After relative Total Shareholder return (TSR) growth, Earnings Per Share (EPS) growth is the most common Long Term Incentive (LTI) hurdle among S&P/ASX 100 companies. Its use is widely supported by investors, with a review of proxy adviser guidelines indicating that EPS growth is one of the preferred measures of long-term performance. The reasons for this include:
· EPS is well defined and understood
· EPS is typically based on statutory NPAT, or other accepted earnings measure
· Analyst forecasts are readily available to benchmark vesting requirements
On face value, an earnings growth measure appears to be a sound indicator of performance, particularly over the medium to long-term. However, the methodology used by most companies can produce unintended outcomes.
We explore the shortcomings and offer some alternatives that may improve the alignment of the outcomes with the intent.
The EPS calculation
EPS is defined as annual NPAT divided by the average number of shares outstanding. Therefore it can be increased in one of two ways:
1. Increasing NPAT
2. Decreasing the number of shares outstanding
For example, a NPAT of $50m and 250 million shares results in an EPS value of $0.20/share. This can be increased by 10% to $0.22 by increasing NPAT to $55, or by a share buy-back of 9% of shares (around 22.5 million). Therefore LTI vesting, based on EPS growth, can occur when profit has not increased.
When shareholders approve LTI plans with EPS growth hurdles, it is likely that they expect vesting to occur based on profit growth, rather than via a reduction in the number of outstanding shares.
The reverse is true if companies issue new shares. That is, the EPS can reduce, even if NPAT increases, if there is an increase in the number of shares outstanding. This can potentially reduce LTI vesting when profit performance has been good.
While it may be possible for Boards to use discretion to adjust the EPS calculation for incentive purposes when capital changes occur, the use of positive discretion (to increase awards) is generally frowned upon. Any adjustment of the EPS calculation for incentive purposes needs to be communicated very clearly and carefully to both investors and management.
The EPS growth calculation
The current EPS growth calculation used by most Australian companies is three-year compound annual growth rate (CAGR). Table 1 shows how the calculation usually works.
Table 1: Standard EPS growth calculation
Measurement Period
| EPS
| Annual EPS Growth
| 3-year EPS CAGR
|
Year 0, the year prior to the performance period
| $0.100
| –
| –
|
Year 1 of performance period
| $0.105
| 5.0%
| –
|
Year 2 of performance period
| $0.110
| 4.8%
| –
|
Year 3 of performance period
| $0.115
| 4.5%
| 4.8%
|
The 3-year CAGR can be calculated by compounding the three annual results, or it can be calculated by simply comparing the year 3 result with the year 0 result. That is, the following equation:
Results in the same answer as:
Therefore, the actual EPS value in years 1 and 2 make no difference to the three-year outcome, because they cancel out in the formula. It is likely that at least some investors believe that all three years count towards the final three-year CAGR, when in fact the year 3 and year 0 EPS values alone dictate the result. This can produce very volatile outcomes. Full vesting is virtually assured three years after a year with poor EPS performance and vesting is very unlikely three years after a very good year.
Table 2 shows an example, based on real data from an Australian consumer company that currently uses EPS growth as a performance measure to determine vesting outcomes of the LTI plan. Fifty per cent vesting occurs at the threshold of 5% EPS CAGR and full vesting occurs at 10% EPS CAGR. The 2015 and 2016 EPS values are based on analyst forecast EPS values.
Table 2: Vesting based on EPS growth
FY
| EPS
| Annual EPS Growth
| 3-year EPS CAGR
| Vesting Percentage
|
2005
| 0.433
| –
| –
| –
|
2006
| 0.377
| -12.9%
| –
| –
|
2007
| 0.413
| 9.5%
| –
| –
|
2008
| 0.524
| 26.9%
| 6.6%
| 66%
|
2009
| 0.605
| 15.5%
| 17.1%
| 100%
|
2010
| 0.660
| 9.1%
| 16.9%
| 100%
|
2011
| 0.781
| 18.3%
| 14.2%
| 100%
|
2012
| 0.601
| -23.0%
| -0.2%
| 0%
|
2013
| 0.105
| -82.5%
| -45.8%
| 0%
|
2014
| 0.356
| 239.0%
| -23.0%
| 0%
|
2015
| 0.526f
| 47.8%
| -4.3%
| 0%
|
2016
| 0.555f
| 5.5%
| 74.2%
| 100%
|
The low 2013 result guarantees full vesting in 2016 and the same is probably true for 2017. In fact, the EPS required in 2016 for full vesting is less than $0.14, representing just 25% of the current analyst forecasts. This is clearly a free kick.
Conversely, vesting was highly unlikely three years after the high EPS value in 2011. The EPS value required in 2014 to achieve full vesting was $1.04, three times higher than the actual result. Unachievable hurdles provide no incentive or retentive power, so this is also an undesirable outcome.
