shorting not allowed. :(
Printable View
shorting not allowed. :(
I am using the calculation method shown on Slide 22 from PR2021 (August)
FY2021 Chorus View (Pre IFRS16) Chorus View (Post IFRS16) Reference Snoopy View Bank and Note Debt (1) $2,339m $2,339m (Slide 21 PR2021) $2,339m add PV of CIP Debt Securities (Senior) $198m $198m (my post 2799) $462m (2) (Slide 21, PR2021) add Net Leases Payable $264m (Note 5 AR2021) equals Total Selected Debt $2,537m $2,801m $2,801m less cash $53m $53m $53m equals Total Net Selected Debt $2,484m $2,748m $2,748m divided by EBITDA $629m $649m $629m equals 'Net Selected Debt' / EBITDA 3.95 4.23 4.37
Notes
1/ Total net bank and other senior debt = $140m+($858m+$511m-$70m)+($200m+$500m+$182m)+$18m = $2,339m.
The $70m decrease in balance sheet value of the Euro bonds is because of a change in end of year value of an associated hedging instrument, the cross currency interest rate swaps used to hedge the Euro bonds (AR2021 p54). Likewise the $18m adjustment upwards in the value of the NZ bonds is because of the offsetting $18m decrease in value of interest rate swaps that are used to ensure hedge effectiveness (AR2021 p54).
2/ I don't like the accounting standard that lets a company discount debt that ultimately has to be repaid. So I am assuming a PV discount rate of 0%.
My 'Snoopy' view of this calculation removes the 'lease interest expense' (a part of rent) from EBITDA (so EBITDA is lowered). But I have also removed the 'Net Leases Payable' (a construct of IFRS16) as a liability, as I don't consider that a debt in a pre-IFRS16 sense. I don't consider my approach better than the approach Chorus has used. But my approach aligns better with the old way of assessing this banking covenant.
The other controversial thing that I have done is to disallow any discounting of CIP debt, in addition to removing the 'Net Leases payable' debt. By pure chance the effect of doing each of those two things separately has exactly cancelled out any overall debt change. Nothing should be read into this, as it is just a case of two unrelated changes unexpectedly neutralizing each other.
The number the banks are on the lookout for is anything over 4.75. Chorus would have you believe that they are doing really well. Anything under 4.2 ( Moody's ) and 4.25 ( S&P ) are the trigger ratings for a credit upgrade. Personally I don't believe Chorus do deserve a credit upgrade, based on my retro calculation that shows the company is a slightly less favourable debt position.
SNOOPY
FY2021 Pre IFRS16 View Post IFRS16 View EBIT $215m $224m divided by Net Interest Expense $92m $112m equals 'Interest Coverage' 2.3 2.0
Notes
1/ Pre IFRS16 EBIT Calculation
Published EBIT of $224m. But as explained in my post 2800, 'lease interest' (a rent cost) has not been removed from this figure. However 'lease interest' is an accounting construct derived from whole of contract rent agreements and is not equivalent to annual rent in any given year. We can best estimate annual rent from the Cashflow statement, as cashflow is not affected by accounting constructs (AR2021 p32).
Rent Estimate (exclusive of GST) is: ($28m/1.15) = $24m
Furthermore we must add back the 'Right of Use Asset Depreciation' of $15m (AR2021 p19). Why? Because although 'D' (depreciation) has already been removed to calculate EBIT, the 'Right of Use Asset Depreciation' is an incremental element of depreciation - an accounting construct created by IFRS16, that did not exist before. So if we want it to not exist again, taking a retrospective preIFRS16 view of the accounts, we have to remove this 'incremental deduction' i.e. add it back
So pre-IFRS16 EBIT is $224m -$24m +$15m = $215m
2/ Pre IFRS16 Interest Calculation
For the net interest bill I am using in this calculation (Ref AR2021 p41):
($153m-$1m) - $34m - $30m + $4m = $92m
I have removed from the net interest bill:
(i) $34m of CIP securities 'notional interest' (AR2021 p20)
(ii) $20m of 'lease expense' interest and $7m amortisation from a swaps reset and a further $3m undisclosed. (AR2021 p20)
And I have added back a reduction in interest:
(iii) $4m from the ineffective portion of a cashflow hedge. (AR2021 p20)
The latest reference i can find to the revised interest ratio coverage indicator is in this 25-05-2017 press release:
https://stocknessmonster.com/announc...nu.nzx-301693/
The interest coverage result of 2.4 is less than the revised 2.75 interest coverage covenant referred to in the link above. This calculation is telling me the 'Interest Coverage Ratio' covenant is broken. Not good news!
3/ Post IFRS16 Calculation
For the interest bill (see calculation above), I am now adding back the 'lease interest expense' that I had previously removed:
$92m + $20m = $112m
Since we are now working with current EBIT figures, EBIT remains as reported at $224m (AR2021 p16)
SNOOPY
Is there a reason you’re using EBIT rather than EBITDA for interest cover calc?
Good question. Answer: Because 'Simply Wall Street' do it?
https://simplywall.st/stocks/nz/tele...nsidered-risky
Maybe that wasn't such a good answer as the "Simply Wall Street" guys do seem to be a bit 'out there' with some of their valuations!
According to these guys:
https://learn.financestrategists.com...overage-ratio/
'Interest Coverage Ratio' can be either EBIT/I or EBITDA/I. But then in the worked example they use EBIT/I. Unlike the 'Net Senior Debt' calculation, I haven't yet found a worked example of 'Interest Coverage Ratio' as calculated by Chorus themselves. Anyone have an opinion as to which calculation method I should use and why?
SNOOPY
It depends on what is agreed with the bank - there are various formulations, including straight net profit. At the end of the day all this really is is a KPI for the bank to track which could indicate serviceability issues.
One would presume that they aren’t using EBIT, otherwise you’d be seeing a disclosed covenant breach based on your own version of workings - particularly give the gap between EBIT and EBITDA is so large for Chorus given their massive depreciation charge. Given the size of non-cash items in their EBIT, this would have a poor correlation with ability to service interest obligations. It would also be unusual to have one covenant being linked to EBITDA, and another being linked to EBIT.
I am not sure this post will end up being coherent, but I thought I would throw out a few 'random thoughts' on this topic, while thinking about a 'housing loan analogy'.
For 1/ 'Net Senior Debt' / EBITDA , this is the equivalent of 'Loan to Income Ratio' borrowing agreement for a homeowner. From a bankers perspective, they don't care if you don't keep the house looking brand new (the DA bit is the amount of money you should be spending on sandpaper, overalls and paint). As long as you are holding down your high paying job and the associated cashflow is coming in to look after the bank loan balance, that is all the bank cares about.
2/ EBITDA / 'Net Interest', is a straight 'cashflow in' to 'cashflow out' measure, and might be thought of as analogous to measuring your ability to service an 'interest only mortgage'.
There is of course a relationship between 'Net Interest' and 'Net Senior Debt', the former being an annual payment that is a fractional multiple of the latter. So in that, sense I might argue that 'Net Senior Debt' / EBITDA (1) and EBITDA / 'Net Interest' (2), although separate numbers, are actually measuring the same thing. Having said that, (1) does not change with interest rates. Whereas (2) will reduce in value and increase in risk when interest rates rise. What (2) is saying is that, as interest rates rise, your loan will require more of your income to service it. I would have said that fact is self evident, and there is no need to contrive a statistic to tell yourself what you already know!
What about:
3/ EBIT / 'Net Interest'? (3) is again an analogous measure to an 'interest only mortgage' home loan, albeit a loan on a tighter leash than covenant (2) . Imagine if you were a taxi driver borrowing against your house to buy your brand new taxi cab. After 5 years, your cab might have clicked over sufficient kilometres -with the associated wear and tear- to be in need of replacement, PLUS you may have to renew your taxi operators licence. IOW the 'DA' part of your annual accounts, is representing money you have to put aside each year, just to stay in business.
One conclusion from this is that if you were in a business with a required high capital replacement spend (e.g. running a taxi) then,
1/ the depreciation on your vehicle and
2/ amortisation of your fixed period taxi licence,
represents cash you will have to stump up every 5 years or so, just so you can keep operating your business as it is now. IOW the 'DA' bit of the accounts is not free cashflow in the strictest sense, but represents money that must be put aside every year just to allow normal operations to continue like now, over the medium term.
In this 'taxi' borrowing example, (3) is a much more useful statistic than (2). Also in this 'high depreciation' situation, (3) is the best measure of whether your bank can expect their payments coming in each week, whereas (1) is a better measure of whether the bank are likely to get all their loan capital back - eventually.
Nevertheless, IF your company owns a very long lived asset, like a neighbourhood fibre network, AND has been granted an indefinite licence to operate that network (IOW there is no need to regularly rebuild the 'book value' of certain company assets, as the cash generating ability of those assets is not being reduced in any meaningful way by 'wear & tear' - somehow Chorus springs to mind), THEN (2) is a better 'cash in', 'cash out' 'business balance measure'. But in those situations, is (2) really measuring anything different to (1)? I might argue that in the current low interest rate environment, and with the majority of interest payments being fixed by hedging, it isn't! Could this be the reason that Chorus doesn't mention 'Interest Coverage Ratio' in its presentations these days?
SNOOPY
The difference between total capital expenditure and sustaining capital expenditure is set to become critical in determining the future dividend returns from Chorus shareholders. From the 17th November 2020 Presentation, Slide 33:
"From FY22 we will transition to a dividend policy based on a pay-out range of free cash flow
▪ free cash flow will be defined as net cash flows from operating activities minus sustaining capex."
The following table is compiled from information in:
a/ Slide 43 of the 23rd August 2021 presentation.
b/ AR2021 p21
Capex over 2021 Sustaining Development Sub Total Fibre Layer 2 Sustaining $31m Fibre Products & Systems Sustaining $11m Fibre Network & 'Other Fibre Connections' Sustaining $22m Fibre Customer Retention Sustaining $11m Fibre UFB Communal Development $147m Fibre Connections & Layer 2 Development $244m Other Fibre Connections & Growth Development $83m Fibre Customer Retention Development $18m Fibre Total $567m Copper Network Sustaining $29m Copper Layer 2 Sustaining $4m Copper Customer Retention Sustaining $11m Copper Connections Sustaining $1m Copper Development $0m Copper Total $45m Common IT Sustaining $46m Common Building & Engineering Services Sustaining $14m Common Development $0m Common Total $60m Sustaining Total $180m Development Total $492m Overall Total $672m
The year to year comparison with the quoted post above shows a $6m reduction is 'sustaining capital expenditure'. But that includes $10m in sustaining costs less spent on the copper network. So sustaining capex for the fibre network has gone up over the year, although that is in the face of a fibre network that is still growing. Nevertheless as long as the overall sustaining capex comes down,
free cash flow = net cash flows from operating activities - sustaining capex
then free cashflow goes up. Since dividends going forwards will be based on 'free cashflow', this is a good thing for shareholders.
SNOOPY
The actual FY2021 revenue figures have been published as I write this. So it is time to update my revenue forecasting table.
Chorus has been locked in a regulatory battle with the Commerce Commission on how much 'maximum allowable revenue' (MAR) their 'regulated asset base' of fibre broadband assets will be allowed to charge their downstream retail customers over FY2022 through to FY2024, otherwise known as the 'Price Quality (PQ) Period - First.' (PQP1)
1/ The initial Commerce Commission proposal on 26-03-2021 was a rising sum starting from $715m (FY2022) rising to $755m (FY2024).
2/ A revised proposal on 27-05-2021 reduced this to a range from $689m to $786m.
