It seems to me that GTK’s earnings have been smashed by a confluence of temporary factors (some inside management’s control, and some outside their control), all of which I’d expect to self-correct in the next few years. The group’s headline EBITDA margin has gone from ~30-32% (up until 1H’18), all the way down to ~21% (2H’19), due to no less than five separate forces:
-On the revenue side, two things have happened:
i.The shift from perpetual license sales to subscription sales results in lower upfront revenue in exchange for greater back-end revenue (and greater overall customer lifetime value), which temporarily pulls down revenue (and can be seen in GTK’s utilities perpetual licence revenue halving from FY18 to FY19), and
ii.Project revenues have gone through the floor as utility clients in the UK and particularly Australia defer major IT/billing system upgrades due to adverse and volatile operating environments, brought about mostly by political idiocy (introduction of price caps in both markets, and then in the UK there’s significant additional uncertainty due to Brexit, the upcoming election (Labour wants to nationalize the big 6 energy distributors), and the fact that the big 6 (4 of which are GTK clients) have been engaging in long overdue but highly disruptive consolidation among themselves which has led to them to delay the rollout of GTK’s services). This is borne out in the 2H’19 results – project revenue was only $7.6m, whereas the average over the last 7 halves had been $12.4m (this ~$5m hole in FY19’s project revenue, I think, is essentially the reason management missed guidance of >$31m EBITDA for FY19 – stripping out bad debt (see below), they delivered ~$27m EBITDA).
-On the cost side, three things have happened:
i.Management scaled up headcount aggressively, particularly in the UK, from 2H’17 to 1H’19 to service the huge rollout of its platforms in the UK (Junifer to service the smaller energy clients, and Velocity to service the big 6 clients). Employee numbers went from 429 (2H’17) with an annualized cost of $44m, to ~560 employees today at an annualized cost of $62m.
ii.In addition to the headcount scale-up for the UK, they’ve put cost into the US airports business (new CEO + support to chase deals), plus opened a Singapore office as a beachhead for Asia (currently next to no revenue, but to be used to service major medium-term opportunities that are coming down the pipe in Japan, Korea and other markets which are starting to deregulate).
iii.Finally, FY19’s headline EBITDA of ~$25m included $2.8m of costs that shouldn’t recur into perpetuity, being: a) $1.8m bad debt expenses due to four smaller UK utility clients going bust as well as a $600k bad debt expense in CA Plus, and b) $400k operating loss in CA Plus which has now been folded into the broader Airports business and is breakeven. Maybe FY20 will be another bad year in which a few small UK utilities clients go to the wall which leaves GTK with some bad debts, but eventually the market will settle down as markets tend to do.
As a result of the above five factors which triggered three profit downgrades through FY19 and ambivalent FY20 guidance, GTK has gone from loved to hated – the analyst briefing (https://www.gentrack.com/wp-content/uploads/2019/11/Audio-GTK-Investor-Presentation-FY19-results.mp3) was one of the more entertaining ones i've heard in a while as you could literally hear the despair and disgust in the voices of the analysts because GTK management didn't spoon-feed them crystal clear guidance for FY20. With GTK having gone from loved to hated, the valuation has mostly round-tripped: the stock went from cheap (IPO through mid-2016), to roughly fair value (2017), to nosebleed valuation (2018, >20x EBITDA), and has now reverted back to what feels like roughly fair value (~13x last 6 months normalized EBITDA, excluding bad debts) for this kind of business.
While today’s headline valuation of ~13x run-rate normalized EBITDA doesn’t sound particularly cheap, if you value GTK on an EV / recurring revenue basis to remove the temporary impacts of their EBITDA margin being crushed from ~30% to ~20%, I think right now it’s marginally on the cheaper side of fair value – given 2H’19 recurring revenue run-rate of $81m, and assuming ~10% growth into FY20 (noting that it organically, sequentially grew at 7.4% from 1H’19 to 2H’19, i.e. a ~15% annualized rate), gets ~$89m FY20 recurring revenue, which yields ~4.5x forward recurring revenue. I think 10% growth in FY20 recurring revenue is pretty conservative considering that utilities RR is organically (i.e. excluding Evolve acquisition) growing at 15-20% (my last 4 halves of utilities recurring revenue, excluding Evolve, are $25.2m, $27.4m, $28.9m, and $32.8m, so they exited FY19 growing organically at 20%!) and Airports recurring revenue should grow strongly with Orlando going live late in FY19 – in total, that 10% could easily be more like mid-to-high teens. If I plug in a 15% FY20 growth in recurring revenue, then GTK is trading today on a ~4.2x forward recurring revenue multiple – that is leaning cheap.
