05/06/2024
In previous years, Guerdon Associates has investigated how the risk profiles of ASXlisted entities relate to their executives’ remuneration. Two pertinent examples include a 2014 article that looked at how share price volatility related to vesting outcomes (see HERE), and a 2022 article that inquired into the relationship between the betas of ASX 100 constituents and their atrisk pay (see HERE).
Following on from these earlier articles, we examined how ASX 200 companies’ risk profiles – as measured by both beta and share price volatility – relate to the amount of incentive pay CEOs receive as equity. As a forerunner to this analysis, we will also shed light on the risk profiles and equitybased remuneration structures that are typical of ASX 200 companies.
For those that like to skip to the ending, we found that the riskiest companies do not pay CEOs with equity. This should not be a surprise. To attract and retain a CEO for a business with a highly uncertain future, payment in cash may be the only solution.
But that is not all. For those who would like to read the whole story, and get a kick out of the maths, read on.
Background
We use the following abbreviations:
 STI: ShortTerm Incentive
 LTI: LongTerm Incentive
 LTE: LongTerm Equity
 TFR: Total Fixed Remuneration
 TR: Total Remuneration
STI and LTI plans are performancecontingent, whereas LTE plans only require continued service. The term “equitybased incentive” refers to the sum of equitybased STI, LTI, and LTE (while LTE does not factor in performance, it still “incentivises” executive retention). TFR refers to the sum of base salary, superannuation, and nonmonetary benefits, while TR is the sum of TFR, STI, LTI, and LTE.
All equitybased remuneration is reported based on the maximum incentive opportunity at fair value (discounting for dividends forgone). The following terminology is used when analysing equitybased incentive structures:
 The equity portion of STI refers to maximum fairvalue equity STI divided by maximum fairvalue total STI;
 The equity portion of TR refers to maximum fairvalue equity remuneration divided by maximum fairvalue TR; and
 The equity “leverage” over TFR refers to maximum fairvalue equity remuneration divided by fairvalue TFR.
In this article, we solely consider remuneration plans for ASX 200 executive directors, focusing on CEOs. Only three companies in our remuneration data sample (n=144) had nonCEOs as the highestranking executive director with available remuneration data. For simplicity, we refer to the entire sample as “CEOs”.
For each constituent’s daily logreturns R_{i} and the daily logreturns of the ASX 200 index R_{m} (based on capitalisationweighted share prices of the index), the beta is given by β = Covariance(R_{i}, R_{m}) / Variance(R_{m}). It measures the extent to which an individual stock varies in correspondence with the overall index.
Each constituent’s volatility is given by σi = √Variance(R_{i}), measuring their stock’s level of variability without reference to the market. Both betas and volatilities are calculated using all trading days over the 3year period spanning from 1 July 2020 to 30 June 2023, and daily logreturns are based on close price. Volatilities are annualised.
When reporting statistical significance results, we follow the convention of a 5% significance level, meaning that we tolerate a risk of committing Type I errors (false positives) of up to 5%. We use the following significance codes for Pvalues:
0.05 ≤ P < 0.1 
0.01 ≤ P < 0.05 
0.001 ≤ P < 0.01 
P < 0.001 
. 
* 
** 
*** 
Exploratory Data Analysis
Only one company in the sample (n=197) had a negative beta (0.054), indicating that this company had a very mild tendency to move against the index, rather than upward/downward together with the index. All other companies in the sample had a positive beta, meaning that they tended to move upward/downward with the index.
The summary statistics in Table 1 indicate that the overall distribution of betas is centred around one, with 50% of betas falling between 0.774 and 1.222, indicating that most stocks tend to move together with the index with a comparable level of volatility to the index. The nearly equidistant positioning of the quartiles, and the similarity of median and mean, suggest that the betas are symmetrically distributed.
Table 1: Summary statistics for 3year beta and volatility of ASX 200 constituents

