SPK - Spark NZ

[Shareguy]

Craig's view

This was a weak result. EBITDAI was $1163m vs guidance issued in May of $1170-1210m. We had forecast EBITDAI of $1180m (core of $1145m plus $35m of asset sales). This result comprised core EBITDAI of $1061m plus asset sale gains of $102m, so the core result was down 8.5% on pcp and well below CIPe $1145m.


SPK will pay a fully imputed final dividend of 14.0c, in line with guidance and taking the total for the year to 27.5c. FCF was $330m and well below the c$500m needed to cover the dividends. The DRP has been reinstated, which based on historical take-up could fund around $100m of the gap. It is also hard to not be critical of the company's capital management when less than a year ago they were buying back stock at over $5 per share and now are planning to issue new shares in the DRP at a price more than 20% lower.

[LoungeLizard]

Craig's view

This was a weak result. EBITDAI was $1163m vs guidance issued in May of $1170-1210m. We had forecast EBITDAI of $1180m (core of $1145m plus $35m of asset sales). This result comprised core EBITDAI of $1061m plus asset sale gains of $102m, so the core result was down 8.5% on pcp and well below CIPe $1145m.


SPK will pay a fully imputed final dividend of 14.0c, in line with guidance and taking the total for the year to 27.5c. FCF was $330m and well below the c$500m needed to cover the dividends. The DRP has been reinstated, which based on historical take-up could fund around $100m of the gap. It is also hard to not be critical of the company's capital management when less than a year ago they were buying back stock at over $5 per share and now are planning to issue new shares in the DRP at a price more than 20% lower.

All true but still doesn't explain a 40% drop in SP. You have to be a FBU or WHS to adequately explain such a fall. Clearly oversold but the real question that Craig's etc don't answer is why.   

[Ferg]

I have been looking into Spark this evening trying to find a decent dividend yielding stock.  Apologies for the long post.

TLDR: I won't be putting any $$ into SPK until I see an improvement in the financials.  EPS looks soggy, dividends look to be unsustainable, and debts have blown out over the past 10 years for no incremental benefit.

Earnings
Earnings per share is summarised per the 10 year graph below.  From 2015-2022 Annual EPS has hovered around 20-23c until 2023 which jumped to 62c* mostly due to the gain on sale of Connexa, and 2024 EPS dropped to 17c.  For those who look at P/E ratios, the fall in 2024 earnings may explain the recent SP fall given FY24 EPS was 23% below the prior 9 year average of 22c (after normalising the FY23 result).  But a share price drop of 36% in the past 12 months feels like an over reaction.

You cannot view this attachment.

Sales per share has a 0.9% CAGR and NPAT as a percentage of sales has a CAGR of -2.6% per annum over 10 years.  Issued shares have dropped by 0.9% over 10 years (from 1,830m to 1,814m) after deducting 60m shares purchased under the buyback scheme in 2023 and 2024.

Putting all that together, EPS has a negative CAGR of -1.6% per annum over the past 10 years.  This will flip to being positive if EPS in 2025 exceeds 20c.  Let's say SPK can get back to 22c per share, would it qualify as a 'no growth' company?

*Looking closer at the high EPS of 62c in 2023: this included 32c of a tax free gain on the sale of Connexa, which leaves 30c which is out of whack with the prior years' range of 20-23c.  There was also an adjustment to deferred tax assets of $126m on the new Connexa leases (7c per share) and other tax adjustments of $42m (2c per share) which leaves earnings of around 21c per share. {NB: late correction to the per share values}

Dividends
Questions persist if the SPK dividend is sustainable.  The following graph shows EPS versus DPS over the past 10 years.  In 8 of the last 10 years, the cash dividend per share (being the nett payment ignoring imputation credits) has exceeded earnings per share.  For those who look at dividend payout sustainability and/or rely on dividends for income, this might also explain the recent drop in SP.

You cannot view this attachment.

Average payout ratio is 109% of EPS for the past 10 years.  Note this increases to 115% when you normalise the 2023 fiscal year to 26c EPS.  Is this currently a dividend yield trap with annual cash dividends of 27c versus a SP of $3.09?  Time will tell.

[EDIT : eagle eyed investors may notice a disconnect between declared dividends and cash dividends which I use.  My models are built showing returns to investors hence the reason I use cash dividends paid during a fiscal year, not dividends declared that pertain to a fiscal year.  These graphs are a useful subset of the wider return calculations I use.]

Debts
We can look at the SPK history of debt raising to see if that has helped SPK with its growth ambitions, and whether we expect that to increase earnings in future.  The graph below has interest bearing debts versus equity over the past 10 years.