3/ A further comment on 19-08-2021 from the Commerce Commission stated.
"The Commission has noted in its draft RAB (Regulated Asset Base) decision today that “If all other aspects of our draft PQ decision remained unchanged, our indicative estimate of the combined impact of these decisions would lead to a 2%-2.5% reduction in allowable revenue over the PQP1 period. This figure also includes the impact of updated WACC values applied in the pre-implementation period.”
The market took this to mean that the revised downwards regulated revenue rates may yet be increased again by the time the final decision in December comes out. The case Chorus made to the Commerce Commission was for a regulated rate range between $720m to $820m. I am doubtful that the Commerce Commission will be that generous in their final regulated outcome figures. So for the purpose of this update I am going to assume that the Commerce Commission raises the allowable revenue figures back to the range of the 26-03-2021 proposal: $715m (FY2022) rising to $755m (FY2024) [Forecast fibre revenue from 26-03-2021 'Initial Asset Value' Presentation, slide 3. (FY2023 values interpolated).]
My table of forecast 'total change of revenue' over the period of interest is as follows.
FY2016 FY2017 FY2018 FY2019 FY2020 FY2021 FY2022F FY2023F FY2024F Fibre Revenue $133m $202m $276m $368m $466m $545m $715m $735m $755m Fibre Revenue Increment +$69m +$74m +$92m +$98m +$79m +$170m +$20m +$20m Fibre Revenue Increment Percentage +51.9% +36.6% +33.3% +26.6% +17.0% +31.2% +2.8% +2.7% Non-Fibre Revenue $875m $838m $714m $602m $493m $402m $332m $272m $222m Non-Fibre Revenue Decrement -$37m -$124m -$112m -$109m -$91m -$70m -$60m -$50m Non-Fibre Revenue Decrement Percentage -4.3% -14.8% -15.7% -18.1% -18.5% -17.4% -18.1% -18.4% Total Revenue $1,008m $1,040m $990m $970m $959m $947m $1,047m $1,007m $977m
Notes
1/ 'Non-fibre revenue' up until and including FY2021 has been calculated by subtracting 'Fibre revenue' from 'Total revenue'.
2/ Chorus have not provided any forecast as to where they expect their non-fibre revenue to go over the next few years. I have inspected the five year revenue trends for:
a/ 'copper connected revenue' and
b/ 'field services, value add network services and infrastructure'
as two groups. The latter group I am forecasting constant revenue of $120m over FY2022 to FY2024 inclusive (Actual figures were $118m over FY2020 and $119m over FY2021). The 'copper connected revenue', comprising 'copper based broadband', 'copper based voice' and 'data services copper' have taken an $91m decline over FY2021. So I am forecasting declining revenue to continue: down $70m over FY2022, $60m over FY2023 and $50m over FY2024. The 'diminishing decline rate' I am modelling to take account of a slowing trend as easy conversions to fibre have happened already.
--------------------
So what to make of this? IF you compare the actual fibre revenue over FY2021 of $545m with
a/ the forecast regulated fibre revenue for FY2021 (slide 24 PRHY2021) $620m (interpreted from graph), OR
b/ my linearly interpolated FY2021 fibre revenue of $592m (being the arithmetic average of the previous year's fibre revenue used figure of $468m and the prescribed first year of MAR $715m).
THEN that actual revenue represents $47m to $75m less revenue than forecast just six months ago. That seems to be an astonishing drop in projected annual fibre revenue in just six months, which has to call into question the forecasting ability of Chorus. Have I got that observation right?
SNOOPY
To get another angle on this, I have decided to rework my interest rate covenant calculations using EBITDA
I consider there are two ways to do this.
1/ The first is to consider what the EBITDA was in pre IFRS16 times. EBITDA does not take into account any interest charges. That means the 'lease interest' part of the 'current accounting standard' annual interest charge - which I class as a rent substitute - has not been accounted for. We need to pull 'the rent' - as a cost - out of the 'interest bucket' and put in back into the 'operating expense bucket'. This will decrease any ultimate interest rate charges reported. But it will decrease EBITDA as well, due to the extra cost of the rent. Nevertheless rent in any year is not exactly the same thing as 'lease interest'. We have to calculate annual rent from the operational accounts of the business, and that means looking at the cashflow statement. This I have done in the quoted post above, and calculated the rent to be $24m.
Furthermore the 'depreciation expense' under IFRS16 is now greater than it was. This is because there is now a new incremental element of depreciation - called the 'depreciation of right of use assets' (actually an IFRS16 accounting construct). Pre-IFRS16, total depreciation was less by this incremental amount. We must compensate for this smaller depreciation pre-IFRS16 if we are to compare today's EBITDA figures with yesterdays. This means for today's EBITDA to revert to the equivalent of the old EBITDA measure, we must increase today's EBITDA value by $15m (figure referred to in quoted post above), being the amount of this incremental depreciation.
With the adjustments in the above two paragraphs made, this means the EBITDA total reduces to:
EBITDA = $649m - $24m + $15m = $640m
I also need to make a complimentary adjustment to the denominator of my banking covenant, the 'Interest' divisor. Because 'lease expense interest' has been removed from the EBITDA figure quoted, to be consistent I also need to remove 'lease expense interest', from the divisor, because 'lease expense interest' is not interest in the pre IFRS16 sense.
For the interest bill I am using in this calculation (Ref AR2021 p41):
($153m - $1m) - $34m - $30m + $4m = $92m
Note that as well as ;lease interest expense', I have removed from the net interest bill:
(i) $34m of CIP securities 'notional interest'
(ii) $20m of 'lease expense' interest and $7m amortisation from a swaps reset and a further $3m undisclosed.
And I have added back a reduction in interest:
(iii) $4m from the ineffective portion of a cashflow hedge.
=> EBITDA/I = $640m / $92m = 7.0 > 2.75 => everything is A O.K.
2/ The second method is to use the EBITDA figure as quoted in the Annual Report. However in this case, the 'lease expired interest' that was equivalent to rent has not been removed. So to measure the ability of EBITDA earnings to cover any interest charges, we must make sure that those 'exaggerated earnings' are able to cover the consummately higher 'exaggerated interest bill' that includes 'lease expense interest' as part of the interest bill total.
EBITDA = $649m as reported.
I = $92m (as above) + $20m = $112m
=> EBITDA/I = $649m / $112m = 5.8 > 2.75 => everything is A O.K.
SNOOPY
For this table I have reverted to the 'conventional' calculations, (Not including the case where the discount rate of debt is zero).
Interest Coverage Debt to EBITDA EBITDA/I (Pre IFRS16) EBITDA/I (Post IFRS16) EBIT/I (Pre IFRS16) EBIT/I (Post IFRS16) 'Net Senior Debt'/EBITDA (Pre IFRS16) 'Net Senior Debt'/EBITDA (Post IFRS16) Value 7.0 5.8 2.3 2.0 3.88 4.23 Target >2.75 >2.75 >2.75 >2.75 <4.75 <4.75 Result Pass Pass Fail Fail Pass Pass
Notes[/
1/ 'EBITDA/I' calculations from my post 2810
2/ 'EBIT/I' calculations from my post 2803
3/ 'Net Debt' / 'EBITDA' calculations from my post 2802
Discussion
What to make of all of this?
There are two accepted ways of calculating 'Interest Coverage'. Since Chorus has not announced to the market that they have 'broken their banking covenants', I think it is fair to assume they are using EBITDA/I rather than EBIT/I. The banks would only accept this 'chosen' calculation method if near future capital expenditure costs were well below depreciation costs. Thankfully, because the assets of Chorus are generally 'long lasting' and nearer the beginning of their design life than at the end, this is the situation.
Post IFRS16, which is the standard that brings building leases onto the Balance Sheet, the 'Interest Coverage' ratio has deteriorated a little. This is a mathematical consequence of the 'Earnings', as a percentage of total earnings, in the numerator not increasing as fast as the 'Interest Payable', as a percentage of all interest payable, in the denominator. In fact under IFRS16, the numerical increment in both the numerator and denominator is the same, because the amount added to both ('lease interest payable') is the same. But as the old 'Net Interest Payable' starts from a smaller base than the old EBITDA, the effect of increasing both by the same numerical amount reduces the 'EBITDA' to 'I' ratio - in most instances. Putting my thinking cap on, if EBITDA was really low to the extent that the company was barely breaking even on a 'day to day' cash basis, but still had significant rent expenses (= 'lease interest payable' under IFRS16), then I could imagine a situation where adopting IFRS16 would improve the[/ Interest Coverage Ratio. However if things had got that bad, the bank would have probably intervened already. So I think it is fair to say that in the real world, adopting IFRS16 has made complying with any fixed Interest Coverage Ratio banking covenant a little tougher.
On the topic of 'Debt to Income' ratios, utility type companies with capital intensive assets yet relatively predictable cashflows tend to have high debts. This is usually the result of seeking 'shareholder capital efficiency', rather than underlying past managerial incompetence. Under IFRS16, multi year rent contracts that traditionally did not feature on balance sheets are now brought onto the balance sheet as 'lease payable' liabilities. You might imagine that:
1/ Having a future ten year capitalised rent bill on the numerator side of this particular covenant, with
2/ the 'offset effect' adding to EBITDA being just one year of 'lease interest' (a proxy for rent) increasing the denominator part of this statistic
might make it harder to repay the debt as depicted. In practice this is what has happened with the 'Debt to EBITDA' ratio getting [/worse on the adoption of IFRS16.
A curious point, which I discuss in post 2807 is that if I use EBITDA/I as 'Interest Coverage' this ratio becomes very closely coupled to the 'Net Senior Debt/ EBITDA' which is in effect a debt to income ratio. In fact if interest rates were permanently fixed at a single interest value, these two covenant ratios would be slightly different ways of looking at the same thing. In the real world interest rates do go up and down though. And I would suggest that interest rates are far more likely to go up from where we sit right now. So with today's very low interest rates, I would expect the Interest Coverage Ratio to be far less of a worry for bankers than the Debt to ear[/nings Ratio, simply because the Interest Coverage Ratio was designed in days where interest rates were generally much higher than today. It is likely the Interest Rate Coverage Ratio has been set by the banks (albeit unconsciously) expecting a rebound in interest rates. Sure enough when you compare the 'value' to the 'target' figure in the table above, EBITDA/I comes out at just over two times the target (a 100% safety margin). Whereas with 'Net Senior Debt'/ EBITDA, the safety margin is just (4.75-4.23) /4.75 = just 11%. Unless interest rate charges triple, we aren't going to get anywhere near the historic Interest Rate Coverage restrictions. So I guess this is why we don't hear much about 'Interest Coverage' as a lending covenant restriction these days. Perpetual very low interest rates have rendered 'Interest Coverage' largely irrelevant from a banker's investment perspective.
SNOOPY
About time. I thought Spark was being a bit misleading years ago when they rang and asked if I would like better broadband. I only clicked when they said I wouldn't lose my landline number but it would be linked to my mobile number.
Very deceptive playing on peoples ignorance. I only realised as I was looking at investing in Chorus.
Although Chorus lazily/stupidly thought it could drop the cables in the ground and wait for Spark and Vodafone to start signing up the customers.
I would have thought the horse has already bolted but with changes coming a few more people might be forced to choose. Hopefully Chorus can educate them before spark and vodafone sign them up to their wireless networks.
https://www.nzherald.co.nz/business/...C3XBTWTICUWW4/
It is a shame we live in a world that accepts lying and deception.