While ~4.2x forward recurring revenue ‘feels’ about right (or perhaps slightly cheap) for a business of this quality (diversified customer base, very sticky revenue and low customer churn), what really piqued my interest from FY19’s results presentation was slides 10-11. Prior to FY19’s results presentation, GTK management had consistently presented “long-term organic EBITDA growth of 15%+” as what I thought was, essentially, an aspirational target that wasn’t grounded in realistic operating assumptions. Slides 10-11, however, changed that significantly:
-Slide 10 says two things:
i.GTK’s meters under management with smaller challenger brands using its Junifer product is expected to CAGR at ~15% as they continue to take share from Big 6 incumbents, meaning today’s ~6.3m meters operated with GTK software will go to ~11m by FY2023; and
ii.GTK’s big 6 clients who are under contract (Eon + nPower, EDF, and Ovo which recently acquired SSE) – but have yet to substantively roll out GTK’s services due to major consolidation activity (Eon buying nPower just closed in Nov 2019, Ovo acquiring SSE occurred in Sep 2019) – are expected to gradually roll out GTK’s product, going from no meters today to ~4m meters by FY2023. The salient point here is, GTK has the contracts with these big 6 clients already but is earning next to nothing currently because rollout has been delayed; this will eventually and inevitably change.
-Slide 11 says:
i.There’s a significant opportunity for GTK to cross-sell its existing customers with additional products, which have either been recently acquired (e.g. Evolve) or which have been developed internally at GTK and are about to be rolled out (e.g. market connector, meter data etc.)
ii.It is significantly easier to cross-sell once customers are on the SaaS/cloud offering, as opposed to the perpetual licence model, so this should become easier over time.
My back-of-envelope math suggests that GTK’s UK utilities business conservatively did about $37m recurring revenue in FY19 (total FY19 utilities RR was $68m, Australia’s $22.7m was stated as including zero project revenue so I conservatively assume it’s 100% recurring, and I assume NZ’s $8m is pretty much all recurring as it’s a mature market for GTK). I say ‘conservative’ because, if I simply take the $68m FY19 utilities recurring revenue and divide it up according to each geography’s meters under management, UK would be closer $45m (i.e. 70% of $68m). At $37m UK RR with an average of ~5.8m meters under management in FY19, I calculate GTK earns about $6.50 recurring revenue per year per UK meter under management currently. If slides 10 & 11 play out as expected, then through FY2023 I’d expect:
-Today’s 6.3m meters under management will CAGR at ~20% (i.e. 15% CAGR of independents, plus Big 6 under contract going from nothing to ~4m); and
-Revenue per meter under management will grow, due to:
i.Cross-sell of new products into existing customer base; and
ii.Big 6 utilities using the Velocity system will, I expect, pay more per meter than the existing UK client base which is almost entirely using Junifer, which is the cheaper ‘out-of-the-box’ solution.
If I run a simple scenario where UK meters under management grows at 20% p.a. for 4 years, and revenue per UK meter under management grows at 5% (due to cross-sell and more expensive Velocity being used by Big 6), then UK recurring revenue will CAGR at a monstrous ~25-30% (going from ~$37m today, to ~$105m by FY2023). Applying a 5x multiple to incremental recurring revenue created (i.e. the ~$70m additional), that’s ~$350m of value creation relative to today’s $400m market cap – meaning you almost double your money just on the UK embedded growth roll-out story alone, ignoring any growth in GTK’s other businesses (Aus utilities has plenty of room to grow as they cross-sell additional features and bring Junifer in for smaller operators, the Airports business is starting to scale, and Asia could be a longer-term growth story), and ignoring the fact a roll-out of this scale, particularly the Velocity platform for Big 6 UK clients, will generate material project revenue along the way.
