Skellerup (SKL) Fundamentals

[Waltzing]

SKL is paying over 4% - at this price.. cheap as chips .. its the market. They will be bringing back there Euros and Pounds..

[Snoopy]

FY2019 was been the first year that dividends have not been fully imputed, and it looks like given the multinational production strategy, this will be the case forever into the future. Granted, the dividends have been increased, which means that dividend hungry shareholders are not worse off in dollars paid out terms. As Liz Coutts highlights in the Chairman's address:

"While much of our product development and design is done in New Zealand, more than three quarters of our products are manufactured overseas"

The calculations to work out the equivalent gross figure for FY2019's and FY2020s unimputed dividends, those actually paid in the FY2019 and FY2020 financial years, are as follows:

FY2019 P1/ 7.0c (55% imputed) = 3.85c (FI) + 3.15c (NI) = 3.85c/0.72 +3.15c = 5.35c +3.15c = 8.50c (gross dividend)

FY2019 P2/ 5.5c (50% imputed) = 2.75c (FI) + 2.75c (NI) = 2.75c/0.72 +2.75c = 3.82c +2.75c = 6.57c (gross dividend)

FY2020 P1/ 7.5c (50% imputed) = 3.75c (FI) + 3.75c (NI) = 3.75c/0.72 +3.75c = 5.21c +3.75c = 8.96c (gross dividend)

FY2020 P2/ 5.5c (50% imputed) = 2.75c (FI) + 2.75c (NI) = 2.75c/0.72 +2.75c = 3.82c +2.75c = 6.57c (gross dividend)

The above is how I have previously dealt with this issue as regards Skellerup. The other way is to take advantage of our knowledge that says:

1/ the way to convert an unimputed 'net dividend' to a 'gross dividend' is to divide the 'net dividend' by one (that is a complicated mathematical way of saying that with no imputation credits a 'net dividend' and a 'gross dividend' are the same thing).
2/ the way to convert a fully imputed 'net dividend' to a 'gross dividend' is to divide the 'net dividend' by 'one minus the company tax rate' (1-0.28) which is 0.72.

Partially imputed dividends are calculated using a divisor that is between the two extremes of '1' and '0.72'. If you think of the distance between the two extremes as 28 steps, then exactly half way between (50% imputation) is 14 steps. That corresponds to a divisor of: 1-0.14= 0.86

For 55% imputation we are going to end up closer to full imputation than the 50% imputation case. The number of steps we have to cover is 0.55x28=15.4. 1-0.154= 0.846

So using this method on the gross dividends I have calculated above

FY2019 P1/ 7.0c (55% imputed) = 7.0c/0.846 = = 8.27c (gross dividend)

FY2019 P2/ 5.5c (50% imputed) = 5.5c/0.86 = = 6.40c (gross dividend)

FY2020 P1/ 7.5c (50% imputed) = 7.5c/0.86 = 8.72c (gross dividend)

FY2020 P2/ 5.5c (50% imputed) = 5.5c/0.86 = 6.40c (gross dividend)

These answers should be identical with the answers I used in my alternative calculation method in my quote bubble above, but they are not. Can anyone explain the difference?

SNOOPY

[Snow Leopard]

The above is how I have previously dealt with this issue as regards Skellerup. The other way is to take advantage of our knowledge that says:

1/ the way to convert an unimputed 'net dividend' to a 'gross dividend' is to divide the 'net dividend' by one (that is a complicated mathematical way of saying that with no imputation credits a 'net dividend' and a 'gross dividend' are the same thing).
2/ the way to convert a fully imputed 'net dividend' to a 'gross dividend' is to divide the 'net dividend' by 'one minus the company tax rate' (1-0.28) which is 0.72.

Partially imputed dividends are calculated using a divisor that is between the two extremes of '1' and '0.72'. If you think of the distance between the two extremes as 28 steps, then exactly half way between (50% imputation) is 14 steps. That corresponds to a divisor of: 1-0.14= 0.86

For 55% imputation we are going to end up closer to full imputation than the 50% imputation case. The number of steps we have to cover is 0.55x28=15.4. 1-0.154= 0.846

So using this method on the gross dividends I have calculated above

FY2019 P1/ 7.0c (55% imputed) = 7.0c/0.846 = = 8.27c (gross dividend)

FY2019 P2/ 5.5c (50% imputed) = 5.5c/0.86 = = 6.40c (gross dividend)

FY2020 P1/ 7.5c (50% imputed) = 7.5c/0.86 = 8.72c (gross dividend)

FY2020 P2/ 5.5c (50% imputed) = 5.5c/0.86 = 6.40c (gross dividend)

These answers should be identical with the answers I used in my alternative calculation method in my quote bubble above, but they are not. Can anyone explain the difference?

SNOOPY

Why would you think that?

