldsymbol{D}}_{\boldsymbol{i}}\right|\right)\hfill & \hfill \boldsymbol{i}=\mathbf{1},\mathbf{2},\cdots \boldsymbol{n}\hfill \end{array}\\ {}\begin{array}{cc}\hfill \boldsymbol{W}\left(+\right)={\displaystyle \sum {\boldsymbol{R}}_{\boldsymbol{i}}}\hfill & \hfill {\boldsymbol{X}}_{\boldsymbol{i}}-{\boldsymbol{Y}}_{\boldsymbol{i}}>\mathbf{0}\hfill \\ {}\hfill \boldsymbol{W}\left(-\right)={\displaystyle \sum {\boldsymbol{R}}_{\boldsymbol{i}}}\hfill & \hfill {\boldsymbol{X}}_{\boldsymbol{i}}-{\boldsymbol{Y}}_{\boldsymbol{i}}<\mathbf{0}\hfill \end{array}\\ {}\boldsymbol{W}=\mathbf{min}\left(\boldsymbol{W}+,\boldsymbol{W}-\right)\end{array} $$

(21.5)

where W(+) is the total of absolute value of serial numbers of positive rank changes and W(−) is the total of absolute value serial numbers of negative rank changes. W is the test statistic.

21.3.2 Test Hypotheses

For comprehensive performance ranks:

• H_0: Performance rating after merger = performance rating before merger (η₁ = η₂).

• H_1: Performance rating after merger ≠ performance rating before merger (η₁ ≠ η₂).

For operating, market, and financial synergy:

• H₀ = No synergy occurred after merger (η₁ = η₂).

• H₁ = Synergy occurred after merger (η₁ ≠ η₂).

We undertake a double end test. Under the significant level of α, we find the critical value W(α) from the appropriate statistical table. The null hypothesis is rejected if

$$ \begin{array}{l}\mathrm{W}\le \left(-\mathrm{W}\left(\alpha /2\right)\right)\kern0.5em \mathrm{or}\ \\ {}\mathrm{W}\ge \left(\mathrm{W}\left(\alpha /2\right)\right)\end{array} $$

(21.6)

This means that merger has produced significant synergy and performance (and credit) rating has affected. The appropriate variables for operating, market, and financial variables are operating cost ratio, ratio of operating income to total assets and market share, and variability of operating risk, respectively.