Contact Energy - CEN Chart

[Ferg]

Snoopy, the gross dividend is 19.88c. The IC's are 3.88c. 3.88c / 19.88c = 19.5%. By re-basing using 28% full IC's you are moving the base hence the maths doesn't work - you have calculated a theoretical gross based on IC's that are not being delivered - so you are comparing 22.22c with 19.88c. A pre-RWT dividend of 16c with 19.5% IC's implies a gross dividend of .16/(1-0.195) = 19.88c, which agrees to their statement.

[Ferg]

Also, 3.88/6.22 = 62% imputation is correct per your note. If we look at the gross dividend of 19.88/22.22 = 89%. .62/.89 = 69% per your later calculation of 19.5/28 = 69%.

Pre RWT	IC's	Gross	IC%
16.00	3.88	19.88	19.52%
16.00	6.22	22.22	27.99%
100.00%	62.38%	89.47%	69.72%
		69.72%

[Snoopy]

I'm a bit slow with this Ferg, so I am reformatiting your data so that it becomes blindingly obvious, even to me.

Dividend Alternative Scenarios	Dividend Declared	plus Imputation Credits {A}	equals Gross Dividend {B}	Imputation Credits as %ge of Gross DividendI {A}/{B}
Partially Imputed 07-04-2020	16.0000c	3.8888c	19.8888c	19.55%
Fully Imputed	16.0000c	6.2222c	22.2222c	28.00%
Partial to Full Ratio	100%	62.50%	89.50%	69.82%

The figure in italics is the one Contact refer to on the dividend statement when they say: "Your dividend has been partially imputed to 19.55%."

Extending the number of decimal places has fixed the rounding error on the 'partially imputed percentage' in post 1792. It didn't occur to me to compare the imputation credits with the gross income because the gross income changes with the imputation credit whereas something like the net income of 16c does not. In fact I took a different path entirely by comparing 'imputation credit' with 'imputation credit' directly across the partial to full scenario. I guess what this table shows is that it is possible to compare two different dividend scenarios in many different ways. And what I think is the logical way to do things might not be the right way for Contact to do it. Thanks Ferg for putting me on the right track and finally allowing me to get my head around all of this!

SNOOPY

[Ferg]

You're welcome. In the end you were both right, you were just measuring different things.

I had forced mine to 2 decimal places hence some of the funky numbers.

[Snoopy]

Below I present my corrected earnings picture for the last ten years. You will note that:

1/ I have deleted the FY2015 50cps special dividend from the record, because it will not be possible to repeat that into the future.
2/ The 'Scenario Dividend Per Share Column' represents a prediction of an ongoing dividend of 100% of free cash flow being paid into the foreseeable future.
3/ The (A) - (B) difference column, if negative, represents the amount of the projected dividend not covered by imputation credits. This is important, because a dividend paid without imputation credits is in accounting terms, equivalent to giving shareholders their own capital back (equal to the amount of the unimputed dividend) complete with a tax bill. This is generally bad for investors. It is necessary to make a negative adjustment to account for any expected tax to be paid on the unimputed dividend component.
4/ The capital component of the dividend is the portion of shareholder equity being returned to shareholders. This will need to be removed from the dividend return calculation. Because to pay it is to return to shareholders money on the balance sheet that they already have, so it isn't a shareholder benefit.
5/ The unimputed component tax bill column, represents the income tax charged on share capital that is expected to be paid by the shareholder. A 28% tax rate is assumed. Note that if the (A)-(B) difference is positive there is no extra tax bill. That's because in such a year, the dividend is fully imputed.
6/ The final column represents the dividend per share adjusted for any extra tax obligation.

Scenario Basis Financial Year	eps (A)	Scenario dps (B)	Difference (A)-(B)	Divie Capital Component (C)	Unimputed Tax Bill (D)	Difference (B)-(C)-(D)
2009	22.2c	36.0c	-13.8c	13.8c	3.9c	18.3c
2010	21.4c	33.0c	-11.6c	11.6c	3.2c	18.2c
2011	21.8c	19.0c	+2.8c	0c	0c	19.0c
2012	24.7c	38.0c	-13.3c	13.3c	3.7c	21.0c
2013	28.2c	40.0c	-11.8c	11.8c	3.3c	24.9c
2014	27.8c	45.0c	-17.2c	17.2c	4.8c	23.0c
2015	22.5c	46.0c	-23.5c	23.5c	6.6c	15.9c
2016	22.2c	51.0c	-28.8c	28.8c	8.1c	14.1c
2017	18.7c	42.0c	-23.3c	23.3c	6.5c	12.2c
2018	18.3c	41.0c	-22.7c	22.7c	6.4c	11.9c
Total	227.8c	391.0c				178.5c

The expected average dividend per year, net of tax is therefore: 178.5 / 10 = 17.5cps (net)

Using a tax rate of 28c this is equivalent to a gross income of: 17.5cps /(1-0.28) = 24.3c

If we assume that a business cycle investment 'gross return' of 5.5% is required, then this equates to a CEN share price of:

24.3c /0.055 = $4.42

So $4.42 is therefore 'fair value'. Naturally this valuation assumes no gross disruption to the market, i.e. Tiwai Point remains a going concern

Readers should note that $4.42 represents 'business cycle neutral' fair value. We could argue that we are currently at the top of a low interest rate inspired valuation cycle. My rule of thumb would suggest a 'top of cycle' value some 20% higher than my calculated fair value.