You cannot view this attachment.

The change in debt versus equity over the past 10 years says to me this is a deliberate tactic by SPK.  Why would they do this?  Debts have increased from $692m to $1,619m over the past 10 years, while equity has fallen from $1,778m to $1,590m over the same time period.

This says 4 things to me:
 1. dividends that exceed earnings are a form of returning capital to investors as evidenced by the falling equity balance
 2. SPK are increasing borrowings to pay dividends
 3. despite an additional $927m in interest bearing debts, there has been no noticeable increase in EPS per the first graph
 4. SPK has spent the $893m proceeds received on the sale of Connexa; debts in 2024 exceed 2022 debt levels prior to the sale of Connexa.  $305m was spent on share buybacks, with the balance likely spent on capex given the cash balance has hardly moved from 2022 to 2024.

Connexa
On the sale of Connexa refer AR23 p95: Spark banked $893m on the sale, entered into new leases of $488m, and after deducting assets foregone etc etc, SPK banked a gain on sale of $583m.  The new leases will add {roughly ($488m-$51m**)/15 years) less tax of 28%} ~$21m of higher annual lease costs to SPK, which is roughly 6-7% of FY24 NPAT before making any adjustments for asset depreciation.  Note the mobile tower sites had a small book value so the effect of less depreciation will be minimal.

**Note the ground leases for the mobile sites were novated to Connexa reducing lease liabilities by $51m hence the -$51m adjustment noted.

We won't talk about the (IMO) dodgy tax treatment on this transaction.

[LoungeLizard]

Great piece of work there Ferg - thanks for all the effort put in.

Can't dispute your figures and obviously there's some pause for thought from investors going on based on what was an underwhelming report and concerns about the dividend going forward. However...

As weak as the result was, it doesn't validate the market response. We're approaching $3 per share, a 40% drop. Something doesn't add up there...

Guidance next year is projecting a return to growth, reduced cap spend, $80m in labour cost/opex reductions and - wouldn't you know it - the same dividend as last year, albeit 75% imputed. Guidance can be changed or not met, but it doesn't sound like a sinking ship to me.

I'm max-ed out on SPK shares, so nothing to do now but wait, watch and hopefully at some point, cheer.

[Ferg]

Thanks LL.

I have been musing over this today and I agree it sounds like the SP drop is overdone.  I mentioned in my post the drop in EPS and questions around the dividend as possible reasons for the SP drop but that doesn't explain all of it.  IMO that 3rd chart paints a picture that would disappoint a number of investors - I know I would be mightily annoyed seeing that graph as an investor after SPK banked $893m for Connexa.

Imagine you come out of the back of FY23 with record EPS, debts down a whopping amount which puts them back at levels from ~5 years prior, and all is looking rosy after the sale of Connexa.  Things are looking up right?  Then FY24 rolls around and debts are at an all time high and EPS a 10 year low.  I would be incredibly annoyed if I were an investor at the wasted opportunity by SPK.  Yes they are likely investing in the future of data centres etc but it takes time to make meaningful contributions to earnings with capex.  For me the proof is in the pudding of prior results that taking on debt (to fund dividends + capex) has not made a significant change to EPS in the last 10 years.

EDIT: and in light of the CFO departure, maybe he was pushed / fell on his sword over the share buy back, which in hindsight was an absolute fizzer and a waste of $305m.

My prediction (did you predict I would say that?    ) is the SP will be depressed for a minimum of 12 months, more likely 2 years before SPK delivers and the professional investors re-rate based on evidence.  We minnows are merely observers to the actions of the whales.

GLTAH

[Ferg]

EDIT: and in light of the CFO departure, maybe he was pushed / fell on his sword over the share buy back, which in hindsight was an absolute fizzer and a waste of $305m.

Looking at this further.....

Spend $305m buying 60m shares saves SPK ~$16m p.a. in cash dividends
versus
Spend $305m repaying debt at 7% saves SPK ~$21m in interest less tax of 28% is ~$15m cash saving p.a.

The interest savings will reduce as interest rates fall, and the dividend savings will increase as dividends increase over time.

Mathematically it makes sense to buy back the shares in an environment where interest rates are falling.  But one of these things is not like the other.....

Imagine a scenario where you repay $305m of debt and reduced the dividend for say 2-3 years.  Assuming half dividends for 3 years saves circa $735m.  Boom debts are down by $1b + change over 3 years.  Take the pain on the SP but get the debts down.  De-risk the business and leave it in better shape than you found it.

But no, do Management get incentives for things like return on equity?  Hence the shrinking equity and increasing debts...

[Basil]

I have been musing over this today and I agree it sounds like the SP drop is overdone.