Minimum 
First Quartile 
Median 
Mean 
Third Quartile 
Maximum 
Beta 
0.054 
0.774 
0.981 
1.018 
1.222 
3.004 
Volatility 
0.166 
0.262 
0.318 
0.352 
0.424 
0.823 
The average stock volatility of 35.2% is higher than the index volatility of 13.7%, which also happens to be lower than the minimum stock volatility of 16.6%. This indicates that the individual constituents’ share prices are more volatile than the capitalisationweighted share price of the index.
This is to be expected, in part due to the variancereducing effects of averaging, and since the constituents will often move in different directions, leading to lower index volatility by cancellation of upward and downward movements.
Using FY2023 executive director remuneration data for the ASX 200 constituents (n=144), we looked at how many of these companies delivered incentives (i.e., STI, LTI, and LTE) in equity. The remuneration data set consisted entirely of CEOs, except for two companies with an executive chair and one company with a CFO as the highestranking executive director with available data. We refer to the sample as “CEOs” for simplicity. We found that 136 company CEOs (94.4%) in the sample of 144 had equitybased incentives, leaving 8 companies that provide the CEO with no equity incentives.
We also looked at the distribution of companies providing the CEO with equitybased STI plans (e.g., deferred shares). Table 2 summarises the frequency of different CEO cash/equity splits for STI, with 0% equity and 50% equity being particularly common, at 30.6% and 29.9% of ASX 200 companies’ CEOs respectively. Only 2.8% of companies had a fully equitybased STI. Note that the ranges “0% to 50%” and “50% to 100%” are exclusive of the endpoints, making each category in the table nonoverlapping.
Table 2: Proportion of CEO STI delivered as equity for ASX 200 constituents in FY2023

0% 
0% to 50% 
50% 
50% to 100% 
100% 
Count 
44 
40 
43 
13 
4 
Percent 
30.6% 
27.8% 
29.9% 
9.0% 
2.8% 
Regression Analysis
We employed logistic regression (detailed in the Methodology section) to examine whether ASX 200 companies with CEO equitybased incentives had a different risk profile from those with cashonly incentives. This was done by regressing a binary response variable, representing the presence of equity in variable pay, against the beta and stock volatility of each company.
Note that logistic regressions do not provide an Rsquared (R^{2}) statistic to measure goodnessoffit of the model, so we only report Rsquared for linear regressions.
Both logistic regressions gave a significant result. The regression against beta (p=0.006) had a negative slope (3.279), indicating that companies with lower beta were more likely to have equitybased incentives. The regression against volatility (p=0.0382) also had a negative slope (5.276), giving a similar conclusion. Overall, these results suggest that higherrisk companies may be more averse to providing variable pay as equity relative to less risky companies.
Figures 1 and 2 illustrate the equity portion of total remuneration (i.e., Equity STI + Equity LTI + LTE all divided by TR) plotted against beta and stock volatility, respectively. These plots highlight how companies without equitybased pay (having zero for the equity portion) tend to have slightly higher risk, reinforcing the regression results. In addition, Figure 1 shows that all equityfree companies except for one had a beta higher than one, indicating abovemarket volatility.
Figure 1: Equity portion of CEO total remuneration against 3year beta
Figure 2: Equity portion of CEO total remuneration against 3year stock volatility
Figure 3 highlights that ASX 200 constituents’ CEOs with equitybased incentives had both lower beta and lower stock volatility on average, relative to constituents without equity incentives. Equityfree companies had an average beta of 1.329, while companies with equity incentives averaged 1.009. This further reinforces earlier conclusions regarding the association of risk with the presence of CEO equity remuneration, while highlighting that companies with equity incentives more closely follow market trends, with an average beta of approximately one.
Figure 3: Average 3year beta and 3year stock volatility for ASX 200 constituents’ CEOs with/without equitybased incentives
For the companies that had equity incentives, neither the equity portion of TR, nor the equity leverage over TFR, were found to be significantly related to 3year beta or stock volatility. The Pvalue and Rsquared of each associated linear regression is shown in Table 3. This implies that while the “presence” of equity in atrisk pay was found to be linked with both beta and stock volatility, the “amount” of equity had no strong association with either.
Table 3: Pvalue and multiple Rsquared for each simple linear regression of CEO equity portion/leverage of TR/TFR against 3year beta/volatility (only considering companies’ CEOs that had equitybased incentives in FY23)

3year Beta 
3year Stock Volatility 
Equity Portion of TR 
P=0.246 (R2=0.010) 
P=0.428 (R2=0.005) 
Equity Leverage over TFR 
P=0.413 (R2=0.005) 
P=0.514 (R2=0.003) 
To compare the risk profiles of different companies based on the level of equitybased STI, the risk measures (beta and volatility) were regressed against a categorical variable representing the level of STI provided in equity, based on the categories in Table 2, with “0%” equity treated as the baseline. The results for the beta regression are shown in Table 4, and those for the volatility regression are shown in Table 5.
Table 4: Coefficient estimates and Pvalues from multiple linear regression of 3year beta against CEO STI equity level

