IFT - Infratil

[Snoopy]

Snoopy wrote:
"We guess that the CDC assets in operation today earn 10% on the original capital outlay. This means the price of assets on the books is $97.4m/0.1= $974m"

Hi Snoopy, in the roadshow presentation from memory it was mentioned that CDC presently return between 10 and 15% on it's different assets, so 10% a bit low I think. Maybe halfway 12.5 % is a better estimate as a average.

Let's go with your suggestion forest, and see where that leads us: "This means the price of assets on the books is $97.4m/0.125= $780m"

Taking that $780m and depreciating it linearly over 40 years gives an annual depreciation and amortisation charge of: $780m/40= $19.5m

SNOOPY

[Snoopy]

I push out my average depreciation time from 20 years to 40 years. The buildings may be on the books with a longer life than that, perhaps 50 years. But all of the cooling equipment will depreciate much faster than that. So I will stick with 40 years as a realistic compromise combined total depreciation schedule number. I expect amortisation will be zero over the period, as no goodwill would have been booked in the first place.

Recalculating Net CDC Profit after once again revising depreciation and now interest charges

iic/ Let's say that USPP funding was secured at a rate of 3% (a pure guess on my part). That means the annual interest bill for CDC would be: 0.03 x $A1,168m= $35m
And the Infratil share of that would be: 0.4824 x $35m = $16.9m

We guess that the CDC assets in operation today earn 12.5% on the original capital outlay (refer post 3661). This means the price of assets on the books is $97.4m/0.125= $780m.

Now, 1/40th of that figure is $19.5m. Assuming 'F' (which is generally reserved for one off non-operational transactions), is zero, we can calculate the implied IFT share of NPAT from the EBITDAF figure quoted as follows:

0.7 x ($130.7m(1) - $16.9m - $19.5m) = $66.0m

(1) This calculation assumes that $130.7m is the Infratil share of the whole CDC EBITDAF profit. This is not clarified in the referred NZX presentation. https://infratil.com/news/infratil-a...-equity-raise/

Despite lengthening the depreciation profile, and reducing the interest cost, $66.0m is still well behind the $97.4m NPAT, the share of CDC's profit owned by IFT actually achieved. So there are possibly still some undisclosed one off adjustments going on within the CDC accounts that outsiders are unaware of.

SNOOPY

[maclir]

Pretty sure the ATO has made some decisions on data centre depreciation, going through the different elements that can be found, floors, cooling, buidlings etc. Not sure about IRD.

[Snoopy]

I push out my average depreciation time from 20 years to 40 years. The buildings may be on the books with a longer life than that, perhaps 50 years. But all of the cooling equipment will depreciate much faster than that. So I will stick with 40 years as a realistic compromise combined total depreciation schedule number. I expect amortisation will be zero over the period, as no goodwill would have been booked in the first place.

Recalculating Net CDC Profit after once again revising depreciation and now interest charges

iic/ Let's say that USPP funding was secured at a rate of 3% (a pure guess on my part). That means the annual interest bill for CDC would be: 0.03 x $A1,168m= $35m
And the Infratil share of that would be: 0.4824 x $35m = $16.9m

We guess that the CDC assets in operation today earn 12.5% on the original capital outlay (refer post 3661). This means the price of assets on the books is $97.4m/0.125= $780m.

Now, 1/40th of that figure is $19.5m. Assuming 'F' (which is generally reserved for one off non-operational transactions), is zero, we can calculate the implied IFT share of NPAT from the EBITDAF figure quoted as follows:

0.7 x ($130.7m(1) - $16.9m - $19.5m) = $66.0m

(1) This calculation assumes that $130.7m is the Infratil share of the whole CDC EBITDAF profit. This is not clarified in the referred NZX presentation. https://infratil.com/news/infratil-a...-equity-raise/

This is getting terribly frustrating. What started out with what I thought was a simple question "How much does a data centre earn?", has turned into a constantly rewriting presentation nightmare.

Recalculating Net CDC Profit after revising earnings

Cross referencing from AR2024 page 80, to the Infratil presentation slide 27 shows 'revenue for CDC' over FY2024 to be $NZ412.3m (AR2024) c.f. $A356m (Infratil Presentation) for the same reporting period. This implies an NZD to AUD exchange rate of $356m/$412.3m= 0.8643. Since the NZD/AUD exchange rate never approached that figure over FY2024, it looks like I am 'missing something'. Some foreign exchange futures contracts going the wrong way and upsetting the exchange rate picture perhaps? In any event, this suggests that 'presentation slide 27' is showing numbers that apply to the whole of CDC, not just the IFT 48.24% share. This means my 'earnings calculation' has to be adjusted again, to this:

iid/ 0.7 x (0.4824x$130.7m - $16.9m - $19.5m) = $18.7m

So I have gone backwards, getting even further away from the implied $94.7m Infratil share of CDC profit, implied by AR2024. Sigh!