2022 Full Year results, annual and sustainability reports - NZX, New Zealand’s Exchange
Summary
• UFB uptake is 69 per cent; Chorus’ fibre rollout is 98 per cent complete
• 88,000 fibre connections added, a total of 959,000 connections
• Over 90% of fibre consumers are on 300 Mbps or above services
• Reported revenue was $965m (restated FY21: $955m)
• Earnings before interest and tax of $248m (FY21: $230m)
• Net profit after tax was $64m (restated FY21: $51m)
• FY22 dividend 35 cents per share; guidance for FY23 and FY24 increased
No imputation credit is a bit rough.
There are three key factors that determine how many imputation credits a company has to distribute to shareholders:
The proportion of profits earned in New Zealand. Companies that generate a substantial portion of their profits in New Zealand will likely have a more highly imputed dividend than those that derive profits offshore and therefore pay tax in other jurisdictions.
The amount of reported profit. Some companies have lower reported profits than the cash flow from which they can pay dividends. This could be due to, for example, high levels of depreciation which is a non-cash expense, lowering profit but not affecting cash flows. As profits are lower, tax is lower, and these companies therefore have fewer imputation credits to attach to their dividends.
Dividend payout ratios. Companies that pay out a lower amount of their profits as dividends are likely to have a more highly imputed dividend than those that pay out more.
Gettin edicated.
Good question from nztx, and a good explanation of possible answers from Nor. There is a fourth possible reason for no imputation credits too. That being that although the company is making a profit and has generated a tax bill, they haven't got around to paying the tax yet. So let's look and some data to try and get to the bottom of this.
FY2018 FY2019 FY2020 FY2021 FY2022 Total Net Profit Before Tax $122m $78m $73m $72m $106m Income Tax Expense ($37m) ($25m) ($21m) ($25m) ($42m) ($150m) Net Profit After Tax $85m $53m $52m $47m $64m $301m Tax rate 29.8%-30.9% 31.2%-32.9% 27.9%-29.7% 33.8%-35.7% 39.0%-40.3% Actual tax Paid ($30m) ($3m) ($12m) ($1m) ($14m) ($60m) Depreciation $283m $303m $319m $331m $335m $1,571m Capital Expenditure $810m $804m $663m $672m $492m $3,441m Operating Free Cashflow $508m $496m $474m $556m $570m $2,604m Dividends Paid $43m $49m $76m $86m $97m $351m Dividend Payout ratio 51% 92% 146% 183% 136%
Notes
1/ A fully imputed dividend may be paid if the tax rate has been assessed as an income tax expense at 28%, and the bill has been paid.
2/ I have calculated the actual tax as assessed within a range, as -given the raw data declared- it is not possible to determine what the assessed tax rate is any more accurately.
2a/ Sample Tax Rate Calculation As an example, over FY2022, the income tax expense was declared at $42m. However due to the rounding error applied to this figure, all we can say is that the actual quantum of tax assessed was between $41.5m and $42.5m. Likewise net profit before tax of $106m as declared, means that the actual net profit was between $105.5m and $106.5m. Thus the actual rate of tax paid was between:
$41.5m/$106.5m = 39.0% (lower bound)
$42.5m/$105.5m = 40.3% (higher bound)
We know for sure that the actual assessed tax rate was between those two percentage figures. But we don't know exactly where it sits. Why was the actual tax bill levied above the 28% levied on NZ companies? It can vary from year to year as the cash impact of a tax bill is not paid entirely in the year in which the underlying earnings were taxed. But generally if your tax bill ends up being higher than the standard tax rate, it means your actual income is shrinking (because provisional tax is based on previous years higher earnings). However earnings are not shrinking at Chorus. So I am mystified as tp why the booked tax rate remains higher than the legislated 28%.
---------------
Discussion
Chorus operates entirely within New Zealand. So we can rule out Nor's first possibility of profits being generated overseas not generating sufficient earnings within NZ to pay fully imputed dividends.
Nor's second conjecture that the available cashflow for distribution far exceeds profits looks to be true from the figures above. But this doesn't tell the full story. Alongside their day to day operations, Chorus have been indulged in a huge build out of their broadband cable network which has been highly consuming of their 'spare' cashflow. It is really only over the last year that Chorus's nationwide broadband roll out has neared completion. Considering the last five years, it is only over FY2022 that operating cashflow has exceeded capital expenditure, thus opening the window for higher dividends to be considered.
The converse of Nor's third conjecture, that companies that pay out more than their earnings as dividends will not have the imputation credits to cover those dividends also applies here. But that only explains why the rate of dividend imputation has to be reduced. It doesn't explain why imputation credits have disappeared entirely.
Finally my own conjecture that Chorus simply has not paid their assessed tax bills in full over recent years appears to also be true. How they have got away with not paying their tax is another unsolved question though.
At last it looks like we have the answer to nztx's question. Chorus must have had a stack of imputation credits that, until the dividend before last, allowed them to pay 'fully imputed dividends'. However once they burned through those, they didn't pay their tax on time and so had no imputation credits to pay out, notwithstanding the fact that one third of their dividend payout was from cashflow above earnings. (Cashflows in excess of earnings have no imputation credits attached to them anyway.)
SNOOPY
With the winding down of the 'Ultra fast Broadband 1' ('UFB1') and 'UFB2' fibre broadband network construction joint agreements with the government, for:
a/ The initial nationwide roll out of a fibre broadband network (UFB1) AND
b/ Subsequent extension of that to smaller centres (UFB2),
Chorus has now entered a new phase of its operation: Owner and operator of New Zealand's largest nationwide wholesale fibre broadband network.
Did I say nationwide? In fact there are are three substantial geographic regions where the fibre broadband infrastructure has been built by others:
1/ Greater Christchurch City: Built and owned by the Christchurch City Council owned 'Enable Networks'.
2/ Northern cities of Whangarei and Kerikeri, built and owned by the local electricity lines company 'Northpower', in turn 100% owned by electricity consumers of those northern regions.
3/ North Island central cities of Hamilton, Cambridge, Te Awamutu, Tauranga, Tokoroa, Hawera, New Plymouth and Whanganui built and owned by 'Tuatahi Fibre' (formerly 'Ultra Fast Fibre') owned by 'Igneo Infrastructure Partners'. Igneo is a sub brand of 'First Sentier Investors Group' created as a vehicle to hold and manage- at arms length - wholly owned businesses owned by First Sentier. Igneo specialises in owning mature infrastructure assets. First Sentier is a leading manager of global infrastructure assets, and manages capital on behalf of more than 120 institutions representing over 365 million pension fund members and policy holders from all over the world (although 37% of investors are geographically domiciled in Australia and 36% are in Europe).
Nevertheless Chorus is also owner of the legacy telecommunications copper network. Thus as the nationwide copper network is gradually and largely retired, we can expect the former copper network Chorus customers in the three regions listed above to migrate to these other fibre network operators.
Chorus faces broadband competition from the so called 'fixed wireless services' operated over the mobile networks owned by big NZ domestic telecommunications retailers Spark, Vodaphone NZ and 2 degrees. But the big telecommunications retailers remain resellers of Chorus's offering as well, an arrangement that can create perverse incentives on which broadband technology a consumer might be steered towards.
Growth prospects for Chorus are the 30% of homes that lie within the current Chorus fibre footprint but have not (yet?) chosen it. Furthermore there is a prospect of greater use of adjunct WIFI technology as more spectrum becomes available.
Conclusion: As
a/ The pre-emanant broadband provider of all broadband technologies by market share AND
b/ Monopoly fibre provider in all regions in which Chorus has their fibre rolled out, AND
c/ As monopoly provider nationwide of the legacy copper technology,
Chorus is the number one player in all markets in which they choose to participate. 'PASS TEST.'
SNOOPY
Time to update the earnings per share trend for the current financial year.
Earnings per share are calculated by taking the normalised net profit after tax and dividing that by the number of shares on issue at the end of the financial year.
FY2018: ($85m + 0.72( $5m+$5m+$7m)) / 429.641m = 22.6cps
FY2019: ($53m + 0.72( $1.5m+$2m+$3m+$3m+$6m+$2m )) / 439.288m = 14.9cps
FY2020: ($52m + 0.72( $2m+$6m+$5m+$2m+$1m+$5m+$3m-$3m ) - $5m) / 444.492m = 13.9cps
FY2021: ($47m + 0.72( $2m+$1m-$4m+$7m ) +$1m) / 447.025m = 11.7cps
FY2022: ($64m - $9m -$3m - $3m + 0.72(-$3m+$1m+$1m+$7m-$7m) ) / 446.512m = 10.8cps
Notes
1/ Normalised FY2018 result removes a $5m labour restructuring charge, removed a $3m interest charge realised from a reset of a GBP Euro Medium Term Note (EMTN) interest rate swap (the interest rate exposure is only partially hedged, explaining the need for a reset). A further interest charge of $7m, based on the amortisation of the ineffectiveness of the EMTN cashflow hedge (closed out on 9 December 2013), has been added back.
2/ Normalised FY2019 removes $1,5m of labour restructuring costs, $2m of consultants fees investigating the forthcoming regulatory regime, and $3m from a set aside implementation charge to get the new regulatory framework in place. Removed a $3m interest charge realised from a reset of a GBP Euro Medium Term Note (EMTN) interest rate swap (the interest rate exposure is only partially hedged, explaining the need for a reset). A further interest charge of $6m, based on the ineffectiveness of the EMTN cashflow hedge has been added back. A $2m one off expense for restructuring two forward dated interest rate swaps has also been removed.
3/ Normalised FY2020 removes a combined $2m Covid-19 relief payment encompassing Fibre and Copper broadband customers, a $6m increase in Covid-19 staff leave provisions and contractors to help make the transition to the new regulatory framework, a $5m payment to contracted service companies to help them through the lock-down periods (adding to a total $13m of Covid-19 relief response), an incremental $2m increase in consultancy fees related to the regulatory transition, a $1m one off expense for restructuring forward dated interest rate swaps. I have removed a $5m interest charge, realised from a reset of a GBP Euro Medium Term Note (EMTN) interest rate swap (the interest rate exposure is only partially hedged, explaining the need for a reset). A final further interest charge of $3m, based on the ineffectiveness of the EMTN cashflow hedge has been added back, and at EOFY2020 and, this 9 December 2013 transaction, was finally 'closed out' of the account books. I subtract $5m from NPAT, to account for the one off $5m tax refund resulting from the resumption of building depreciation. Finally my normalised FY2020 now removes a favourable $3m one off settlement (not mentioned in AR2020 specifically, but see AR2021 p17))
4/ Normalised FY2021 profit adjusts for a $1m value Added Network one off dispute settlement (assumed non taxable), a $2m one off restructuring labour costs, as a result of the ongoing transition from a 'build it' to a 'manage it' network, and $1m of costs relating to the decommissioning of legacy copper network equipment in Spark exchanges. I have subtracted a $4m reduction of interest costs relating to the "Ineffective portion of changes in fair value of cash flow hedges" and added back $7m interest charge arising from interest rate swap resets.
5/ Normalised FY2022 profit subtracts $9m, from a one off benefit following a judicial review of a one off interpretation of the Holiday's Act, subtracts $3m from a surplus property sale and subtracts $3m form a legal settlement. Further adjustments where tax effects should be considered include a $3m benefit from a change of lease contract (normalisation reverses this), adding back $1m of make good costs relating to office moving and renovation, adding back $1m spent on external consultants advising on transitioning to the new regulated operating framework, adding back the $7m finance expense credit resulting from ineffective cash flow hedging and subtracting $7m of finance expenses relating to an adjustment of interest swap resets.
