Spark – historically a company with a take-to-the-bank record of meeting its forecasts – suffered a rare guidance miss as its FY2024 net profit fell 21% (or 71% if you didn’t adjust for the one-off windfall from its cell tower sale) to $316 million. The telco blamed the soft economy and, in particular, a slowdown in its IT services revenues as big firms and government departments put projects on hold.

The economic slowdown – touch wood – is a transitory pressure on Spark’s bottom line and profit payout. But Jarden’s Arie Dekker has also raised a more structural issue: Spark wants to invest more in data centres – a rare bright spot in its FY2024 result – but such investments could see its profit payout trimmed. And that would come at a time when interest rates are falling, sending investors back on the hunt for dividend-paying stocks. In the telco sector, analysts see Chorus as a safe bet to keep increasing its dividends as it enters its lower-spending, post-fibre rollout costs era.

Dekker had been expecting Spark to cut its full-year dividend when it announced its full-year result in August. In the event, it met its full-year dividend guidance of 27.5 cents per share (cps).

He thinks Spark has overpaid its dividend in recent years and sees a “reset” on the way.

2. $1b spend, but no special advantage in data centres

Data centres were already a booming business worldwide before AI (artificial intelligence) came along and super-charged growth. Spark’s data centre revenue increased 54.2% to $37m in FY2024 after it completed an upgrade that boosted its Takanini facility to 12 megawatts (data centre size is measured by peak power consumption). It also recently won consent to build a 10MW data centre at Dairy Flat, with the potential to boost that to 40MW. Chief executive Jolie Hodson says there are plans to build capacity to 118MW – which will cost up to $1 billion over the next five to seven years.

The data centre business holds a lot of potential to help Spark replace the areas where revenue has dried up (including the well-catalogued decline of its once cash-cow voice business and the disappearance of the nearly $1b it cumulatively banked in near-monopoly profits from its 50% share of the transpacific Southern Cross Cable between 2000 and 2019).

“But it’s not an easy industry. It’s not like Spark has a competitive advantage in New Zealand in data centres by virtue of being the main telco. Data centres are a completely different business and there are a lot of gorillas in that industry doing exactly the same thing,” Han says.

A couple of those gorillas, Microsoft and Amazon, are building giant data centres in northwest Auckland – their first in New Zealand, with multibillion-dollar budgets – while the regional big dog CDC Data Centres (half owned by One NZ owner Infratil) already has two up and running, with expansion plans.

3. Short-selling

While Spark’s shares have been trending down this year, with its profit downgrade in May then its profit miss reported in August, its recent nine-year low has been driven, in part, by an apparent short-selling campaign – essentially, a group of investors betting the stock will fall.

In our market, that’s a relatively rare phenomenon, analysts say.

The exact extent can’t be quantified because, unlike many markets, the NZX does not maintain a register of stocks that have been short-sold, and the ASX register (where Spark is dual-listed) does not include the number of shares in firms that have been shorted across the Tasman. That means analysts have to rely on intel directly gathered from the market.

While short-selling is transitory, it comes at a bad time for Spark, with a major global index, based on market cap and used as a guide by some fund managers, on the brink of its annual revision.

But in a confidential note to clients this month – sighted by Business Herald stablemate BusinessDesk – Jarden says around 40 million Spark shares were short-sold.

Spark had dropped out of the FTSE All-World Index and would be expected to leave the MSCI index in the coming weeks, which would inevitably cause selling in Spark shares.

“The pricing for the [MSCI] change will be determined over the last 10 days of October, with the change being implemented on the last Friday of November,” wrote Jarden director Jeremy Ashworth.

“This is expected to result in the sale of 90 million Spark shares.” (While Spark has a free float of some 1.8 billion shares, its average daily trading volume over the past 12 months has been just over 1 million, with its busiest day 18.4 million.)

‘Oversold’

“The negative factors and headwinds have basically all come at once, one on top of another, over the past six to nine months,” Han says.

“There are plenty of reasons for people to be negative about Spark. And looking at the share price, the negative view is winning at this point.

“But my view is that it is oversold,” he says. His fair value estimate is $4.90 – 60% above Spark’s recent trading price.

“All the downside risks are more than reflected in the current stock price.”

And, just quietly, the depressed share price has pushed up the yield, for those inclined to load up for the promised 27.5cps FY2025 payout.

“Spark’s dividend yield at its current price is about 8%, which is quite attractive,” Han says.

Three things that could steady the ship

“I know the New Zealand economy is suffering and business conditions are tough,” Han says.

That was out of Spark’s hands.

“But there are three factors within management’s control to steady the ship in terms of cashflow – and that, in turn, will allow them to maintain the dividend at the current level.”