Clearly, that’s a bull scenario case assuming all goes to plan, but even if you take a hacksaw to those forecasts and assume much more conservatively that: i) independents scale at only 7% instead of 15% p.a., ii) Big 6 only goes to 2m rather than 4m meters, and iii) 2% annual growth in revenue per meter, the UK utilities recurring revenue base still almost doubles from now through FY2023 and, at 5x recurring revenue value creation multiple, that’s $1.60 per share (enough for a ~9% CAGR in share price), and again this assumes GTK’s various other businesses do nothing for four years.
In summary, while GTK doesn’t look especially cheap today on an EBITDA basis, I think it’s on the cheaper side of fair value based purely on a recurring revenue multiple which looks through their temporarily crushed EBITDA margin. The UK embedded growth story (i.e. ‘embedded’ because they already have all these clients under contract, and are simply waiting for them to expand over time in the case of the independents, or, in the case of Big 6 customers, to hit the “go” button on implementation), combined with cross-sell (more revenue per client) makes this a pretty compelling entry point, I think – if the UK utilities roll out goes even remotely to plan as slides 10-11 suggest it might, then GTK has years of double digit growth in its revenues ahead of it, and if that plays out then I suspect it’ll trade at a valuation in excess of today’s 4.5x recurring revenues (probably somewhere between 5-6x). Even if UK bumbles along far worse than what GTK management think, you still probably make a high single digit or low double digit return from here – heads you win big, tails you still make out ok. In the highly unlikely event the UK roll-out / imbedded growth story turns out to be a figment of GTK management’s imagination and never happens, then I’d expect GTK would take corrective action in the form of redundancies – reversing the big headcount growth of the last two years – and the EBITDA margin would track back up toward 30% over time (even though revenue growth would be low single digits).
As for risk, it’s worth mentioning that the board here is about as good as you’ll ever see for a $400m market cap business – it speaks for >25% of the equity so no questions on alignment, contains two private equity software veteran investors with particular expertise in this space, a software business veteran exec, plus GTK’s ex-CEO – and all except the ex-CEO participated in last year’s rights issue at $6.19 so I think they’re pretty motivated to restore and grow shareholder value here. That $6.19 rights issue looks like a stroke of genius now, too – it cleaned up the balance sheet such that GTK is now completely unlevered with capacity to debt-fund small acquisitions, and they sold the equity at a high valuation so it wasn’t horribly dilutive. They’re clearly managing the business for the long-term (e.g. ramping up cost base in UK, Singapore & US to prosecute growth opportunities), the acquisitions they do tend to have industrial logic (e.g. cross-sell to other existing clients, expand offering etc.), and they understand that the way to grow value here is to drive recurring revenue. HG is one of the world’s foremost software PE investors and they’ve invested $45m into GTK at an average of $4.08, and given they target 20%+ returns I’m pretty confident they together with Leigh Clifford (own >20% combined) will be holding management’s feet to the fire to ensure the market opportunity in front of them is properly prosecuted.
As a postscript, it’s worth mentioning that having EoN, Ovo and EDF as clients in the UK is potentially quite significant – EoN and EDF especially are multinational behemoths that have businesses across Europe and other parts of the world, so GTK having its foot in the door in the UK gives them a significant opportunity to help these major clients perform IT/billing upgrades in other international markets as they deregulate over time. This has already happened with EoN, as GTK is contracted to do Romania for them as well as the UK.
Lastly, while I personally don’t think FY20 EBITDA really matters much in the grander scheme of things given the opportunity here is significant on a 3-5 year time span, it’s seems like the guidance for FY20 EBITDA to be flat relative to FY19 is ‘soft’ guidance. Management hated having to guide down three times throughout FY19 – “wedefinitely do not enjoy downgrading our guidance. And I don't want to be doingthat again” was the quote from the conference call – so logic would suggest they’ve deliberately set the bar of expectations low out of the gate this year to avoid having to guide down again. They also inferred that most of the bad news is behind them: they said about bad debts / customers going bankrupt that “ It's our expectation that's mostlybehind us and that we've seen the worst of that”, and that they’re “hoping to get better news as we go throughthe year and put the uncertainty behind us”.
(20min delay)