Conclusion: FAIL TEST
SNOOPY
Normalised profit is divided by shareholders equity at the end of the financial year
FY2018: $97.2m / $1,022m = 9.5%
FY2019: $65.6m / $979m = 6.7%
FY2020: $62.0m / $927m = 6.7%
FY2021: $52.3m / $948m = 5.5%
FY2022: $48.3m / $1,029m = 4.7%
At no time over the last five years has 'Return On Equity' exceeded 15%. And I don't like the multi-year trend.
Conclusion: FAIL TEST
SNOOPY
'Net Profit Margin' is the 'Normalised Net Profit After Tax' divided by 'company sales' over the financial year
FY2018: $97.2m / $990m = 9.8%
FY2019: $65.6m / $970m = 6.8%
FY2020: $62.0m / $959m = 6.5%
FY2021: $52.3m / $947m = 5.5%
FY2022: $48.3m / $965m = 5.0%
Pretty hard to see any ability to raise margins when the trend is relentlessly down over five years.
Conclusion: FAIL TEST
SNOOPY
Ok time to come clean. Evaluating Chorus against he Buffett criteria was a set up for failure. Why? Because the key to a Buffett style halo company is efficient use of assets. Clearly if you have just made a massive capital investment, as Chorus have done in their broadband network, -ahead of demand-, that network will not be used efficiently from day 1. Furthermore other 'network companies', like the power distribution companies, have built up their assets over decades. Many of those assets will be on the balance sheet at 'historical cost'. It doesn't take a genius to figure out that if your costs are largely historical, but your income is in present day dollars, that is a superior business model to Chorus.
Like the gentailers, Chorus is now a 'cashflow story'. If we evaluate Chorus against the more traditional metric of 'underlying earnings' (or 'normalised earnings' as I put it), Chorus, at $7.85, is currently trading on a PE ratio north of 70. No that is not a misprint!
If we look at the headline profit trend, Chorus is now climbing out of the profit dip that occurred when the network building costs were at their peak. However, once I normalise that profit for each year, this is not the case. Normalised Chorus profits hit new lows over FY2022. The fact that imputation credits on dividends have vanished is indicative of a company making next to nil real money in 'profit' terms. I contacted the CFO of Chorus about this and he claimed it was a 'depreciation issue', and that ultimately imputation credits would return. He was very coy as to when that might be though!
For reasons I have just outlined, I am not too worried about return on equity being low. What is of more concern is the ever reducing net profit margin on revenue that is not growing. I can forgive profit going down if it is a depreciation issue. But I am concerned about the flat revenue. One aspect of this is that -perversely-, the more people nationwide that move to fibre broadband (the barrow that Chorus is pushing), the more people who will move away from Chorus copper onto one of the regional fibre networks that Chorus does not own. And when 'regions' not fiberised by Chorus encompass such populous areas as Christchurch and Hamilton, that will hurt Chorus.
From my brief brush with members of the Chorus leadership team, my feeling is that the company is being run on a 'build it and they will come' mentality. There is nothing wrong with that, provided forward sales projections are realistic. The Australian equivalent company, NBN (National Broadband Network) is not so technology tunnel visioned. NBN use a couple of geo-centered satellites and wireless broadband to cover remote communities. So NBN is a 'technology agnostic' company. Personally I think the threat of wireless broadband is being underestimated by Chorus, and that they will have trouble convincing many 'hold outs' from fibre to switch over. On top of this we have the heavy hand of government regulation that caps overall revenue that may be taken by Chorus over a financial year. And the, according to Chorus, unrealistic cost of capital assumptions that have lead to such revenue caps.
The wrap: Chorus is clearly not a Warren Buffett style target investment. Whether it is a good cashflow style investment will require some more digging on my part. Stay tuned.
SNOOPY
Some more thoughts on this more than 18 months on.
I have established that all of the Crown Infrastructure Partners (CIP) funding that has been used to partially fund the construction of the Chorus broadband network will have to be repaid to the crown in full. For IFRS accounting reporting purposes, interest free funding (which is what the CIP funding is, until repayments become due) is deemed to distort the operational picture of Chorus. Thus, for IFRS reporting purposes, Chorus is forced to introduce the concept of 'notional interest' (which in the operating world is never charged nor paid, as evidenced by the fact that the IRD has ruled it is not tax deductible) to reflect what would happen if these funds were borrowed on more commercial terms.
Whatever is presented via the annual report to Chorus shareholders, the equivalent year annual accounts submitted to the IRD will be different set: a set with all of the 'notional entries' removed (or more correctly never created in the first place). This is IMV, one of the reasons that Chorus is 'apparently' paying a higher tax rate than the statutory 28% corporate rate. One driving part being the 'notional interest' that is so carefully added to the interest bill in the shareholder accounts, - is (from the perspective of the IRD), actually supplementary 'hidden profit' that needs to be taxed. More correctly 'notional interest' should be referred to as a non-deductible expense. IOW it is an expense in the books that cannot be claimed as a deduction for the purposes of calculating the year end income tax expense and would not form part of anything submitted to the IRD. Simplistically, the tax calculation is expressed: NPBT + non-deductible items = taxable profit.
Similarly the 'reduction in depreciation', brought about by crown funding via CIP, I do not believe is real either. (AR2022 p20: "depreciation credit recognised in the profit and loss in relation to CIP securities are non‑taxable"). It is untenable to me to imagine that the boffins at IRD should agree that the rules on depreciation of assets should be altered, simply because some of the funding of the building of those assets is 'interest free' via CIP. From a Chorus shareholder perspective, this means that the 'discounted' portion of depreciation is being time shifted further out into the future. The discounted depreciation will still be claimable in the future. But for now, profits declared today from a Chorus shareholder perspective are increased because of this IFRS mandated adjustment.
My next step will be to create an 'alternative profit picture', similar to that I believe is seen by the IRD, where I add the notional interest back onto income and restore the depreciation claimed back to what would have been claimed had no CIP implied 'depreciation discounts' ever existed.
SNOOPY
FY2018 FY2019 FY2020 FY2021 FY2022 Total Net Profit After Tax (Declared) $85m $53m $52m $47m $64m add CIP Notional Interest Deducted 0.72x $17m 0.72x $22m 0.72x $29m 0.72x $34m 0.72x $39m subtract Depreciation Discount 0.72x ($22m) 0.72x ($25m) 0.72x ($27m) 0.72x ($29m) 0.72x ($27m) equals IRD Adjusted NPAT $81m $51m $53m $51m $73m Implied Income Tax Expense (@28%) ($32m) ($20m) ($21m) ($20m) ($28m) add Reinstatement of tax on Building Depreciation $5m subtract Prior Period Tax Adjustment ($6m) add Other Non-taxable adjustments ($2m) ($3m) ($5m) ($5m) ($6m) equals IRD Adjusted Tax total (As calculated) ($34m) ($23m) ($21m) ($25m) ($40m) ($143m) Income Tax Expense (Declared) ($37m) ($25m) ($21m) ($25m) ($42m) ($150m) Tax Paid Actual (Declared in cashflow) ($30m) ($3m) ($12m) ($1m) ($14m) ($60m) Current Tax Expense (Tax Note 14) ($16m) ($6m) ($1m) ($1m) $3m ($21m)
Notes
1/ To calculate the 'Implied Income Tax Expense' at 28% from NPAT: (NPAT)/(1-0.28) - (NPAT) = 'answer'
2/ Numbers referred to in the above table as 'declared' are from the respective Profit & Loss statements.
3/ 'Other non-taxable adjustments are taken from section 14 'Taxation' of the respective annual reports.
--------------
Discussion
Please note that I am using some of the information from Ferg's post 2826 (thanks Ferg) to more correctly explain some of the terms and concepts in more recognised accounting terms.
I am working on the assumption that 'income tax declared' cannot be fudged by IFRS. IOW the amount of tax paid to the IRD is the same, no matter what perspective a company takes on how, and over what timeframe, other costs should be attributed. That means the figures of the second to last two rows of the table above should be the same.
One explanation of a deviation in the numbers reported between each of the two referred to rows could occur because I haven't accounted for tax payments from one year spilling over into adjacent years. Specifically I am thinking about payments made in advance called 'provisional tax' and payments in arrears called 'terminal tax'. The timing of such tax payments differ to the income tax expense declared in the annual report. But I am hoping that because Chorus is a closely defined predictable business, any such errors caused because of tax payment timing will be minor.
'Other non-taxable adjustments' (as tabulated above) that increase the actual tax paid above the statutory rate are unspecified but could include items like:
a) Permanent differences which are non-deductible items (such as some legal fees, 50% of entertainment, some IFRS adjustments etc). Note that I have included one class of permanent difference 'notional interest' in its own separate row, further up the table, AND
b) Timing differences claims which vary 'year to year' and usually unwind to a net impact of zero over a long enough time horizon (e.g. asset depreciation may be accelerated or retarded for IRD purposes vs accounting purposes, and the deductibility of holiday pay liabilities are subject to non-straightforward rules etc.) Note that I have reversed the contrary 'slowed depreciation rate', that is a consequence of CIP crown funding, in a separate line in the table above.
A separate issue is the 'Rural Broadband Initiative' (RBI) assets, funded by non‑taxable government grants. The accounting amortisation of RBI government grants and RBI accounting depreciation recognised in the profit and loss are non‑taxable and tax depreciation is not claimed (Refer AR2022 p20).
Conclusion
My measure of how well I understand the tax, and hence tax credit, position of Chorus is how well the third and fourth to last lines in the above table agree. FY2020 and FY2021 are in agreement, FY2022 and FY2019 are out by $2m and FY2018 is out by $3m. Reporting rounding errors could account for around $1m of those differences. Temporary differences from moving tax expenses could account for the rest. However, if temporary differences were the explanation, I would expect that over the years these differences in overpayments and underpayments would balance out. If you inspect the picture of the last five years by comparing the totals of the second and third to last table rows, it looks like underpayment of income taxes is entrenched.
Furthermore I would expect a Rural Broadband Initiative (RBI) effect, where construction is funded by grants and no depreciation is claimed. From AR2022 p20:
"RBI assets were funded by non‑taxable government grants. The accounting amortisation of RBI government grants and RBI accounting depreciation recognised in the profit and loss are non‑taxable and tax depreciation is not claimed."
If depreciation of RBI is not claimed, does that not mean that profits will increase in that arm of the business? Thus more tax will be paid on any RBI profits, thus increasing the tax bill above the statutory 28% rate?
I admit that I am speculating and I don't fully understand what is happening with the tax bills here. Nevertheless, I consider that I have identified enough 'wriggle room' in the taxation picture that could explain why my calculated 'IRD Adjusted Tax total' and the declared 'Income Tax Expense' are not in perfect agreement. Thus I believe my adjusted net profit figures, where notional interest and depreciation discounts are removed, provide a better representation of where the company is headed profit and tax wise than the IFRS reporting sanctioned and massaged 'official' NPAT figures.
SNOOPY
Snoopy
I imagine the CNU tax will take quite some time to unpick so I'm not going to try, but I have a couple of pointers.
1/ Ensure your phraseology is correct in that tax "declared" and tax "paid" will be 2 different numbers. Payments made in advance are called provisional tax plus there are payments in arrears called terminal tax. The timing of such tax payments differs to the income tax expense declared in the annual report. Information on payments will be sourced from the cash flow and should not go anywhere near your analysis (I suspect they have not). Instead you want to work with the numbers in the P&L and the notes.
1a/ I would say ignore the Balance Sheet numbers and accompanying notes - this is where timing differences between the declared expense in the P&L and tax payments appear. But on reflection, if this shows the amount declared to the IRD for the year, before the deduction of any payments for that year, you might get the amount declared to the IRD from these notes.