1. Cut costs, especially in IT

Spark needs to cut costs, especially in the IT business, where revenue potential has not been as great as management thought it would be.” Spark’s IT services teams have been a point of focus as the telco pushes to cut 10% or $50m from its labour costs (which increased slightly in the year to June) in the current financial year. But Han says Spark “needs to completely reconfigure its IT cost base. I do believe there are a lot of costs to be taken out in that whole IT sphere.”

And he blames IT services for a lot of the telco’s woes. “Spark is your market prototype defensive stock. And, uh, that defensiveness is underpinned by its strong and very resilient mobile business, which I think is about half of group revenue and more than half in terms of earnings,” Han says. (See revenue breakdown below.)

“I think that foundation excused a lot of subpar performance in Spark’s IT businesses; cloud and managed data and security. They have been struggling for a while,” Han says.

Forsyth Barr’s Aaron Ibbotson had a similar take in a September research note. “Spark’s historically dominant position within private cloud in general, and government cloud in particular, is also deteriorating, driven by a combination of structural, technical and competitive factors,” he wrote.

Jarden’s Dekker said the rise of public cloud (hosted by the likes of AWS, Google and Microsoft) was squeezing Spark’s private cloud business.

“These businesses were supposed to provide a growth angle. Instead, they were responsible for a downgrade,” Han says.

“They’ve taken a lot of time, money and management resources.”

2. Pull back on billion-dollar data centre plans

Hodson’s push toward 118MW doesn’t hold a lot of truck with Han given, as noted, he sees it as a highly competitive, Big Tech-dominated market where Spark, in his view, holds no special advantage.

“Pulling back on the data centre growth strategy is another thing that’s within the control – whether it’s in terms of the expansion timeframe or in terms of the capital allocated to it within that timeframe.”

Where Jarden sees potential for Spark to cut its dividend to help fund data centre investment (and arrest the trend of the profit payout increasing while free cashflow shrinks), Han says hands off.

He sees the dividend as central to Spark’s appeal (which is where we came in). It should curb its data centre ambitions if that’s what it takes to maintain the profit payout at 27.5cps.

“Hopefully that will say something along the lines of, ‘We are going to quarantine our dividend-paying ability from the data centre strategy’.

“As things stand, they haven’t earned the right to spend a few hundred million dollars on their data centre strategy.”

Says Dekker: “There have been some super-meaningful data centre transactions in Australia. But the New Zealand market is a lot smaller.” Could Spark’s data centre revenue approach mobile? “Certainly not any time soon,” he said. Spark’s data centre market share was in the region of 20% to 25%. But it was not a small, fixed number of players like mobile.

3. Manage capex more tightly

“The third thing is to manage working capital and capital expenditures more tightly,” Han says.

Spark has already leaned in this direction. Capital expenditure is also being cut in the new financial year (which started on July 1), from $513m in FY2024 to between $460m and $480m (finance director Stefan Knight will leave on December 20, so the final tally will be driven by his yet-to-be-appointed successor).

At the full-year result, chair Justine Smyth says, Spark planned to spend $1b on data centre expansion over the next five to seven years.

Over the current year, the telco will spend between $70m and $90m on data centres.

Beyond that, Smyth says, a potential hybrid capital notes issuance would help fund its growth investments in the near term. “We will also explore other equity funding options such as capital partnerships.”

The 10MW phase one of the surf lagoon data centre at Dairy Flat, for example, will be funded from existing capex, while the potential upgrades to 40MW could swing on roping in a third party.

“That mechanism will have to involve capital partners. They can’t do this by themselves,” Han says.

Spark revenue breakdown

• Mobile $1.474b (2023: $1470b)

• Broadband $613m (2023: $626m)

• Procurement and partners $548m (2023: $584m)

• Voice $180m (2023: $231m)

• IT products $527m (2023: $509m)

• IT services $165m (2023: $194m)

• High-tech $79m (2023: $$65m)

• Data centres $37m (2023: $24m)

• Other operating revenues $136m (2023: $172m)

A big part of the revenue and profit tumble was down to the fact the telco’s FY2023 result benefitted, to the tune of $583m, from the sale of 70% of the passive assets of its cell tower network (there was also a negative one-off, with $54m in costs associated with the closure of Spark Sport).

On an adjusted basis, revenue was down 1.2% to $3.861b, net profit fell 21% to $342m and earnings before interest expense, tax, share-based payments, defined benefit expense, depreciation, amortisation and impairment (ebitdai) dropped 2.5% to a below-forecast $1.163m.

Chris Keall is an Auckland-based member of the Herald’s business team. He joined the Herald in 2018 and is the technology editor and a senior business writer.