The next point of bafflement is to explain why EBITDAF at CDC increased from $A215m to $A271m between FY2023 and FY2024 (slide 27), while the net surplus after tax declined from $NZ762.7m to $NZ201.9m over the same period (AR2024 p80).

SNOOPY

[Snoopy]

Pretty sure the ATO has made some decisions on data centre depreciation, going through the different elements that can be found, floors, cooling, buildings etc. Not sure about IRD.

Yes I am sure you are right about the the ATO I don't see any reason why NZ's IRD would not be 'on the ball' in the same way. The ironic thing is the 'removal of structural building depreciation' as a tax deductible item in NZ could mean that despite Australia being nominally a 'higher taxed' 'higher cost' tax residency, it might be cheaper for CDC to own their NZ properties on their Australian tax book now. Perhaps an unforeseen consequence of changing the NZ building tax depreciation regime?

SNOOPY

[Snoopy]

The next point of bafflement is to explain why EBITDAF at CDC increased from $A215m to $A271m between FY2023 and FY2024 (slide 27), while the net surplus after tax declined from $NZ762.7m to $NZ201.9m over the same period (AR2024 p80).

I reckon I can make a fist of unravelling this one. Since NPAT must always be less that EBITDAF, except for the 'one offs', the FY2023 NPAT must contain a lot of 'one offs', (like a revaluation upwards of the whole portfolio)? I would guess that new expenditure on new data centres would be capitalised. So I don't think that 'the intent to build more data centres at scale' and 'executing such a plan' should mean that such increased expenditure would necessarily force declared profits down.

SNOOPY

[Bjauck]

Yes I am sure you are right about the the ATO I don't see any reason why NZ's IRD would not be 'on the ball' in the same way. The ironic thing is the 'removal of structural building depreciation' as a tax deductible item in NZ could mean that despite Australia being nominally a 'higher taxed' 'higher cost' tax residency, it might be cheaper for CDC to own their NZ properties on their Australian tax book now. Perhaps an unforeseen consequence of changing the NZ building tax depreciation regime?

SNOOPY

That is an interesting point. By avoiding the elephant in the room a CGT, as well as depreciation, NZ has some other taxes - financial arrangements rules, FIF etc. - to bring to income what many other jurisdiction do not.

[Snoopy]

CDC had approximately $A3bn in outstanding debt arrangements as at 1st December 2022.
https://www.thelawyermag.com/au/prac...ts-help/429195

Part of the refinancing package at the time included $A308m of US private placement (USPP) issuance. The above reference does not give any hint as to what the agreed interest rate might have been. Without this information I cannot calculate the interest rate bill at CDC. So any hints on where I can find such information, even a ballpark figure, would be most welcome.

Next we move onto the CDC funding strategy today (slide 14 of the referenced Infratil presentation).
"CDC has committed or undrawn liquidity of ~A$2.2bn following the recently executed ~A$860m USPP issuance (June 2024), weighted average tenor now ~5.7 year"

So what this is telling us is that, following a reshuffle to commit more debt to USPP the 'committed and undrawn facilities' are still about $A3bn 18 months down the track. there must be some reason to commit to USPP debt over Australian bank debt. So I am guessing that the total amount of USPP debt disclosed is equal to the complete amount of debt drawn: $A308m + $A860m = $A1,168m

Let's say that USPP funding was secured at a rate of 3% (a pure guess on my part). That means the annual interest bill for CDC would be: 0.03 x $A1,168m= $35m

And the Infratil share of that would be: 0.4824 x $35m = $16.9m

That is very different to the $31.9m (wrong) figure that I was using before. Lots of assumptions used. I am not really sure how good they are, so best I sleep on it before continuing.