With a declared dividend of N imputed to a percent P then the Tax T always is (at the current corporate tax rate of 28%):

T = N * P/100 * 0.28 / 0.72

and the gross dividend G is

G = N + T

Just us that formula and trust me, I am not an accountant

[Snoopy]

With a declared dividend of N imputed to a percent P then the Tax T always is (at the current corporate tax rate of 28%):

T = N * P/100 * 0.28 / 0.72

and the gross dividend G is

G = N + T

Just us that formula and trust me, I am not an accountant

Ok Snow Leopard, Let's tax the latest 5.5c dividend and apply your tax calculation formula:

T = N * P/100 * 0.28 / 0.72 = 5.5c * 50/100 * 0.28/0.72 = 1.07c

Now work out the gross dividend

G = N + T = 5.5c + 1.07c = 6.57c

Now compare that to my first effort below

FY2020 P2/ 5.5c (50% imputed) = 2.75c (FI) + 2.75c (NI) = 2.75c/0.72 +2.75c = 3.82c +2.75c = 6.57c (gross dividend)

and you will see that we both have the same answer. But is it the right answer?

---------

I want to introduce the algebraic entity 'TD' for 'Tax divisor', I am defining this as the number you must divide into the Net Profit to get the gross profit. For a fully imputed dividend TD =0.72. For an unimputed dividend TD = 1

T = G-N = N/TD - N = N( 1/TD - TD ) = N (1-TD)/TD = N *(0.28/0.72) for a fully imputed dividend.

Now compare that to your formula SN for a 100% imputed dividend

T = N * P/100 * 0.28 / 0.72 = N *100/100 * 0.28/0.72 = N *(0.28/0.72)

So for 100% imputation we get the same answer. This is getting promising. We have now agreed twice within the same post!

----------------------

But look what happens when we change the imputation rate to only 50%, which equates to 'TD =0.86'

T = G-N = N/TD - N = N( 1/TD - TD ) = N (1-TD)/TD = N *(0.14/0.86) for a 50% imputed dividend.

Now compare that to your tax formula SN for a 50% imputed dividend

T = N * P/100 * 0.28 / 0.72 = N *50/100 * 0.28/0.72 = N *(0.14/0.72)

and you will see that we no longer agree. So which is the right approach?

SNOOPY

[Snow Leopard]

Snoopy

The formula I provided is ALWAYS correct.

[Snoopy]

Snoopy

The formula I provided is ALWAYS correct.

Here is the information from my March 19th 2020 dividend statement, normalised down to one share:

Payment Rate	5.5000c
less Withholding Tax	1.0985c
equals Net Dividend	4.4015c
NZ Imputation Credits	1.0694c
Withholding Tax	1.0985c
Gross Dividend	6.5694c

Using the SN formula for calculating imputation credits

T = N * P/100 * 0.28 / 0.72 = 5.5c * 50/100 * (0.28 /0.72) = 1.0694c

This is the same imputation figure printed on my dividend statement. So it looks like the SN formula is consistent with the information that Computershare is giving me.

SNOOPY

[Waltzing]

remark removed.

[Snoopy]

Using the SN formula for calculating imputation credits

T = N * P/100 * 0.28 / 0.72

I still don't get the derivation of the above formula. However, given it has the Snow Leopard 'black spot' of approval, and the answers line up with the dividend statements coming out of Computershare, I will carry on with it.

Tax isn't such an issue from a Skellerup management perspective. They invest in different centres all over the world, manufacture in different centres all over the world, and pay tax in different jurisdictions all over the world. The problem is that our Inland Revenue department only recognises imputation credits on that part of the profit generated in New Zealand. If we shareholders get a part of dividend from profits earned offshore, then that part of the profit is 'double taxed' in our hands. By this I mean that if Skellerup profit is earned in the United States (as an example) it will be tax in the United States. But if that U.S. profit is passed on to resident New Zealand shareholders, then that bit of the profit that as already been will be taxed again by order of the New Zealand government as if no prior tax has been paid.

The partially imputed dividends that have suffered from this effect, I requote in the bubble below:

The calculations to work out the equivalent gross figure for FY2019's and FY2020s unimputed dividends, those actually paid in the FY2019 and FY2020 financial years, are as follows:

FY2019 P1/ 7.0c (55% imputed) = 3.85c (FI) + 3.15c (NI) = 3.85c/0.72 +3.15c = 5.35c +3.15c = 8.50c (gross dividend)

FY2019 P2/ 5.5c (50% imputed) = 2.75c (FI) + 2.75c (NI) = 2.75c/0.72 +2.75c = 3.82c +2.75c = 6.57c (gross dividend)

FY2020 P1/ 7.5c (50% imputed) = 3.75c (FI) + 3.75c (NI) = 3.75c/0.72 +3.75c = 5.21c +3.75c = 8.96c (gross dividend)