$4.42 x 1.2 = $5.30

Contact Energy is trading at $6.79 as I write this post. This technique would suggest that Contact Energy is now significantly overvalued (28% above fair valuation). But does a capitalised dividend valuation give the full picture?

Below I present my corrected earnings picture for the last ten years. You will note that:

1/ I have deleted the FY2015 50cps special dividend from the record, because it will not be possible to repeat that into the future.
2/ The 'Scenario Dividend Per Share Column' represents a prediction of an ongoing dividend of 100% of free cash flow being paid into the foreseeable future. Howewver, in the two years this policy has been in existence, only 39cps has been paid out. So where the 100% of operating free cashflow exceeds that figure I have capped the dividend payout to 39cps.
3/ The (A) - (B) difference column, if negative, represents the amount of the projected dividend not covered by imputation credits. This is important, because a dividend paid without imputation credits is in accounting terms, equivalent to giving shareholders their own capital back (equal to the amount of the unimputed dividend) complete with a tax bill. This is generally bad for investors. It is necessary to make a negative adjustment to account for any expected tax to be paid on the unimputed dividend component.
4/ The capital component of the dividend is the portion of shareholder equity being returned to shareholders. This will need to be removed from the dividend return calculation. Because to pay it is to return to shareholders money on the balance sheet that they already have, so it isn't a shareholder benefit.
5/ The unimputed component tax bill column, represents the income tax charged on share capital that is expected to be paid by the shareholder. A 28% tax rate is assumed. Note that if the (A)-(B) difference is positive there is no extra tax bill. That's because in such a year, the dividend is fully imputed.
6/ The final column (Column D) represents the 'effective' dividend per share adjustment for any extra tax obligation from paying tax on unimputed distributions.

Scenario Basis Financial Year	eps (A)	Scenario dps (B)	Difference (A)-(B)	Divie Capital Component (C)	Unimputed Tax Bill (D)	Difference (B)-(C)-(D)
2011	21.7c	19.0c	+2.7c	0c	0c	19.0c
2012	24.7c	38.0c	-13.3c	13.3c	3.7c	21.0c
2013	28.1c	39.0c	-10.9c	10.9c	3.1c	25.0c
2014	27.7c	39.0c	-11.3c	11.3c	3.2c	24.5c
2015	22.4c	39.0c	-16.6c	16.6c	4.6c	17.8c
2016	22.1c	39.0c	-16.9c	16.9c	4.7c	17.4c
2017	18.7c	39.0c	-20.3c	20.3c	5.7c	13.0c
2018	18.2c	39.0c	-20.8c	20.8c	5.8c	12.4c
2019	24.3c	39.0c	-14.7c	14.7c	4.1c	20.2c
2020	17.7c	39.0c	-21.3c	21.3c	6.0c	11.7c
Total	225.6c (E)	369.0c (F)				182c
Business Cycle Imputation Rate (E)/(F)		61.14%

.

The expected average dividend per year, net of tax is therefore: 182 / 10 = 18.2cps (net)

Using a tax rate of 28c this is equivalent to a gross income of: 18.2cps /(1-0.28) = 25.3c

Now we come to a critical point in this analysis - the choosing of an indicative interest rate that allows us to value Contact on the basis of being an ongoing income stream, albeit not quite a 'bond equivalent'. I have previously used a figure of 5.5%. But that was in a climate of investment interest rates of some 3.5%. Interest rates have fallen by at least 200 basis points since then. OTOH the previously unthinkable is happening in that the Tiwai Point aluminium smelter is closing, and the power market is consequently in a state of flux as to how particularly Contact will be able to extricate the consequential surplus energy from their Clutha River power stations at Clyde and Roxburgh while transmission north is constrained. To balance these competing factors, I have assessed that a gross return of 4.5% iThe bulk of these tax credits came about in FY2016, because of $204m of impairments net of tax relating to the closure of the Otahuhu Power Station, an assessment that the Taheke Geothermal field is unlikely to be developed in the foreseeable future (from a June 2016 perspective) and a write down in the value of inventory gas. I have calculated above that $63m of those tax credits have been used up, which means that $97m - $63m = $34m are yet to materialise. That means there are probably two years worth of 'Superimputed dividends' for Contact Energy shareholders to come. What value of superimputation is appropriate? I am going with the average that I calculated in the first quoted post on this thread of 71.64% (post 1799) as now appropriate for Contact Energy.