2/ When you say this:
Quote:
One driving part being the 'notional interest' that is so carefully added to the interest bill in the shareholder accounts, - is (from the perspective of the IRD), actually supplementary 'hidden profit' that needs to be taxed.
Whilst the concept is correct, the phraseology is not quite right. It is not "hidden profit" per se; rather it is a non-deductible expense. IOW it is an expense in the books that cannot be claimed as a deduction for the purposes of calculating the year end income tax expense and would not form part of anything submitted to the IRD. Simplistically, the tax calculation normally looks like this: NPBT + non-deductible items = taxable profit.
3/ Remove the amount (if any) classified as deferred tax that has been included in the income tax expense note that supports the income tax expense value in the P&L. Normally it is split between "deferred tax" (more on that later) and "current tax". You want to work with the "current tax" value only. I believe that will usually be the figure declared to the IRD, unless there are other 1 off adjustments which I see you have noted. Per note 1a above this should match the value that was added to the "income tax liability" account per the notes to the Balance Sheet.
4/ There will be other non-deductible expenses (much like the CIP interest) that will muddy the waters. There are two types, being
a) permanent differences which are non-deductible items (such as some legal fees, 50% of entertainment, notional interest, some IFRS adjustments etc), and
b) timing differences claims which vary year to year and usually unwind to a nett impact of zero over a long enough time horizon (e.g. asset depreciation may be accelerated or retarded for IRD purposes vs accounting purposes, and the deductibility of holiday pay liabilities are subject to funny rules etc.) Such timing differences are usually detailed in the deferred tax notes and these differences give rise to the deferred tax liability (and in some cases a deferred tax asset).
I suspect IFRS 16 might also muddy the waters. And sometimes there is non-assessable "income" (per the RV's) etc. which is unlikely in this case. But the point is that there may be any number of adjustments which are not separately disclosed in the accounts, but you may find a summary of these in the notes to the accounts that support the value for "income tax expense" in the P&L.
5/ I recommend find the tax liability declared to the IRD, which is probably the "current tax" portion of the income tax expense (note this is not the prima facie value per the tax calc note). Once you have that work backwards to see what is the taxable NPBT versus the declared NPBT per the annual report. Then work on pre-tax values for the differences between the two (pre tax is easier) as CNU will have done in their note supporting income tax expense in the P&L. Also double/triple check you have all your signs round the right way.
6/ I presume CNU is 100% NZ income and has no income from a higher tax jurisdiction? If there is overseas income, this could be at a different tax rate.
Lots to unpick so good luck.
Yes the 'declared profit' I tabled was straight from the respective 'Consolidated Income Statement(s)'.
The tax note (AR Note 14) contains 'Current Tax Expense' split between the 'current year' (which I am guessing is provisional tax) and 'adjustments in respect of prior periods' (which I am am guessing is terminal tax). I have summed these two figures for each of the five years and added the totals to the bottom of my table, as a curiosity. However, these figures are not representative of the 'Declared tax' from the income statement.
I have added your corrected description (thanks) to my original note, to make sure there is no misunderstanding as to what I am talking about.
SNOOPY
FY2018 FY2019 FY2020 FY2021 FY2022 Total Deferred Tax Expense (Adjustments wrt Prior Periods) $4m $2m ($14m) ($8m) add Deferred Tax Expense (Depreciations, provisions, accruals, leases & other) ($25m) ($21m) ($20m) ($24m) ($31m) ($121m) add Current Tax Expense (Tax Note 14) ($16m) ($6m) ($1m) ($1m) $3m ($21m) equals Income Tax Expense (Declared) ($37m) ($25m) ($21m) ($25m) ($42m) ($150m)
Notes
1/ Deferred tax figures taken from Note 14 'Taxation'
Yes, no overseas income. All income billed by Chorus comes from New Zealand sources.
SNOOPY
The bold line per your analysis titled "Current Tax Expense" is what has been declared to the IRD. That is exactly the line you want to work with if you are trying to work out taxable profit etc. I recommend take the $3m refund for the last fiscal year, divide by 0.28 to get a pre tax loss of $11m. I think you are trying to reconcile that number to the pre tax profit figure per the annual report. Given that is a loss and prior years are so low, that will explain why there are no imputation credits - IOW there is no taxable profit hence no tax payments -> no imputation credits.
Per this earlier post:
Quote:
The tax note (AR Note 14) contains 'Current Tax Expense' split between the 'current year' (which I am guessing is provisional tax) and 'adjustments in respect of prior periods' (which I am am guessing is terminal tax). I have summed these two figures for each of the five years and added the totals to the bottom of my table, as a curiosity. However, these figures are not representative of the 'Declared tax' from the income statement.
WRT a specific financial year, provisional tax payments are payments made in advance on the current tax amount. And terminal tax payments are payments made in arrears also on the current tax amount. Deferred tax is never paid - that is why it is 'deferred'.
This very sobering moment in the analysis of Chorus's FY2022 result deserves its own one liner, particularly in the context of the AGM addresses which were along the lines of: "Dammit we are doing well!"
"Pre-tax profit declared to shareholders: $106m. Pre tax loss declared to Inland Revenue: ($11m)"
What the ..??##$$!!!!!!.....
SNOOPY
You think right, and here are the last five years of results as declared to the IRD, compared to what shareholders were told.
FY2018 FY2019 FY2020 FY2021 FY2022 Total Current Tax Expense (Tax Note 14) ($16m) ($6m) ($1m) ($1m) $3m ($21m) Pre-tax profit declared to IRD $57m $21m $3.6m $3.6m ($11m) $74m Pre-tax profit declared to shareholders $122m $78m $73m $72m $106m $451m
Notes
1/ Pre-tax profit declared to IRD = 'Current Tax' / 0.28
2/ Pre tax profit declared to shareholders is from the respective Annual Reports
------------------
I have to admit that I am in shock at the diversity of the two profit pictures being prepared: For the IRD and shareholders respectively.
SNOOPY
Wow, that is a lot of timing and permanent differences. Very sobering indeed.
That is too much (and the wrong way round) to be attributable to non-deductible expenditure.
There must be something classified as revenue in the P&L that is not income that is assessable for income tax. It might be those Government grants we discussed some months/years ago - perhaps they are being counted as income in the AR but are not taxable income...?
IOW : reported profit - non taxable income = taxable profit
I presume there is only the NZ tax jurisdiction? (which means the /0.28 method is correct) Secondly, I'm guessing there are no imputation credits - is that correct and is that what prompted your analysis?
Yes, all income was earned in the NZ tax jurisdiction. And yes, every Chorus dividend since the first in 2012 has been faithfully fully imputed. Except for this last dividend paid on 11th October 2022 where the imputation credits suddenly disappeared completely. And despite forecasting increasing dividends into the future (HYR2022 p1), there is no expectation that any form of dividend imputation will return within the next two years.
Yes the AR2022 accounted for CIP funding 'interest repayments', which we have both agreed are 'not real' (merely accounting constructs), mean that over the current year the 'declared profit', as supplied in the income statement of the Annual Report 2022 will be lower than the IRD submitted profit. But offsetting that effect, is the 'depreciation discount' which has also come into the accounts because of crown funding. If the depreciation is reduced over a year, that means the resultant 'declared net profit', as seen in the annual report, will increase over and above what it otherwise would be. But I don't believe the 'depreciation discount' is real either (see my post 2825 as evidence of this). The 'depreciation discount' in the annual accounts means that the AR2022 income statement profit is higher than it would otherwise be.
I have a feeling a third factor could be coming into play here as well: accelerated depreciation. Chorus are saying with the success of their broadband roll out (p19 AR2022)
"With the commencement of the Chorus copper withdrawal programme, Chorus has revised the depreciation profile of copper cables in areas where there is fibre available (Snoopy comment: i.e. most areas). Depreciation of copper cables will be accelerated from FY23 so that those in UFB2 areas will be fully depreciated by June 2025 and those in local fibre company areas will be fully depreciated by June 2026."
That follows on from a similar comment made on p19 of AR2021.
"Chorus has considered the useful life of copper cables in UFB1 and UFB2 areas. Due to strong fibre uptake depreciation of these cables is being accelerated at a rate of approximately $11m and $4m per annum respectively. This means copper cables will be fully depreciated by UFB1 by 30 June 2025 and UFB2 by 30 June 2027."
Given the tone of those comments, I am not sure why the extra depreciation on copper in UFB1 areas has been rolled back by a year (now being fully depreciated by 30-06-2026, not 30-06-2025). But I suspect that, short of Chorus declaring itself bankrupt, the IRD would not allow any such 'accelerated depreciation' in accounts submitted to them. If Chorus 'declares' to shareholders that their depreciation is much higher than it really is, that means that Chorus will also 'declare' to shareholders a profit much lower than it really is (from an IRD perspective). And we all know the IRD is the ultimate 'pulpit of truth' for finances, right?
So if Chorus really has slowed again their copper depreciation in UFB1 areas (i.e. most of the country) over FY2022 (as indicated in my two quotes above), then that would increase their AR2022 reported profits (while there would be no change in their IRD reported profits).
Yes, I agree, I must be missing 'something'. Perhaps if I produce a detailed income statement from both FY2018 and FY2022, that would offer a clue as to where such 'non taxable' revenue might have come from?
SNOOPY
Income Statement Comparison
FY2018 FY2022 Operating Revenue Fibre $276m $614m Operating Revenue Copper $581m $211m Operating Revenue Backhaul Infrastructure $126m $128m Operating Revenue Other $7m $12m Operating Expenses ($337m) ($290m) EBITDA $653m $675m Depreciation Fibre Cables ($78m) ($122m) (*?) Depreciation Copper Cable ($51m) ($61m) Depreciation Network Electronics ($65m) ($62m) Depreciation Other ($111m) ($117m) (*) Depreciation Crown Discount $22m $27m Amortisation (Software & Customer Retention Assets) ($104m) ($92m) EBIT $266m $248m Finance Income $7m $0m Finance Expense (excluding CIP notional interest) ($134m) ($103m) (*) Finance Expense: CIP notional interest ($17m) ($39m) NPBT $122m $106m Income Tax Expense (Current) ($16m) $3m (*) Income Tax Expense (Deferred) ($21m) ($45m) NPAT $85m $64m
Notes
1/ The above numbers are sourced from the respective annual reports.
2/ The lines marked with an asterisk, I suspect are the ones that the IRD would regard as 'not real' for income tax purposes.
3/ p19 AR2021 notes that depreciation of the copper network has been increased by a total of $15m from FY2021. So it could be that $15m of that copper cable depreciation for FY2022 is 'not real' either (from an IRD perspective).
4/ Depreciation of fibre cables is not an 'apples with apples' comparison between FY2018 and FY2022..
Operational Year FY2016 FY2017 FY2018 FY2019 FY2020 FY2021 FY2022 Capital Expenditure Fibre $486m $503m $620m $664m $548m $567m $403m
The above table shows that since EOFY2018 the $1.782b of fibre on the books has more than doubled in value, with an extra: $664m + $548m + $567m + $403m = $2,182b in fibre investment made over the four ensuing years. We expect the depreciation charge of fibre to increase consummately over that time.
5/ Total debt (excluding CIP debt) was $2.322b at EOFY2022, verses $1.807b EOFY2018. Despite the 29% increase in the debt balance over four years, the weighted average interest rate (including the effect of derivative financial instruments and facility fees) dropped from 5.96% to 3.77% over that same period. That equates to a 37% reduction in debt servicing costs on a constant balance basis. The net effect being 1.29 x 0.63 = 0.81, a 19% fall in debt servicing costs.