FY2020 P2/ 5.5c (50% imputed) = 2.75c (FI) + 2.75c (NI) = 2.75c/0.72 +2.75c = 3.82c +2.75c = 6.57c (gross dividend)

The imputed tax bill of each of these dividends is as follows:

FY2019 P1/ 7.0c (55% imputed) = 7.0c * 55/100 * 0.28 / 0.72 = 1.50c

FY2019 P2/ 5.5c (50% imputed) = 5.5c * 50/100 * 0.28 / 0.72 = 1.07c

FY2020 P1/ 7.5c (50% imputed) = 7.5c * 50/100 * 0.28 / 0.72 =1.46c

FY2020 P2/ 5.5c (50% imputed) = 5.5c * 50/100 * 0.28 / 0.72 = 1.07c

A New Zealand shareholder on a 28% tax rate (e.g. an NZ Company investing in Skellerup) will have the following total tax deducted from their respective dividends:.

FY2019 P1/ 7.0c (55% imputed) = 8.5c * 0.28 = 2.38c

FY2019 P2/ 5.5c (50% imputed) = 6.57c * 0.28 = 1.83c

FY2020 P1/ 7.5c (50% imputed) = 8.96c * 0.28 = 2.51c

FY2020 P2/ 5.5c (50% imputed) = 6.57c * 0.28 = 1.83c



So the extra tax levelled on NZ shareholders, due to the dividend not being fully imputed, was:

	NZ Shareholder Tax	less Imputed Tax	equals Additional NZ Resident Tax
FY2019 P1: 7.0c (55% imputed)	2.38c	1.50c	0.88c
FY2019 P2: 5.5c (50% imputed)	1.83c	1.07c	0.76c
FY2020 P1: 7.5c (50% imputed)	2.51c	1.46c	1.05c
FY2020 P2: 5.5c (50% imputed)	1.83c	1.07c	0.76c

For a 50% imputed dividend, the Additional NZ resident tax amounts to 14% of the net declared dividend.

SNOOPY

[Snoopy]

For this model I am using an ROE of 17.8% (the actual average of the last 9 years) and a dividend payout ratio of 62% (the actual dividend payout of the last 9 years).

using a PE of 12.6 (actual average over the last 9 years)

The last time Skellerup qualified for this kind of analysis was FY2014, incorporating the previous five year perspective that went with this date. There were three crucial parameters involved in the modeling which I have quoted above. For the FY2020 edition of the Buffett growth model, I have recalculated these parameters as below.

	FY2016	FY2017	FY2018	FY2019	FY2020	Average
		New Wigram factory opens		Nexus Foams (NZ) & 35% of SimLim (USA)	Silclear (UK)
Return on Shareholder Equity	14.7%	12.3%	15.2%	16.4%	15.6%	15.0% (rounded up from 14.8%)
Dividend Payout Ratio	92%	88%	71%	83%	87%	84%
PE Ratio at 30th September	11.5	16.6	15.7	15.2	19.9	15.8

The dividend payout ratio is based on the dividends actually paid out in the financial year under question - normally the final dividend for the previous year and the interim dividend for the current year, (not the dividends declared relating to the results of that year).

The number of previous years that I use to generate my data is a judgement call. Last time I used nine years of data. The more years of data that you use, the better longer term picture you get. But over time a business evolves. So the longer series of data may be less representative of the business today, and going forwards. And it is the future that is of most interest when we are making future projections. FY2016 marked the start of a 'new era' for Skellerup. I quote from the Chairman's address in the FY2016 Annual Report.

"The FY16 year included a number of notable milestones. The most significant is the completion of the of the base build of our new facility at Wigram which has enabled us to commence the careful and gradual relocation of our Agri business from Woolston to Wigram."

"Another notable milestone has been the growth we have achieved in international markets."

So this time I have elected to use my 'minimum period' of just five years, to keep my Return on Equity, Dividend Payout Ratio and market rated PE position most relevant.

Some financial analysts might see the idea of a 10 year projection forwards as absurdly unreliable, because so much can happen in that time. Buffett argues that for a special subset of businesses, that have strong internal fundamentals, it is actually easier to predict where that business will be in ten years than two. In two years any short term shock might hit. But over the much longer time period of 10 years, the underlying competitive advantage of this select group of businesses that can pass the Buffett tests are unlikely to be derailed.

SNOOPY

[percy]

I think Ggcc's post #870 helps explain why the market is prepared to pay above the average PE of 15.6 ie 19.7.It is certainly not based on eps growth.
'The market however is factoring in negative interest rates that are set to arrive in the coming years. SKL still gives a better return than banks give in a term deposit. Of course no guarantee this will continue of course.'