KAPE Technologies (KAPE): Contract assets (MelloBASH)

I sat in on this Mello Event that was promoted by David Stredder during this ShareSoc webinar.

The BASH (Buy, Avoid, Sell, Hold) session was an enjoyable part of the event. Kape Technologies (KAPE) was among the shares discussed and Leon Boros highlighted some unusual accounting involving deferred contract assets. The issue is worth further investigation.

Essentially KAPE’s reported earnings are not backed by cash flow, because its marketing costs are capitalised onto the balance sheet and subsequently expensed through the income statement as an amortisation charge.

The accounting in question follows the implementation of IFRS 15, which altered the way contract revenue is recognised. Here is the relevant explanation from KAPE’s 2018 annual report:

You will see the scope of the policy was changed to cover every product for 2019:

Sales commissions can indeed be capitalised under IFRS 15 as an incremental cost. Other companies (e.g. Avast, see below) capitalise such marketing costs, too.

But missing from KAPE’s small-print is the useful life (or ‘expected customer relationship period’) of the amortised marketing spend. I could not find the exact useful life at all within the reports, which is bad form.

KAPE pays commissions to affiliates for successful referrals, and recognises such costs over its best guess as to how long the customer will remain a customer (note the text “includes the initial contract period and expected renewals” within KAPE’s accounting policy).

KAPE’s amortisation policy looks wrong to me. The duration of commission amortisation ought to match the associated revenue – e.g. if a customer purchases a one-year software subscription, then the associated commission payment ought to be amortised over one year, too.

KAPE’s current amortisation policy would only be justified if the company paid commission in advance to affiliates based on expected renewals – e.g. the customer buys a one-year software subscription, but KAPE generously pays, say, 3 years’ worth of commission upfront based on expected renewals.

I am quite sure KAPE does not pay commissions to affiliates in advance based on expected renewals. As such, KAPE’s reported profits are somewhat flattered by the longer amortisation period the company applies.

The policy has a significant accounting impact. KAPE’s 2019 cash flow statement shows net marketing cash of $17m being diverted to the balance sheet…

… and these contract assets at the 2019 year-end were carried at $29m (i.e. marketing costs of $29m had yet to go through the income statement):

This note shows the balance-sheet movement:

Cash marketing spend of $32m versus amortisation of past marketing costs of $15m gives the earlier $17m cash flow movement. This $17m difference is material to revenue of $66m and adjusted Ebidta of $15m, and left KAPE with negative operating cash flow despite positive reported profits.

Cross-checking how similar businesses account for marketing costs provides extra perspective. Avast (AVST) for example also uses affiliates to sell software and its accounting policy feels correct to me:

Unlike KAPE, AVST amortises its marketing costs “over the licence period, consistent with the pattern of recognition of the associated revenue”.

AVST’s approach means total additions (i.e. the marketing cash spent) and the associated amortisation were reasonably similar for both 2019 and 2018, and the $33m 2019 year-end value carried on the balance sheet represented less than 4% of revenue.

In comparison, KAPE’s $29m year-end value represented a substantial 44% of 2019 revenue.

All told, KAPE’s accounting looks wrong to me and flatters the company’s reported profits. KAPE ought to amortise its marketing costs over the same duration as the associated revenue relating to the commission.

(As an aside, I notice KAPE-owned Cyberghost offers its affiliatesup to 100% in commissions” . No wonder cash flow is negative and marketing spend is so high if up to 100% of a customer’s initial purchase goes to the affiliate!).

Maynard

3 Likes

Great analysis Maynard, it would be interesting to know how 'sticky ’ KAPE’s clients are - they clearly need very high retention rates if the entire initial contract is given away in commission!

Quite the red flag.

Roger

1 Like

I think Webselenese’s audience is indeed uninitiated and simply find its websites through searching for ‘best vpn’.

The logic of the deal seems straightforward enough. KAPE gets an inside view on customer demand for other VPNs, could perhaps influence the editorial (a bit naughty) and will capture the CyberGhost/PIA affiliate payments made to Webselenese. Looking at my original post, KAPE’s marketing costs of $29m compared with revenue of $66m – so KAPE seems to be paying a huge proportion of income to affiliates such as Webselenese.

A corporate website constructed using WordPress is not unusual. Vpnmentor uses WordPress, too, which is impressive for 105 million visits a year. Lots of other popular sites use WordPress (40+ Most Notable Big Name Brands that are Using WordPress)

Agreed, the job ad is worded badly. But I get the impression the role is one of collecting information about the leak, rather than interrogating the actual compromised database.

My searching for Webselense Ltd led me to D&B and an address (132 Begin Menachem Rd, TEL AVIV-JAFFA , 6701101) in the middle of Tel Aviv that corresponds to the office location stated on the Webselense website. The address on the Google My Business sidebar seems odd. With Globo, which proved to be a complete fraud, the official addresses on its website and documentation turned out to be very peculiar locations.

The articles are formulaic and written purely for SEO purposes to find people searching for ‘How to watch [show] for free’. Easy way to attract traffic for conversion into affiliate sales. The site’s product reviews seem pretty comprehensive.

I have doubts about KAPE, too – somewhat shady accounting, somewhat shady products, somewhat shady marketing, somewhat shady owner. But overall you have not convinced me that Webselense has “no substance”!

Maynard