Table 2 illustrates how a company with reasonably consistent EPS over 12 years (except for a relatively poor outlying result in 2013) can produce LTI vesting that is generally all or nothing. While different vesting outcomes should be a consequence of different levels of performance, we would expect partial vesting to occur more often in a consumer sector company.
Perversely, this is actually rewarding earnings volatility. If the EPS in 2013 had been higher (say the same is in 2012 $0.601), then the shareholders would have been significantly better off, but the forecast vesting in 2016 would be 0%, rather than the likely 100%. This presents a clear disconnect between what is in the best interests of shareholders and what is in the best interests of executives with LTI vesting linked to EPS growth. It could, for example, encourage behaviours that produce volatile earnings, rather than steady growth.
An additional complication arises if EPS is ever negative. If this occurs, then three years later the EPS growth percentage cannot be calculated at all.
We conclude that only companies with very stable earnings can appropriately use EPS growth using the standard growth formula.
Alternative approaches to calculating EPS growth
The two shortcomings in the standard approach to calculating EPS growth are that:
1. Only the final year EPS is used to determine CAGR and intermediate results have no impact
2. The base year EPS (prior to the performance period) represents only one year of historical performance and volatility in this value tends to produce all or nothing LTI vesting outcomes three years later, and it must never be negative or zero
The first issue can be addressed simply be averaging the actual EPS over the performance period, before growth is calculated. The second issue can be addressed in the same way, by averaging the actual EPS for the three years prior to the performance period. This also eliminates the likelihood of EPS being negative or zero.
Table 3 illustrates the impact this is likely to have by comparing the standard formula outcomes with the alternatives outlined here. The vesting criteria applied are the same as the real example above in Table 2. That is, vesting is 50% at 5% EPS CAGR and 100% at 10% EPS CAGR with a straight-line formula in between these points.
Table 3: Impact of changing the EPS formula on vesting outcomes
Year
| Actual Annual EPS
| 3-year EPS CAGR
(Standard)
| Vesting Percentage
| 3-year Rolling Average EPS
(Alternative)
| CAGR in average EPS
| Vesting Percentage
|
Historical EPS
| 9.9
| –
| –
| –
| –
| –
|
10.0
| –
| –
| –
| –
| –
|
10.2
| –
| –
| 10.033
| –
| –
|
11.0
| 3.6%
| –
| 10.400
| –
| –
|
13.2
| 9.7%
| –
| 11.467
| –
| –
|
1
| 13.7
| 10.3%
| 100%
| 12.633
| 8.0%
| 80%
|
2
| 10.0
| -3.1%
| 0%
| 12.300
| 5.8%
| 58%
|
3
| 9.8
| -9.5%
| 0%
| 11.167
| -0.9%
| 0%
|
4
| 14.0
| 0.7%
| 0%
| 11.267
| -3.7%
| 0%
|
5
| 14.0
| 11.9%
| 100%
| 12.600
| 0.8%
| 0%
|
6
| 13.1
| 10.2%
| 100%
| 13.700
| 7.1%
| 71%
|
7
| 14.5
| 1.2%
| 0%
| 13.867
| 7.2%
| 72%
|
8
| 15.5
| 3.5%
| 0%
| 14.367
| 4.5%
| 0%
|
9
| 17.6
| 10.3%
| 100%
| 15.867
| 5.0%
| 50%
|
10
| 16.7
| 4.8%
| 0%
| 16.600
| 6.2%
| 62%
|
| Average:
| 40%
| Average:
| 39%
|
The 3-year rolling average EPS is the simple average of EPS over three years, so 16.600 in year 10 is the average of 15.5, 17.6 and 16.7 in years 8, 9 and 10. The CAGR of average EPS compares two three-year average values. For example the 6.2% in year ten is the CAGR required to increase the average over years 5, 6 and 7 (13.867) to the average over years 8, 9 and 10 (16.6).
In Table 3, the average vesting over time is similar, whether using the standard formula or the alternative method. However, the standard formula produces full vesting in four of the ten years and zero vesting in the other six. Whereas, the alternative rolling average EPS growth produces partial vesting in six of the ten years and no vesting in the remaining four years.
Figure 1 shows the result graphically and overlays the rolling three-year EPS value in cents/share.
Figure 1: Comparative vesting outcomes
The only restriction on the use of the alternative methodology is that six years of EPS history is required. It is therefore unsuitable for a newly listed company, although it may be possible to use a proxy, based on historical earnings and the number of shares at IPO.
Suggestions
1. If you are considering the use of EPS growth as part of an LTI, think carefully about how the calculation works in practice and whether or not it is aligned with shareholders interests. If you currently use EPS growth, and calculate it using the standard methodology, consider using the growth in average EPS instead. Since investors and proxy advisers may not be familiar with this approach, it will also be important to engage with both and communicate the reasons for the change.
2. Consider the consequences of the growing phenomenon of share buybacks and amend your policy accordingly. Perhaps if you do not allow your executives a free kick, they may find another use for the cash (such as regular dividends, or, even better according to the central bankers, investing in R&D or production).