Buffett Test: Overall Conclusion (FY2022 view)
Quote:
Originally Posted by
Snoopy
First, an apology. It looks like I wrote the equivalent of a short story about Spark over 2021 and never penned the final chapter! This post is the missing 'summary post' for my previous posts 1806, 1807, 1609 and 1810 on this thread. With my forum settings, I have to go back 17 pages to find these posts.
Spark fails the Buffett test, because earnings per share were largely flat over the five year period test period (no eps growth). Meanwhile, net profit margin was showing a declining trend, except for a one off 'profit bump' between between FY2018 and FY2019. This was principally due to falling labour costs, the result of both standardization of processes and automation. But it becomes hard to continue such savings year after year.
Failing the Buffett test does not necessarily mean that Spark is a poor investment though. It just means that we need to look at the investment under different valuation models. If we assume that Spark is a 'no growth' business over the business cycle, then we can look at 'capitalised dividend valuation' as an investment yardstick. Post 1819 summarizes this, based on a gross yield of 6%. I should add this 'capitalised dividend analysis' was done on 24th August 2021, before the dream of interest rates dropping to a permanently lower notch for the foreseeable future was shattered. The Spark share price was $4.84 on that day. That looked favourable against a $5.58 capitalised dividend valuation.
In the last few days (post 1068) I have redone my 'capitalised dividend valuation' for FY2023 using a gross interest rate of 6.5%. If I had known then what I do now, I would not have dropped my acceptable gross yield down to 6% back in August 2021. Reworking my historical gross yield calculation from post 1819 using today's 6.5% gross yield expectation gives me a revised fair historical 'capitalised dividend value' of:
33.46c/0.065 = $5.15 (on 24-08-2021)
So back on 24th August, that $4.84 market value was still below my 'with hindsight' 'capitalised dividend value' of $5.15. But much of the earlier 'margin of safety' from my post 1819 former valuation of $5.53 was gone. This is a good reason why I would never buy a share at what I deem 'fair value'. I always buy for a purchase price 'below fair value' (10% discount minimum), that will allow for any overoptimism I put into my 'fair value' calculations.
A great advantage of using "capitalised dividend valuation" is that. being a 'zero growth' model, any real growth that does occur, we as investors get 'for free'. This brings me to my post 1811 where I look at the contribution of the growth side of the Spark business: Qrious (data processing and AI) , Internet of Things and Spark Sport. That shows me that for FY2021 these 'growth businesses' delivered $10m to the bottom line from normalised profits of $375m. That works out as 2.66% of profits, which I see as margin of error stuff. (The recent closure of 'Spark Sport' puts a further cloud over Spark's growth ambitions. But this is judging historical growth plans with the benefit of hindsight that was not available 'back in the day'). With the growth engine at Spark not really firing, I would say using a 'capitalised dividend revaluation model' for valuing the Spark share is reasonable.
Despite the failure of Spark to jump all the Buffett test hurdles using a FY2021 perspective, I would say the on market trading price of $4.84 'back in the day' offering a 6% discount on my revised $5.15 'capitalised value' was a fair price. The Spark share back in August 2021 was not a compelling buy. But it was certainly not something a portfolio investor looking for income should be looking to sell either.
This post is a result summary of posts 1806, 1994, 1995 and 1997. These four posts are a good snapshot of the way Warren Buffett might look at Spark as a potential investment, by looking at a previous five year investment review window.
The Good: Spark is in a very strong market position as a top three retailer of fixed line, mobile line (both in voice and data) and ancillary services (BT1). Spark also has a strong return on shareholder equity, consistently around the 25% level (BT3).
The not so Good: A very flat normalised profit over five years (BT2), combined with a very flat normalised profit margin over the same period (BT4).
Good enough?: No. All four Buffett Test boxes must be ticked for a Buffett test pass. Only two were achieved.
The result is a 'test failure' by the Buffett test criteria. The Buffett test numbers are telling me this is a 'no growth company'. But that doesn't mean that Spark is a bad investment for a non-Buffett style investor, when the investment metrics are comsidered in other ways.
----------------
Spark is an efficient provider with a peer beating total shareholder return of 11.7% per year (dividends compounding) over the last three years (Spark PR2022, Slide 27). This return beats all but one of the 14 comparative businesses in the Spark peer group comparison.
My recent capitalised dividend model exercise (post 1968) gives a 'fair value' for Spark of $5.24 (based on a 6.5% averaged gross dividend yield). The share closed at $5.05 on 14-04-2023. The crudeness of the capitalised dividend value technique means that no tertiary trained financial analyst would ever admit to using it. But the fact that it is:
a/ Entirely based on real results , with no fudge factors, AND
b/ It incorporates the company directors view (more reliable than any analyst), on what dividend level is appropriate - irrespective of the declared earnings numbers.
These are the reasons why I think 'capitalised dividend valuation' is such a useful tool, often confirmed by Mr Market's valuation of the same company being not so far away. My existing portfolio holding of Spark certainly paid dividends, both literally and figuratively, on enhancing the value of my NZX investments over tax year FY2023. Or perhaps more correctly meant my portfolio returns were less negative than they otherwise would have been :(. But my valuation does come with a caveat.
Whether you look at Spark earnings 'as declared' or 'normalised', of late, they don't cover the dividends paid out (refer post 2032). Since FY2022, Spark have received a 'capital top up' from the sale of a controlling stake in their cell tower network. So any crack in the balance sheet from what I have termed 'over-payment of dividends' has been filled in. But Spark have more or less reached the end of the road for 'surplus assets to sell'. So my advice to Spark share buyers is to keep a close eye on how well earnings cover dividends going forwards.
On the positive side I am assuming the newly blossoming identification verification SaaS arm - MATTR - has zero value. If this turns out not to be so, then a possible future trade sale, or even a float, of this business would provide a welcome cash bonus for shareholders.
SNOOPY
discl: hold