Awesome work Ferg, thank you.  You and KW have made the case very comprehensively this is a no growth company.  Sure, there might be some growth in the future with the data center.  Only time will tell with the current cost of capital if this is eps accretive and when.

Even with equity risk premiums reducing considerably since Ben Graham's day when he argued a no growth company is worth no more than a PE of 8.5 when the risk free rate is 4.0%, I find it hard to get enthusiastic about paying any more than a 10 handle for a no growth company with only a modest prospect of some modest possible eps growth in the years ahead. 10 times say average eps of 22 cps = $2.20.

Oversold or reality starting to bite?  You folks be the judge, but I'm not rushing in anytime soon.

[Ferg]

Thanks Basil

Oversold or reality starting to bite?  You folks be the judge, but I'm not rushing in anytime soon.

Same.  This is both oversold and reality, but with a huge caveat.  It may be considered oversold based on knowledge to hand today, but new knowledge may result in more reality continuing to bite.  Speaking of which, someone very generously shared some research with me today that predicted cuts to the SPK dividend in FY26 and commentary around data centre earnings.  That other dog championing SPK needs to zoom out for a bit given long term investments require a long term view.

[BlackPeter]

Great piece of work there Ferg - thanks for all the effort put in.

Can't dispute your figures and obviously there's some pause for thought from investors going on based on what was an underwhelming report and concerns about the dividend going forward. However...

As weak as the result was, it doesn't validate the market response. We're approaching $3 per share, a 40% drop. Something doesn't add up there...

Guidance next year is projecting a return to growth, reduced cap spend, $80m in labour cost/opex reductions and - wouldn't you know it - the same dividend as last year, albeit 75% imputed. Guidance can be changed or not met, but it doesn't sound like a sinking ship to me.

I'm max-ed out on SPK shares, so nothing to do now but wait, watch and hopefully at some point, cheer.

I guess one could argue the market used to be crazy before and just came to its senses.

But no matter whether the market overvalued Spark in the past and is only coming now to its senses (as I think), or whether the market used to value it correctly in the past and is undervaluing it now (as you seem to think) ... remember:

The market can stay irrational longer than you can stay solvent - i.e. always dangerous to wait for the market to correct its (perceived) "mis"-valuations, and if you just look at the fundamentals and the future earnings potential, than Spark looks still dear today.

Ah yes, and hope is not a good investment strategy, but I am sure you knew that ;
 

[CG]

Ah yes, and hope is not a good investment strategy, but I am sure you knew that ;
 

In the universe where nobody can predict the future hope is the only way forward

[Basil]

In the last decade the low point is $2.63 in May 2015.  Holders will be hoping it doesn't test or really hoping it doesn't break down below that level.

[KW]

The SPK pivot to data centres isnt going to happen - nor will any NZ company looking to build data centres in this country.  Our electricity prices are simply too high.  And the pricing model for data centres is moving to power consumption, not square footage of space (much like how telcos moved from charging by call minutes and texts to charging for data).  Not to mention that NZ simply does not produce enough electricity.

https://www.nzherald.co.nz/nz/politics/what-went-wrong-in-the-electricity-market-and-why-it-will-get-worse-before-it-gets-better/EXTDX7DD6JGNJJFR7UEATLLFRA/

[Red Baron]

The SPK pivot to data centres isnt going to happen - nor will any NZ company looking to build data centres in this country.  Our electricity prices are simply too high.  And the pricing model for data centres is moving to power consumption, not square footage of space (much like how telcos moved from charging by call minutes and texts to charging for data).  Not to mention that NZ simply does not produce enough electricity.

https://www.nzherald.co.nz/nz/politics/what-went-wrong-in-the-electricity-market-and-why-it-will-get-worse-before-it-gets-better/EXTDX7DD6JGNJJFR7UEATLLFRA/

More 'bendz in ze rails' for ze Spark datacentre express?   Just as vell Infratil's CDC eez not zimilarly effected!  Phew!

RB

[Buzz]

The SPK pivot to data centres isnt going to happen - nor will any NZ company looking to build data centres in this country. 

"Pivot to", "isn't going to happen"? Spark already has 16 data centre sites, they've been in the market for many years, and are aggressively growing and expanding that capability. They have all sorts of innovative partnerships around access to energy.

[mfd]

Not to mention, our data centers are not necessarily competing with international centers. I know that Christchurch hospital uses a spark data center and I doubt if they would be happy to outsource to another country for privacy and speed reasons. Similarly, we are waiting to outsource a piece of software to the cloud until there is sufficient capacity in New Zealand - for this use, an Australian center would introduce too much lag.