Coefficient 
Standard Error 
PValue 
Significance 
0% to 50% 
0.156 
0.062 
0.013 
* 
50% 
0.159 
0.061 
0.010 
** 
50% to 100% 
0.091 
0.090 
0.311 
– 
100% 
0.196 
0.148 
0.187 
– 
The regression of beta against STI equity level was found to be collectively significant (p=0.013, R2=0.087), with only the equity levels “0% to 50%” (p=0.013) and “50%” (p=0.010) found to differ significantly from the baseline zeroequity case. The coefficients for these levels were 0.156 and 0.159, indicating that companies with over 0% but at most 50% of STI in equity had a lower beta relative to the equityfree companies. However, companies with more than half of STI given in equity did not differ significantly in beta from equityfree companies.
Table 5: Coefficient estimates and Pvalues from multiple linear regression of 3year stock volatility against CEO STI equity level

Coefficient 
Standard Error 
PValue 
Significance 
0% to 50% 
0.086 
0.022 
0.0002 
*** 
50% 
0.100 
0.022 
8.96E6 
*** 
50% to 100% 
0.067 
0.032 
0.039 
* 
100% 
0.090 
0.053 
0.093 
. 
The regression of stock volatility against STI equity level was also collectively significant (p<0.001, R2=0.194), with all three of the levels “0% to 50%” (p<0.001), “50%” (p<0.001), and “50% to 100%” (p=0.039) found to differ significantly from the baseline zeroequity case. However, companies with fully equitybased STI did not differ significantly in volatility from those with fully cashbased STI.
Figure 4 illustrates that, on average, the riskiest ASX 200 companies were those that either had fully equitybased STI or fully cashbased STI, with both brackets having abovemarket volatility as indicated by betas above one, and higherthanaverage daily return volatilities (refer to Table 1). The other “midrange” brackets had risk profiles more representative of the overall index, with average betas of approximately one.
Figure 4: Average 3year beta and 3year stock volatility for ASX 200 constituents’ CEOs with varying levels of STI delivered as equity
The equityfree companies form a subset of the companies with zeroequity STI, but the converse is not true (not all companies with zeroequity STI would be equityfree overall). The average beta for companies with cashonly STI is 1.121, which is lower than the average of 1.329 for equityfree companies overall. This indicates that many of the companies with cashonly STI provide equitybased LTI or LTE, likely to offset the lack of longterm focus in their STI plans.
Conclusion
Our findings for the ASX 200 over the threeyear period spanning from 2020 to 2023 indicate that higherrisk companies, as measured by beta and stock volatility, tend to be more averse to CEO equitybased remuneration compared to less risky companies.
Companies that had CEO equitybased incentives followed the market more closely than those with CEO cashonly incentives.
The companies that deviated the most from the market were those without any equity pay, or those with 100% equity STI, with abovemarket volatility in both cases.
While the presence of equity in CEO incentive pay seemed to be associated with companies’ risk profiles, the amount provided in equity by those with equitybased plans was not found to be linked with company risk.
Methodology
The “ASX 200” refers to all 199 current constituents of the ASX 200 index, as last rebalanced in March 2024. Financial data over the 3year period spanning from 1 July 2020 to 30 June 2023 is considered for only these companies, without any adjustments made to account for index rebalancing in prior quarterly periods.
After appropriate data cleaning, the sample consisted of 197 index constituents with available stock market data that were used to compute financial statistics (e.g., beta, volatility), and 144 constituents with remuneration plan data for FY2023.
Logistic regression involves regressing the “logodds” of a binary response variable against a linear combination of independent variables, like in linear regression. For a binary response variable with “success” probability , the logistic regression model with one independent variable can be expressed as
log[(p/(1p)] = β_{0} + β_{1}x
where log is the natural logarithm. The ratio “p/(1p)” of the success probability over the failure probability is called the “odds”.
The slope coefficient β_{1} has the interpretation that a unit increase in x is associated with an increase in the logodds of success by β_{1}, on average. In other words, for a unit increase in x, the odds of success will change by a multiplicative factor of exp(β_{1}).
In our analysis, p would represent the proportion of companies with equitybased incentives, with the independent variable x being either the 3year beta or the 3year stock volatility.
© Guerdon Associates 2024 Back to all articles