---------------
Discussion
The purpose of looking at these two income statements five reporting years apart is to answer the question:
"What has changed, that has seen Chorus go from a fully imputed dividend payer, to a company that has no imputation credits?"
The trite answer is "Because they haven't paid enough tax." But why would that be? I was hoping a 'line by line' comparison between an income statement from Chorus's full tax paying days back in FY2018, and FY2022, where apparently no tax was owing, might provide a clue. Unfortunately after going through the two income statements line by line, I am none the wiser about the 'time shifting' or the 'quantum of tax payments' changing over that four year comparative period.
If I adjust the 'declared net profit' by the asterisked row entries in the table, then I get an IRD sanctioned net profit before tax for FY2022 of:
$106m + ($15m - $27m +$39m) = $133m
So my adjusted profit figure of $133m lines up against the tax figure implied NPBT figure of an ($11m) loss. Needless to say my reconciliation has not gone well :-(.
SNOOPY
The difference between 'total capital expenditure' and 'sustaining capital expenditure' is now critical in determining the future dividend returns for Chorus shareholders. From the 17th November 2020 Presentation, Slide 33:
"From FY22 we will transition to a dividend policy based on a pay-out range of free cash flow
▪ free cash flow will be defined as net cash flows from operating activities minus sustaining capex."
The following table is compiled from information in:
a/ Slide 40 of the 22nd August 2022 presentation.
b/ AR2022 p21
Capex over 2022 Sustaining Development Sub Total Fibre Layer 2 Sustaining $29m Fibre Products & Systems Sustaining $7m Fibre Network & 'Other Fibre Connections' Sustaining $23m Fibre Customer Retention Sustaining $13m Fibre UFB Communal Development $77m Fibre Connections & Layer 2 Development $166m Other Fibre Connections & Growth Development $74m Fibre Customer Retention Development $14m Fibre Total $403m Copper Network Sustaining $27m Copper Layer 2 Sustaining $3m Copper Customer Retention Sustaining $7m Copper Connections Sustaining $1m Copper Development $0m Copper Total $38m Common IT Sustaining $32m Common Building & Engineering Services Sustaining $19m Common Development $0m Common Total $51m Sustaining Total $161m Development Total $331m Overall Total $492m
The year to year comparison with the quoted post above shows a $19m reduction is 'sustaining capital expenditure'. That includes $7m in sustaining costs less spent on the copper network. So sustaining capex for the fibre network has gone down over the year by $12m. This is an indicator as to why management has decided they can increase their dividend for FY2022. The reason being that -at last- the broadband network is generating more cash than it consumes
Dividends going forwards are based on 'free cashflow'. Note that 'based on free cashflow' does not mean paying out 100% of free cashflow as dividends!
Free cash flow = net cash flows from operating activities - sustaining capex = $570m-$161m = $405m
On a per share basis: $405m/446.512m = 92cps.
This more than covers the 14cps (or 14cps/0.72 = 19.4cps before tax) and 21cps unimputed dividend based on FY2022 free cashflows: 92cps - (19.4cps + 21cps) = 51.6cps 'cashflow left over'.
That cashflow also makes the 40cps dividend forecast for FY2023 and the 45cps dividend forecast for FY2024 look do-able (even if probably without imputation credits). Although I do note that projected increase in dividend over the year from 35c to 40c for 2023 will be entirely negated by the loss of imputation credits! At today's $7.83 closing share price, that means we are looking at a forecast gross yield of 5.1% for FY2023, rising to 5.7% for FY2024.
SNOOPY
I had a brief chat with senior management about the upcoming $170m of CIP debt and CIP preference share maturity coming in 2025. The thinking at this stage is that both the CIP debt and the CIP preference shares will be refinanced with other debt, with no net company debt reduction. There is no hurry to do this as no commercial deal can match 'interest free' government funding. But the CIP debt coupon rate, when it comes into force in 2025, is the 180 day bank bill rate plus 6% (a condition set eleven years ago). I can no longer find any reference to a 180 day bank bill rate in NZ (anyone know what happened to this metric?). But taking the 90 day bank bill rate as a proxy, (4.17% as I write this), that equates to a coupon rate of over 10% - positively usurious, even in these days of higher interest rates.
Chorus seemed pretty pleased with their most recent EURO note financing deal, taken out over FY2020: EURO 300m in notes at 0.88%, with repayments fully hedged back to the NZD (from FY2021 at least). The implication being that -come 2025, and the first CIP coupon activation date-, Chorus will be able to strike a similarly good refinancing deal. I sure hope that is true. And hope is a legitimate investment strategy - I learned that from this forum!
International agencies seem keen to give Chorus more and more debt rope. This was the last remark in the 'capital management' section of HYR2022 p3
"In early 2022 Moody's Investors Service and S&P increased their debt/EBITDA credit rating downgrade thresholds to 5.25 times and 5.0 times respectively.for Chorus. Following this increase, it is Chorus's intention that in normal circumstances the ratio of net debt to EBITDA will not materially exceed 4.75 times."
This remark comes after Chorus and their bankers had previously announced revised banking syndicate arrangements on 25th May 2017.
"The facility has also been repriced to reflect current market rates and the covenants have been revised from 4.0 to 4.75 times debt to EBITDA and 3.0 to 2.75 times interest coverage, to better align with Chorus’ rating thresholds."
It looks like those bankers and their credit rating sidekicks have raised their 'borrowgestioning' strategy as regards Chorus to new heights: "Borrow, borrow, borrow!" "There is no risk for we bankers here." (as long as you pay us your interest payments on time, of course).
SNOOPY
AR2022 p20 tells us that :
1/ 'Other interest expense' includes $15m of 'lease interest', from a total of $23m.
Our objective: To ensure that the effect of 'lease interest' is not 'double counted' when calculating banking covenants. 'Lease interest' is part of what used to be considered 'rent' in the pre-IFRS16 days. So when we use EBITDA as part of a banking covenant calculation, we should calculate this for Chorus as : 'Declared EBITDA' - 'Lease Interest'.
For Chorus over FY2022, this means, for banking covenant purposes: EBITDA = $675m - $15m = $660m
SNOOPY
This has become a bit of an 'old chestnut' of mine: The representation of the 'Crown Infrastructure Partner Securities', incorporating both 'CIP Debt' and 'CIP Equity' (actually preference shares) on the Chorus balance sheet. First of all, hats off for Chorus listing the 'preference share equity' as debt. But as for the way the balance is discounted back to present day values using an ultra high (IMO) 8.5% discount rate....
CIP Debt Facilities
Chorus have used the face value of the debt facilities issued (Slide 27, August 2022 Presentation) and I have used an 8.5% annual discount factor (AR2021 p43, discount rate not disclosed in AR2022) to get their 'present value' of CIP debt:
$85m / 1.085^3 + $90m / 1.085^8 + $137m / 1.085^11 + $174m / 1.085^14 = $224.7m
The undiscounted value of that CIP debt is: $85m+$90m+$137m+$174m = $486m, more than the total value of the UFB1 debt. The UFB1 construction project is now complete, with $462m of CIP debt incurred. Therefore, I presume any additional debt is because some UFB2 debt has been brought under the 'Chorus Debt Facilities' umbrella?
On a separate note, I see the UFB2 equity drawn by Chorus has increased from $265m at EOFY2021 (PR2021 s21) to $306 + $24m = $330m as at EOFY2022 (PR2022 s17), an increase of $65m. At the same time the UFB2 roll out passed another 1,324,000 - 1,282,000 = 42,000 premises during the financial year (PR2022 slide 5).
CIP Equity Facilities
The total of $613m on the balance sheet listed as "Crown Infrastructure Partners (CIP) securities" represents both 'CIP equity' ($346m) and 'CIP debt' ($267m) (See AR2022 p44). The 'CIP equity' is actually preference shares, which in my view are more closely classed as a form of debt (it is listed as a liability in the balance sheet after all). Nevertheless, the listed total "Crown Infrastructure Partners (CIP) securities" liability of $613m is discounted from the ultimate repayment amount, as accounting standards allow. Nonetheless all of this 'CIP debt security' 'discounted debt' will eventually have to be repaid at face value. As for the 'CIP equity' (preference shares) these will either:
1/ Have to be repaid at face value OR
2/ Be settled by converting the preference shares to ordinary shares, and that process will diluting existing shareholders' interests. OR
3/ Be refinanced as additional debt on commercial terms.
This debt does not come into any banking covenant calculations, because the bank as they are entitled to do, regards the CIP Preference Shares as 'equity'.
I am a little nervous of companies with large amounts of debt in general on their balance sheet in a rising interest rate environment. Management's preferred formula for ultimately exiting CIP funding is 'option 3', to refinance it all as 'additional debt'. Furthermore the banks seem prepared to play along with this strategy, at least for now (see my post 2836). Whether 'global financiers' feel the same, we shareholders get to find out before October 2023. That date is when a 500 million euro note funding arrangement, with a coupon rate of 1.13%, comes to an end. I wonder what coupon rate that 500m euro quantum of debt will be able to be refinanced at?
SNOOPY
Some fair comments, but I want to add to this part:
...that I would be more than nervous of a NZ-domiciled company with 100% on-shore earnings, borrowing in a currency other than NZD. AFAIK there is no "natural hedge" available to CNU when borrowing in Euros. That makes zero sense to me and exposes CNU (and shareholders) to FX risk where none should exist.
It is important to respond to my quote immediately above, because even an established utility provider can come unstuck if the debt position gets out of hand. To check this out, I will use the same yardstick that the banks use. That consists of figuring out how well the total bank debt taken on by the business covers (or smothers?) its gross earning capacity (EBITDA). The lower the (Net Debt)/EBITDA ratio the better.
I am using the calculation method shown on Slide 18 from PR2022 (August 2022)
FY2022 Chorus View (Pre IFRS16) Chorus View (Post IFRS16) Reference Snoopy View Bank and Note Debt (1) $2,389m $2,389m (Slide 18 PR2022) $2,389m add PV of CIP Debt Securities (Senior) $225m $225m (Slide 18, PR2022) $486m (2) (my post 2838) add Net Leases Payable $187m (Note 5 AR2022) equals Total Selected Debt $2,614m $2,801m $2,875m less cash $88m $88m $88m equals Total Net Selected Debt $2,526m $2,713m $2,787m divided by EBITDA $660m $675m $660m equals 'Net Selected Debt' / EBITDA 3.83 4.02 4.22
Notes
1/ Total net bank and other senior debt = $190m+($828m+$464m-$42m)+($200m+$500m+$154m+$45)m = $2,339m.
The $42m decrease in balance sheet value of the Euro bonds is because of a change in end of year value of an associated hedging instrument, the cross currency interest rate swaps used to hedge the Euro bonds (AR2022 p55). Likewise the $45m adjustment upwards in the value of the NZ bonds is because of the offsetting $45m decrease in value of interest rate swaps that are used to ensure hedge effectiveness (AR2022 p55).
2/ I don't like the accounting standard that lets a company discount debt that ultimately has to be repaid. So I am assuming a PV discount rate of 0%.
My third column 'Snoopy' view of this calculation removes the 'lease interest expense' (a part of rent, now reclassified as an interest payment under IFRS16) from EBITDA (so EBITDA is lowered). But I have also removed the 'Net Leases Payable' (a construct of IFRS16, representing rent contracts committed to over several periods) as a 'finance liability', as it wasn't a finance debt in a pre-IFRS16 sense. I don't consider my approach better than the approach Chorus has used. But my approach aligns better with the old way of assessing this banking covenant.