If we assume that a business cycle investment 'gross return' of 4.5% is required, then this equates to a CEN share price of:

25.3c /0.045 = $5.62

So $5.62 is therefore 'fair value'.

Readers should note that $5.62 represents 'business cycle neutral' fair value. We could argue that we are currently at the top of a low interest rate inspired valuation cycle. My rule of thumb would suggest a 'top of cycle' value some 20% higher than my calculated fair value.

$5.62 x 1.2 = $6.74

Contact Energy is trading at $6.25 as I write this post. This technique would suggest that Contact Energy is now 11% overvalued (above fair valuation), not unexpected given the plunge in interest rates of late is probably worrying yield investors, and not out of line with the broad overvaluation of the NZX as a whole.

But does a 'capitalised dividend valuation' give the full picture?

SNOOPY

[Snoopy]

Fair enough guys.

For those interested in the maths here's what it looks like for CEN.
I have used the average forecast figure of 33 cps future dividends off market screener and assumed 65% imputation credits and that an investor looking at this now is treating the almost imminent dividends of 23 cps as a partial return of the purchase price.

Looking at the medium term yield.

$6.38 - 23 cents = $6.15 net purchase price on a medium term view.

33 cps per annum in dividends with 65% imputation credits = 33 / 0.818 = 40.34 cps gross. 40.34 / 615 = 6.56% Gross Yield

Some of you may be puzzled where that '0.818' divisor came from in Beagles partial imputation calculation.

Sometimes it is easier to understand what is going on 'in the middle' (i.e. with partial imputation) if you can understand the extremes. Let's take an example where our net yield from holding the investment is 33c.

Extreme 1/ If there was no imputation the gross income calculation is easy because 'net yield' = 'gross yield' and the divisor is '1': 33c/ 1 = 33c

Extreme 2/ If there was full imputation credits, that means tax has already been paid on the 33c payment at the company tax rate of 28%. Conversely if 28% has gone in tax that must mean that 72% is left. So our divisor to calculate the gross income before the tax is taken out becomes 0.72: 33c/ 0.72 = 45.83c

Now we know the result in both extreme cases, we know the divisor for any 'partial imputation' will be between these two figures. Imagine a straight line starting out at number 72(%) -representing full imputation- with 28 single steps going forwards until you reach 100% - representing zero imputation. Now let's pick an imputation percentage you want to check out, let's say 65% ;-). We know from our imaginary line that the difference between nil imputation and full imputation (a total journey of 100%) is 28 steps. So how many steps on our journey do we need to walk if we only want to go 65% of the way?

Answer: 65% of 28 steps = 0.65 x 28 = 18.2 steps

Now we observe that 65% is closer to 100% than 0%. So starting at the 100% end of the line and walking backwards by 18.2 steps will be the way to get nearer to the 72 figure, representing 100% imputation. By starting at step 100 and walking backwards what step do we come to a halt at? Answer: 100-18.2 = 81.8. Step 81.8 therefore represents the divisor that you need to work out the gross payment that leads to the net payment of 33c: 33c / 81.8% = 40.34cps. Incredibly, this is exactly the same figure that the other Beagle got!

Of course you can repeat the basic procedure in this post with any partially imputed percentage figure that you choose. I am putting this post up because I know not everyone is a whiz banger at numbers. I hope this post helps de-mystify the calculation.

SNOOPY

[Ferg]

Snoopy

Is the deduction of the unimputed tax bill (D) double counting? It is funded personally from the unimputed portion of the dividend (C) received by the shareholder, noting that (C) has already been deducted from the declared dps (B), which excludes imputation credits. I agree with what you are doing - just this one part had me scratching my head.

[Snoopy]

Snoopy

Is the deduction of the unimputed tax bill (D) double counting? It is funded personally from the unimputed portion of the dividend (C) received by the shareholder, noting that (C) has already been deducted from the declared dps (B), which excludes imputation credits. I agree with what you are doing - just this one part had me scratching my head.

This is the way I think of it Ferg.

If you own a share you can sell it at any time and get your capital back (or more fully get your capital back at the market rate). If you are paid an unimputed dividend that is the same as getting some of your capital back. Why? The reason the dividend is unimputed is that no income has been earned by the company before they pay you that dividend. Because if they had earned some money to pay it, the dividend would have come with an imputation credit attached. Of course the tax man sees you getting your unimputed dividend and says "Hey I want my slice of that". So the tax man takes his slice of your unimputed dividend leaving you effectively with less capital than you started with.

The tax efficient way to handle this payment, from a shareholder perspective, would be for the company to make a capital return of the exact same amount of the unimputed dividend payment., Done this way, the company would reduce its capital by a set amount that would go into the pockets of shareholders and our tax man would be nowhere to be seen, in this perfectly legal alternative payment scenario.

