Thread by @stocks102

THREAD: Long Calnex #CLX

Calnex is a recent IPO on AIM (Oct 2020). It’s already had one ‘trading ahead of expectations’ statement. It seems unknown – based on there being no ‘Discussions’ at all on Stockopedia about it at the time of writing.

The company provides test kits to

telecom operators like Verizon. Verizon uses these devices to make sure, ahead of time, that their telecom systems can deal with the huge ramp up in telecom traffic which 5G is about to bring.

What is 5G? This is the new standard for telecom networks, with greater-than-ever

bandwidth, and faster-than-ever download speeds. This will facilitate a whole host of new technologies (not just mobile phones) - for example autonomous vehicles will communicate with their control centres via 5G. Much more data will flow through telecom systems as a result,

and telecom operators must be certain, ahead of time, that their systems can cope with this rush of data. The advent of 5G is a 10 year tailwind for Calnex; but after that there’ll be 6G (probably from 2030 onwards) then 7G etc.

Calnex owns this niche (of providing test kits

for telcos to make sure their systems don’t crash ahead of the data ramping up for 5G).

Gross margin (c 80%) and EBIT margin (c 30%) metrics are strong – as is cash generation - because, a bit like Fevertree, everything (mainly manufacturing and sales) is outsourced.

But, crucially, R&D is not – this is all in-house.

Customers are diverse and sticky. The top 10 customers by revenue (who together account for c 50% of group revenue) have an average 9-year relationship, with repeat customer business averaging over 80% of revenue between FY18

and FY20. Repeat business exists in part because software upgrades – following the initial hardware purchase by customers - are drip fed out over years.

The sole analyst (at Cenkos) has 0% revenue growth in for H2 2021 versus H2 2020 (the company has a March year end, so H2 is

Sept – Mar). Yet H1 2021 growth was 37% versus H1 2020. There is no reason for revenue growth to collapse in H2; in fact the business is H2-weighted. Thereafter analysts have 6% revenue growth in for 2022 - yet revenues have been accelerating in the last few years,

not decelerating (growing at 25% in 2019, 30% in 2020 and 37% in H1 2021). The shift towards 5G is now well underway so, again, this 6% revenue growth prediction in 2022 is likely to be too low. Since it floated in October 2019, the company has already said it is trading ahead

of expectations (in the November 2020 interims) – so it is adept at managing expectations already, it seems. Also the tailwind - of telecom operators and others needing to test their systems ahead of their launching 5G - is multi-year, so upgrades may continue for years.

At 126p, Calnex shares trade at 33x March 2022 eps of 3.8p; but this eps number is likely to be upgraded and the operational gearing inherent in the business model (with manufacturing and sales largely outsourced) means the upgrades could be substantial. After all, if most of

your cost base is salaries to the 50 or so people in the R&D department, then these do not go up just because you have sold more testing kits. This is why, in H1 2020, when revenues rose 37%, PBT was up by some 90%.

Finally, a new kind of customer is emerging for Calnex: non-telecoms companies. For hyperscale companies like Amazon and Facebook have ever-expanding data centres which need precise timing measurements that Calnex’s testing kits can also be used (ensuring any data glitches can be

caught ahead of time). These customers have moved from 14% of revenue in 2018 to 22% in 2020. If this type of customer really takes off, this is another reason for eps forecasts to rise materially.

If the world falls apart again then it has net cash on the balance sheet

(£9m net cash in 2021, say Cenkos). Also in 2008 and in 2009, Spirent plc, a sector bellweather in telecoms testing equipment, actually grew revenue in both years - so the sector may be economically resilient.

To find out more, go to piworld.co.uk for the results

presentation, and researchtree.com for the 84 page Cenkos note.

/end