The other controversial thing that I have done is to disallow any discounting of CIP debt.
The covenant number that Chorus are targeting is anything under 4.75.. Anything over 5.25 ( Moody's ) and 5.00 ( S&P ) - see my post 2836 - are the trigger ratings for a credit downgrade. To avoid such a thing, Chorus would have to take some remedial action, like raise some more capital. My Snoopy retro calculation shows that - even if the company replaced their CIP debt with an equivalent quantum of commercial debt-, they would still be noticeably below this specific debt covenant ceiling. That's good. But it is all on the precept that net revenue remains at current levels. Is this likely?
SNOOPY
I suspect the lure of EURO funding may be 'supply side initiated'. Getting 1.13% on your invested capital may not sound like a good deal to you in NZ. But if you are a Belgian dentist, who is staring down the barrel of a 'negative 1%' coupon rate on an investment bond, that dentist might think she is onto a good deal!
Of course, FX exposure at Chorus is managed. From AR2022 p41:
"Chorus holds cross currency interest rate swaps to hedge the foreign currency exposure to the European Medium Term Notes (EMTN). The cross currency interest rate swaps entitle Chorus to receive EUR principal and EUR fixed coupon payments for NZD principal and NZD floating interest payments (*)."
(*) This sentence doesn't read right to me. Chorus are receiving EUR principal to debt fund their business (I get that bit). But I would I assume they are making EUR fixed coupon (i.e. interest?) payments because of this, not receiving EUR fixed coupon payments. Either that quoted sentence is poorly worded, or I am just dumb, not really understanding how these interest rate swaps work (a possibility I do not rule out).
The slightly strange wording then continues.
"The EUR cross currency interest rate swaps (notional amount EUR 800 million) are partially hedged for NZD interest rate payments using interest rate swaps."
EUR 800 million is the actual amount of the euro debt. This quote is saying that an equivalent sized interest rate swap is - I think- brought into being to allow this money to be accessed as NZD, without capital exchange rate risk. But I am guessing that this cross currency interest rate swap, while cancelling itself out against the underlying loan at the end of the loan contract, must be valued in the intermediate years as a non-net zero (accounting rule priced) offset. Thus for annual account valuation purposes, it has an accounting value adjusted from EUR 800 million, even though the ultimate 'destination value' of this interest rate swap, once the EURO notes are due for repayment, remains at EUR 800m.
This is my best effort at explaining how this EURO funding at Chorus works. I am happy to be corrected if I have got it wrong.
SNOOPY
The actual FY2022 revenue figures have been published as I write this. So it is time to update my revenue forecasting table.
From 1st January 2022, Chorus has been operating under the 'regulated utility model', that regulates how much Chorus will be allowed to charge their downstream wholesale fibre customers over CY2022 through to CY2024, otherwise known as the 'Price Quality (PQ) Period - First.' (PQP1).
The aggregate Maximum Allowable Revenue (MAR) for fibre for the 3 years is $2.2 billion, split up into $690.2m (CY2022), $747.4m (CY2023), and $789.5m (CY2024). Forecasts by Chorus themselves for CY2022 were for fibre revenues of $657m (AR2022 p10). Where actual revenues fall below the MAR cap, the difference is added to the MAR for the next period.
There is a six month gap between the end of the reporting year and the end of the calendar year. This means we can make an educated guess at the MAR revenue in the financial year by splitting the difference between forecasts for adjacent calendar years.
MAR Revenue for FY2022: ($545m+$690.2)/2= $617.6m
MAR Revenue for FY2023: ($690.2+$747.4)/2= $718.7m
MAR Revenue for FY2024: ($747.4+$789.5)/2= $768.5m
My table of historic and forecast 'total change of revenue' over the years of interest is now as follows.
FY2016 FY2017 FY2018 FY2019 FY2020 FY2021 FY2022 FY2023F FY2024F Fibre Revenue $133m $202m $276m $368m $466m $545m $614m $719m $769m Fibre Revenue Increment Percentage +51.9% +36.6% +33.3% +26.6% +17.0% +12.7% +17.1% +7.0% Non-Fibre Revenue $875m $838m $714m $602m $493m $402m $351m $291m $241m Non-Fibre Revenue Decrement -$37m -$124m -$112m -$109m -$91m -$51m -$60m -$50m Non-Fibre Revenue Decrement Percentage -4.3% -14.8% -15.7% -18.1% -18.5% -12.7% -17.1% -17.2% Total Revenue $1,008m $1,040m $990m $970m $959m $947m $965m $1,010m $1,010m
Notes
1/ 'Non-fibre revenue' up until and including FY2022 has been calculated by subtracting 'Fibre revenue' from 'Total revenue'.
2/ Chorus have not provided any forecast as to where they expect their non-fibre revenue to go over the next few years. I have inspected the five year revenue trends for:
a/ 'copper connected revenue' and
b/ 'field services, value add network services and infrastructure'
as two groups. The latter group I am forecasting constant revenue of $120m over FY2023 and FY2024. (Actual figures were $118m (FY2020), $118m (FY2021) and $128m over FY2022). The 'copper connected revenue', comprising 'copper based broadband', 'copper based voice' and 'data services copper' have taken an $91m decline over FY2021 and a further $69m over FY2022 So I am forecasting the decline of non-fibre revenue to continue: down $60m over FY2023 and $50m over FY2024. The 'diminishing decline rate' I am modelling to take account of a slowing trend, as 'easy conversions' to fibre have happened already.
'Total revenue' for the forecast years is calculated by adding the forecast fibre revenue to my own modelling of the non-fibre revenue as described above.
--------------------
So what to make of this? IF you compare the actual fibre revenue over FY2022 of $614m with
a/ the forecast regulated fibre revenue for FY2022 (slide 24 HYR2021, 22-02-2021) $680m (interpreted from graph), AND
b/ A more recent linearly interpolated FY2022 fibre revenue of $618m (being the arithmetic average of the previous year's fibre revenue used figure of $545m and the prescribed first year of MAR $690.2m).
THEN that actual revenue for FY2022 represents a $62m revenue drop, compared to the forecast just a year earlier. This seems very similar to FY2021 when a big jump in revenue was projected, However, when that big jump did not occur, the big jump forecast was pushed out into he subsequent year. Nevertheless even if revenue in FY2023 disappoints (as happened when forecasting 'next years' revenue over the previous two years), I don't see an FY2023 scenario where the revenue does not increase, at least a little. So it looks like from a revenue perspective, Chorus is on track to raise their dividend for FY2023. Yet ultimately we shareholders want to know about profit trends, as well as revenue trends. What is happening to ARPU (Average Revenue per User)?
SNOOPY
The following revenue figures for 'fibre broadband' and 'copper broadband' are largely taken from the 'Management Commentary' section of the respective annual reports. The 'In summary' table lists the number of connections, whereas the 'Revenue Commentary' lists the revenue attributable to each type of broadband.
Year Broadband (GPON) Revenue Broadband (GPON) Connections ARPU (F) Copper Broadband Revenue Copper Broadband Connections ARPU (C) FY2015 $29m 75,000 $? $268m 1,132,000 $? FY2016 $60m 167,000 $37.19 $456m 1,059,000 $34.69 FY2017 $123m 292,000 $44.66 $501m 894,000 $42.75 FY2018 $198m 433,000 $45.52 $421m 754,000 $42.58 FY2019 $294m 599,000 $47.48 $344m 597,000 $42.44 FY2020 $393m 740,000 $48.92 $271m 466,000 $42.49 FY2021 $477m 860,000 $49.69 $203m 320,000 $43.04 FY2022 $548m 949,000 $50.49 $153m 240,000 $45.54
Notes
1/ Sample ARPU calculation for Fibre Broadband 'Gigabit Passive Optical Network' (GPON) for FY2022:
ARPU = 1/12 x ($548m/ (0.5x(0.949m + 0.860m)) = $50.49
2/ Fibre GPON customers for FY2016 = Fibre 'Mass market' Customers for FY2016
Fibre GPON Revenue for FY2016 and FY2015 taken from Appendix 3, Slide 36 , in PR2017.
3/ Fibre GPON customers for FY2015 estimated as: 88,000 - 13,000 = 75,000
4/ Copper broadband numbers for FY2016 and FY2015 are from Slide 5 of the Chorus PR2016.
5/ Copper broadband revenues for FY2016 and FY2015 are from Appendix 3, Slide 36, in PR2017
6/ Actual ARPU for FY2021 listed on slide 15 of PR2021 to be $49.87, a little higher than the $49.69 that I have calculated above. However Chorus has used a different calculation method
"ARPU is total GPON revenue for the June month, divided by the average of May and June connection."
I have used an average connection for the full year, not just May and June.
Discussion
Why is this table important? The first ARPU column is representative of the business being gained as fibre broadband is being rolled out across the country. The second ARPU column represents business being lost as copper broadband is gradually retired and replaced with fibre broadband. It is heartening to see the ARPU being gained is greater than the ARPU being lost, even if such a superficial statement does not tell the full story.
I talk of copper broadband as a 'legacy technology'. Yet it is surprising how recently the likes of copper VDSL was the latest in broadband technology and still in its growth phase (revenues sharply rose between FY2015 and FY2017). The calculated ARPU is consistent with the wholesale /Dial Up (remember that?)/ ADSL/VDSL/ bill being fairly steady over the time-frame of this table up to FY2022 at least. By contrast the ARPU for Fibre GPON is steadily rising. This is consistent with, on average, with allowable inflation adjustments and a higher number of higher specification Fibre Broadband plans being sold as the years go by.
SNOOPY
Snoopy
You reckon this is a hold or even a buy now?
I had the impression that you were negative earlier on.
It is a good question Nor. Looking back over the last year, I think it is true that I was looking to justify that Chorus was trading at a price where I should take my profits. You obviously picked up that vibe. But with another year of operating results out now, I am not so sure. The key post from the many I have made over the last few days is the one below
I don't want anyone to get too excited, because this isn't a get rich quick opportunity. However, I did roll over a one year bank term deposit at 4.5% the other day, a rate I thought 'quite good'. And with a prospective comparative yield of my Chorus shares sitting at 5.1%, possibly rising to 5.7% within two years I am thinking 'that looks a little better'.
Then I look across to my other utility type investments in Contact Energy and Mercury, which I believe have better growth prospects. But if Onslow gets the go ahead, then this could really take the shine off some of those electricity market trading opportunities that have boosted the gentailer profits in recent years. So all of a sudden my smallest NZX investment (my shares in Chorus) is looking relatively more attractive. The credit rating industry seems more comfortable with those high Chorus debt levels too, now that it is transforming from a 'growth' company into a 'utility' company, and they take heart from the CIP government funding backing too.
So to answer your question, I think I have moved from a slightly worried/negative headspace into a more neutral position. It wouldn't surprise me, once I finish my 2022 deep dive into Chorus, that I come out with the finding that 'Mr Market' has the share price about right. Some would say that means I have wasted a lot of time finding out what Mr Market could have told me by just looking at the CNU trading price. But my mission is to try and find out what is driving this value. I see that as the edge I need if I am to sleep well holding Chorus shares into the future. In the short term, I think that rising interest rates are the biggest threat to a rising CNU share price.
SNOOPY
Addressing my comment immediately above, 'the full story' must address the numbers of copper customers being lost, offset against the number of fibre customers being gained.