Trying to work out a gearing ratio when you aren't given a balance sheet is an interesting exercise. In the case of CDC we are given partial information in Infratil AR2024 (31st Match 2024 balance date) p80. More clues are in the June 2024 Infratil capital raising presentation.
https://infratil.com/news/infratil-a...-equity-raise/

The report timing differences between the two references means we can't be sure that we can take data from one report and be sure it is current for the other. But AR2024 shows for CDC:

Total Assets: $6,820.7m
Total Liabilities: $4,251.1m
Net Assets: $2,575.6m

This gives a CDC debt ratio of: $4,251.1m/$6,820.7m = 62.3%

Go to the presentation, slide 14, and a 'CDC key metric' is 7-10 times leverage (the ratio of a company's loan capital (debt) to the value of its ordinary shares (equity) ). This implies a debt ratio of between 7/8= 87.5% and 10/11= 90.9%.

Stacking up the 'key metric' debt ratio range against the actual debt ratio, it looks like 'the plan' for CDC is to borrow a lot more money. And that was before the equity injection (of which Infratil's share is expected to be $600m - ref slide 14) from the capital raise arrives, which should allow CDC to borrow even more!

Given all of this, my guesstimate of the total borrowed capital by CDC of $1,168m (post 3662, referred above) at EOFY2024 looks very light. I find it confusing when a company talks about 'borrowing facilities', because such a number just means that such facilities are available. It doesn't mean that they have been drawn down to the maximum amount. But some presentations do not make it clear which number (debt which can be drawn vs debt that is drawn) is being talked about. Nevertheless balance sheets, or the scant information we have from them (AR2024 p80) do not lie. And even if the total declared liability balance, $4,251.1m, is all long term bank or USPP debt (there isn't enough detail disclosed to know this), in the big picture, CDC regards that as 'conservative' (!).

In most companies with a high debt ratio of 62%, which management were seeking to increase further to near 90%, this would be a big red flag for me. However, Infratil is a very skilled manager of long term infrastructure assets. The 30 year data centre contracts that Infratil, via CDC, is able to write gives a secure cashflow from what is a largely recession resistant asset class. And if CDC have a secure cashflow over a long time period, then this allows the company debt profile to be leveraged up with greater safety.

SNOOPY

[Snoopy]

My average depreciation time remains at 40 years. The buildings may be on the books with a longer life than that, perhaps 50 years. But all of the cooling equipment will depreciate much faster than 40 years. So 40 years is my compromise 'combined total depreciation schedule number'. Amortisation is assumed to be be zero over the period, as no goodwill would have been booked on the balance sheet for a 'green field build' data centre.

Recalculating Net CDC Profit after once again revising interest charges

iie/ Total debt, assumed as a constant value across the year was $4,251.1m. Let's say that USPP funding and other bank funding was secured at a rate of 3% (a pure guess on my part). That means the annual interest bill for CDC would be: 0.03 x $A4,251.1m= $127.5m

And the Infratil share of that would be: 0.4824 x $127.5m = $61.5m

We guess that the CDC assets in operation today earn 12.5% on the original capital outlay (refer post 3661). This means the price of assets on the books is $97.4m/0.125= $780m.

Now, 1/40th of that figure is $19.5m. Assuming 'F' (which is generally reserved to symbolise one off non-operational transactions), needs no further adjustment, we can calculate the implied IFT share of NPAT from the EBITDAF figure quoted as follows:

0.7 x (0.4824x$130.7m - $61.5m - $19.5m) = -$12.6m

I have lengthened the depreciation profile, and increased the interest cost to something more realistic, now that I have realsed the debt burden at CDC is much greater than I had previously thought, The result is that the net profit I thought I had, has turned into an operational loss (-$12.6m). Yet the declared profit was +$97.4m NPAT, the share of CDC's profit owned by IFT actually achieved. Despite this being my fifth attempt at calculating it, I am not guaranteeing that I have got my operational profit calculation right. But even with some more imagined future tweaks, I see no way to turn a $12.6 loss into a $97.4m profit.

The loss is determined by the gearing of CDC which will have been decided upon by Infratil and its ownership partners. My inkling is that CDC has been designed as a company with near zero earnings for tax reasons. It could easily become profitable by having less debt in its underlying structure. But to avoid paying back some of Infratil's injected capital as a 'tax bill', Infratil have kept the interest bill high. This is what accountants call 'tax efficiency'. So it looks like CDC is losing money, or making a very modest profit on 'normal operations'. I certainly would not have predicted that, prior to me tackling this calculation, starting at the weekend just gone.

SNOOPY

[kiora]

There being a lot/too many asumptions there I'd say

One of my collegues always said assumptions are the MOAF