The way I see it, paying an unimputed dividend is simply giving shareholders their own capital back with an extra tax bill. So I subtract from the 'total partially imputed dividend' all the unimputed bit (because giving shareholders the capital that they own already back is not really income from where I sit) AND I also take off the extra tax paid to the taxman for the pleasure of shareholders being given their own capital back. That isn't double counting the way I see it.

SNOOPY

[Snoopy]

For those interested in the maths here's what it looks like for CEN.
I have used the average forecast figure of 33 cps future dividends off market screener and assumed 65% imputation credits

Imputation credits are sometimes a hard thing to pin down as companies can keep them on their books for years, and distort the imputation credits attached to future dividends with those imputation credit reserves. Companies can even front load imputation credits by paying a tax bill to the IRD before it is due. Yet over a longer period of time it becomes harder to distort imputation credits paid if shareholders consider a multi year company overview. In the case of Contact Energy, they cleaned out their stash of imputation credits just before Origin Energy gave up their controlling shareholding by paying a special 50c dividend on 23rd June 2015 that was fully imputed. That means that, post that special dividend payment, Contact Energy started with a 'clean imputation credit slate'. There have been ten dividend payments since that time. So what payments, with what percentage of imputation credits attached, have been made since Origin Energy left the share register?

Dividend Payment Date	Amount	Dividend Imputed to... {A}	Imputed to Maximum (B)	Dividend Imputation Rate (A)/(B)
25-09-2015	15cps	0%	28%	0%
23-06-2016	11cps	19.84%	28%	70.9%
23-09-2016	15cps	15.36%	28%	54.9%
17-03-2017	11cps	22.04%	28%	78.7%
19-09-2017	15cps	28.00%	28%	100.0%
06-04-2018	13cps	28.00%	28%	100.0%
18-09-2018	19cps	28.00%	28%	100.0%
09-04-2019	16cps	19.55%	28%	69.8%
17-09-2019	23cps	20.23%	28%	72.3%
07-04-2020	16cps	19.55%	28%	69.8%
Average				71.64%

I thought about what would be the best way to calculate the average imputation rate. Each dividend would have been individually considered by the board, and a decision made on what the appropriate imputation rate should be at each dividend payment time. So I have decided to just average the raw figures, exactly as they were presented to shareholders at the time.

By contrast, the dividend imputation rate that I used in my own capitalised dividend model (my post 1795) was somewhat lower.

Scenario Basis Financial Year	eps (A)	Scenario dps (B)
Total	225.6c(E)	369.0c (F)
Business Cycle Imputation Rate (E)/(F)		61.14%

So there we have it, three perspectives on what the imputation credit rate should be: 61.14%, 65% (as assessed by Beagle, post 1811), 71.64%

Which to choose?

SNOOPY

[Ferg]

Thanks for clarifying. I understand what you are saying 100% and that a return of capital via unimputed dividends is not income, and that then results in a real cash outflow for the shareholder. I agree a share buyback would make more sense. And in the absence of a share buyback, then debt reduction would also be beneficial. Unimputed dividends are not an efficient use of capital, nor are they tax efficient. There is no absolutely debate from me on that.

However, I guess I was looking at it from the angle that if we are going to exclude the unimputed cash given to the shareholder, then the payment of tax from those same proceeds should also be excluded. Why? Assuming returning capital via unimputed dividends does not impact upon future profitability, P/E ratio or SP, then the overall personal and enterprise value should remain unaffected. The funding of tax payments on imputed dividends is independent of all of these factors. Therefore by excluding the unimputed dividend from the valuation, then the tax payment on the unimputed dividend should also be excluded. Yes unimputed dividends erode NTA but, in my opinion, NTA is the floor price for a share valuation and is not highly correlated to the share price or its movements for entities such as MEL. Such entities have a low ratio of capex to depreciation which generates FCF, and the non-retention of this cash erodes the NTA but not the share price or the shareholder wealth. Accordingly, the tax on the unimputed dividend should have no bearing on the overall share, or enterprise, valuation under this train of thought.

Another angle was that if the share price is $6.25 and MEL is returning capital of say $0.20c per share per annum via unimputed dividends, then after 31 years you have all your capital back (admittedly before tax), plus an ongoing imputed dividend flow and an unimputed payment ad infinitum. Shouldn't that be worth something over above the imputed dividends...? Maybe others are not making the adjustments you are making to their calculations, hence they are comfortable with the SP being higher than your calculations. I'm merely exploring other avenues of thought, not saying any method is more or less accurate.

Anyway, I have a numerical proof exploring the financial impact of various scenarios but I'm struggling to retain the formats from Excel....it looks like an unreadable wall of numbers, so I will spare you.