Wholesale Broadband Operator %ge Nationwide Connections EOFY2022 Connections {A} EOFY 2022 Premises Passed {B} %ge Regional Take Up {A}/{B} Chorus 73% 919,000 1,324,000 69% Northpower fibre 2% 22,624 33,000 69% Tuatahi fast fibre 14% 173,000 240,000 72% Enable 11% 143,331 193,000 74% Total 100% 1,258,000 1,790,000 70.3%
Notes
1/ The end of the financial year is, for all four operators, on 30th June.
2/ Ultrafastfibre was in May 2020 sold to Sentier Investments based in Australia for one of their managed global infrastructure funds . Ultrafastfibre now trades under the name 'Tuatahi fast fibre' in New Zealand. The UFB2 and UFB2+ roll outs at Tuatahi were completed early in December 2019.
The most important figure in the above table, for Chorus investors, is the one in bold. Chorus owns 100% of the legacy copper network. So as users progress to fibre, they will eventually lose almost all of their copper legacy network customers, but only retain a maximum of around 73% of the nationwide replacement fibre network customers. This means that although transitioning from the copper network will overall mean a boost in customer receipts for pure fibre network owners, the expected 'gain' in revenue at Chorus is a little different.
Weighted Average Fibre Monthly Revenue Gain Per User 73% x $50.49= $36.86 less Weighted Average Monthly Copper Revenue Lost Per User 100% x $45.54= $45.54 equals Weighted Average Overall change in MRPU (Monthly Revenue Per User) -$8.68
Yes you did read that right. As the fibre transformation wave wafted over the country in FY2022, Chorus has lost on average $8.68 MRPU for every customer that transitioned to fibre! In Australia they are forecasting around 70% of old copper network premises will eventually transition to fibre. In New Zealand we will now track above that 70% 'migration to fibre' figure, (because of our superior fibre broadband infrastructure compared to Oz, we are at 70% now). But as a counterpoint to that, we now have three 'fixed wireless' broadband retailers (Spark, Vodaphone and 2 degrees) marketing their internally owned alternatives to fibre broadband at good discounts to fibre (albeit at reduced peak time speed). At least Chorus will get a meaningful inflation adjustment from their staunch fibre customers that will strengthen their cashflows this year.
I estimate the total revenue lost to Chorus over the FY2022 financial year, as a result of copper to fibre transition, to be:
0.27(1,258,000 - 1.151,000) x ($8.91/month x 12months) = $11.4m
If you regard all of that 'decremental revenue loss' as profit forgone, then this migration has reduced Chorus profit for the year by:
0.28 x -$11.4m = -$3.2m
Actual normalised profit drop between FY2021 and FY2022 was: $48.3m - $52.3m = -$4m
I could make an argument from this that most of the 'normalised profit drop' at Chorus over FY2022 has occurred as a result of the success of NZ's transition to fibre over that time!
SNOOPY
I have been struggling to complete the above table in the previous post. So to avoid future angst, I thought it worth putting 'keyboard to the ether' to record for future reference how I solved the problem.
For the record, the above table was difficult to compile this year. The main problem was that Tuatahi, despite the native nomenclature, is now part of an unlisted company in 100% foreign hands. That means I couldn't find any reporting data as to the growth in network utilisation over FY2022. In the end it was 'Crown Infrastructure Partners' that came to my rescue. ('Crown Infrastructure Partners' is the government organisation that funded the whole UFB roll out ahead of consumer network demand.)
https://www.crowninfrastructure.govt.../publications/
Even then, the rather tardy release of the CIP FY2022 (ending 31st June 2022) annual report (4.5 months on and it isn't available) left me scratching around the site for information that I could use.
Method 1
The most recent 'quarterly connectivity update' in June 2022 told me that overall UFB uptake for the country was 70%. Given that
i/ The total incremental uptake was the sum of all the individual incremental uptakes AND
ii/ Given the total number of premises passed by fibre by Tuatahi had not significantly changed by EOCY2018
I could use the variables I did know to calculate the variable I didn't know. 'f' in the equation below is the fraction of premesis passed by fibre taken up by Tuatahi customers as at the 30th June 2022 reporting date.
0.69(1.324m) + 0.69(0.033m) + f(0.240m) + 0.74(0.193m) = 0.7(1.790m) => f=0.724 or 72.4%
Once I had calculated the utilisation rate at Tuatahi, and knowing the size of the network, that means I could calculate the number of Tuatahi fibre customers by EOFY2022. Furthermore, because I knew the number of customers from EOFY2021, the difference was the number of customers gained over FY2022.
Method 2
If I looked at each of the four CIP connectivity quarterly update reports relating to the full year ending 30th June 2022, the new UFB connections over the whole country add up as follows
30,056+28,797+25,061+24,518 = 108,074
If we look at the increase in connections over the same time period from the other three network owners we know about we get
Chorus 919,000 - 837,000 = 82,000 Northpowerfibre - = 1,624 Enable 143,331 - 132,000 = 11,331 Total 94,955
I am calling that total 95,000. If we subtract that number from the total number of new connections across all networks :
108,000 - 95,000 = 13,000
That 13,000 is the number of new fibre connections from from the one fibre market player we don't know about - Tuatahi. Because we know the number of customers at Tuatahi from the previous year, we can now work out the Tuatahi network utilisation:
(161,000 + 13,000) / 240,000 = 72.5%
Conclusion
Note that the result from method 2 is almost exactly the same as the result from method 1. When you come at the same problem from two different angles and come to the same result, this gives me confidence the result is right! It also makes sense that the market penetration at Tuatahi is closer to that of Enable than Chorus, because Chorus covers many more smaller rural towns that were further behind in their fibre network 'build to completion' program. To be conservative, I have rounded down the Tuatahi market penetration to 72% (instead of rounding it up to 73%).
SNOOPY
https://www.nzx.com/announcements/406952
Key results
• Fibre connections increased by 38,000 to 997,000
• Fibre uptake 71% in completed UFB areas
• 24% of mass-market fibre connections on gigabit or higher plans
• Operating revenue $487m (HY22: $483m)• EBITDA $342m (HY22: $347m)
• FY23 EBITDA guidance increased to $675m to $690m
• Net profit after tax $9m (HY22: $42m)
• Unimputed interim dividend of 17 cents per share
Nice concise explanation in the half yearly 2023 newsletter on page 2:
"Our dividends are expected to be unimputed for some time, because the substantial investment we have made in the fibre roll out has created a difference in timing between tax and accounting depreciation. This means tax payments are effectively deferred to the future and we do not have the imputation credits in the short term."
--------------------------
And now for the long answer.....
While I am not disagreeing with the explanation above, there are other ways to spin this story. Page 1 of the newsletter gives us a good clue on what is happening:
"Net profit decreased .....
Part 1/ largely due to interest rate rises and the need to refinance a large tranche of debt due for repayment later in 2023(*). AND
Part 2/ The accelerated depreciation of the copper cables in areas where fibre is available also contributed to the reduction in net earnings."
(*) AR2022 shows Euro Medium Term Notes on the books at EOFY2022 valued at $NZ828m needing to be refinanced, making up 35.7% of Chorus long term debt as at EOFY2022. This debt was taken out over FY2017 and I quote from p31 of that Annual Report:
"On 18th October 2016, Chorus issued EUR 500m of Euro Medium Term Notes at a rate of 1.125%. They will mature in October 2023 and have been swapped back to $NZ785m using cross currency interest rate swaps."
Now, 1.125% of EUR 500m is EURO 6.25m.
The added capital cost of the loan to be repaid six years later is $828m-$785m= $43m, or $7.2m each year for six years. At that time $NZ1 = Euro 0.59. So the effective extra interest paid each year as a result of the interest rate swap was: $NZ7.2m x 0.59 = Euro 4.2m. This means the effective euro interest rate being paid on this debt is:
(6.25m+4.2m)/6.25m x 1.125% = 1.881%,
Refinancing of this bond happened at 3.625%
https://chapmantripp.com/about-us/ne...ful-refinance/
This means the 'interest payments due', to all of those friendly European Note Holders that are 'helping Chorus out' on what is now a 7 year bond has effectively doubled to: (3.625/1.881) x EUR10.45m/0.59 = $NZ34m per year.
Part 2 of the NPAT after tax is the success of broadband roll out undermining the success, and therefore value of the legacy copper network. The IRD have their rules own depreciation that presumably do not include slashing the value of the legacy copper network on Chorus's books at the request of management.
So in my view the explanation for the disappearance of imputation credits could be rewritten as follows:
"The real life of the copper network is dramatically shorter than we thought. But we have been continuing to pay low depreciation charges that do not reflect the reducing revenue that we still get from this network. So actually we are now making a substantial loss on copper, which combined with the significantly higher interest payments on our 'long terms loan portfolio' have wiped out the overall unadjusted Chorus profit -from the point of view of the IRD- to nothing (hence no imputation credits). IOW 'we no longer make any money at all', but -ah- we have free cashflow!"
I think that is an equally good explanation as to what has happened. It doesn't read as well as the paragraph written by Chorus that I opened this post with though!
SNOOPY
https://www.nzx.com/announcements/416645
Highlights
• Fibre uptake: 73% in UFB areas, with UFB fibre rollout now complete
• Fibre growth: added 72,000 fibre connections in FY23, totalling 1,031,000
• Fibre plans: 91% of residential and business connections above 300Mbps
• Revenue: grew to $980m from $965m in FY22
• EBIT: down to $226m from $248m in FY22
• Net profit: down to $25m from $64m in FY22
• Dividend: 42.5 cents per share, unimputed for FY23
Not sure about the long term future for this - Spark was trumpeting they managed to move 30% of their broadband customer base to their fixed wireless product (which users their own cellular network rather than fibre), with plans to convert much more over (cutting out the main cost of providing broadband - Chorus).
Well, I think about it this way. I have Fibre to my house, (I use my own router for better wifi performance in my own home) and I have had it, Fibre, for a number of years now. It's about 99% reliable, and any issues I do get seem to have more to do with my ISP than with Chorus' fiber itself. It's plenty fast enough for me too, even when I watch Netflix or YouTube or other content in 4K, I've never had any lag ever. The price seems reasonable too. And if the price I pay to my ISP is within $5 or $10 a month of everyone else, well, who cares?
So, why would I give up fixed fibre to have a wireless connection instead? What would be the problem I was trying to solve?
Even if I got pissed at my ISP and decided to go someplace else, my fixed fibre infrastructure is already in place, so why would I go wireless?
And if my experience is like most others, and my hunch is that it is, I don't think Chorus has much to worry about from wireless connections. There is room and a place in the market for both.
https://www.nzx.com/announcements/426882
Key results
• Increase in fibre connections: by 31,000 to a total of 1,062,000
• Fibre uptake increased to 70.6%• 25% of residential fibre connections on gigabit or higher plans
• Operating revenue $503m (HY23: $487m)
• EBITDA $347m (HY23: $342m)• Net profit after tax $5m (HY23: $9m)
• Unimputed interim dividend of 19 cents per share
• JB Rousselot steps down as CEO in April 2024. Mark Aue appointed CEO
"Better customer experience key factor"
"New Zealand IoT Alliance says a key part of a connected New Zealand is the acceleration and uptake of IoT, with non-premise IoT representing a significant portion of this. Non-premise IoT connections are external infrastructure assets that can be transformed into IoT applications, for example, traffic lights, bus stops, outdoor billboards and CCTVs. “We estimate that better use of IoT could create at least $2.2 billion in net economic benefit for New Zealand over the next 10 years. We are starting to see its impact across smart city infrastructure, electric vehicle charging stations, utility, security monitoring, and digital billboards,” he says."
https://businessdesk.co.nz/sponsored...5451-446239310