Day 6: Fringe81 (6550)

Category: Ad tech / HR tech, Mkt cap: ¥4.2bn, NTM EV/Sales: 0.92x, Gross margin: 29.1%, 3yr Sales CAGR 10.5%, Founded: 2012

This is installment #6 of my 200-day challenge to do write-ups on 200 Japanese small caps, at the end of which I plan to publish a book (tentative title: Japan Small Caps Handbook). I’m writing these unedited versions in a stream of consciousness style, and they will be refined before being finalized later on.

If you don’t have time, I encourage you to jump to the end of the post for the Key Takeaways. The goal of this newsletter is to fill the information gap on Japanese small caps that have little or no analyst coverage.

Click on the button below to join 219 others to receive these posts in your email.


I’ve been unable to do write-ups over the past two days. I was too bullish when I said I would do one every day. My near-term goal is to do 4 every 7 days.

Last week, I was contacted by a portfolio manager of a multi-billion dollar investment firm. He took interest in my 200-day challenge. We chatted for about 30 minutes. But there was nothing we found in common. He was looking for value stocks trading below net cash. I am focused on growth and tech.

People living in Tokyo have also reached out to me. Feel free to do the same if you have some time for coffee. Substack seems to be connecting me with people I would’ve never met. My life is being distorted by the creator economy.

I woke up this morning to an interesting email from an investor. It was about Fringe81, an ad tech company worth $38 million. The deck felt compelling to me. So, I’m taking a break from real estate tech today, and I will instead try to dig deeper into this micro cap. If this is something that interests you, read on.

Fringe81 (6550): Peer Bonus SaaS “UNIPOS” Gaining Traction, But Shares Discounted Due to Cash Burn

Business overview

Fringe81 used to be a wholly-owned subsidiary of UNITED, Inc., an online media subsidiary of major ad agency Hakuhodo DY Holdings (2433). It was founded with a different name in 2005, changed names to Fringe81 in 2010, and was spun off by means of an MBO in 2013. It was listed on TSE Mothers in June 2017.

CEO Yuzuru Tanaka was a founding member of the Internet Department at Softbank, where he was involved in launching (present-day Yahoo! Videos). He left to be part of the executive team of Netyear Group (3622), with which he was involved from the founding days until pre-IPO preparations.

In 2005, Tanaka became the CEO of UNITED, Inc.’s subsidiary (predecessor of Fringe81 prior to MBO), and he has served as the CEO of Fringe81 ever since.

The company operates two types of businesses:

  • Internet advertisement, which breaks down to (1) ad agency business and (2) media growth business.

  • Unipos, a SaaS peer bonus service, which enables peer-to-peer bonuses between employees.

Q3 FY20 (out Feb. 12) sales weighting in the Internet Ad Business:

  • Ad agency: ¥153mn (44% of segment total)

  • Media: ¥158mn (46% of segment total)

  • Others: ¥36mn (10% of segment total).

The figure below (in Japanese) shows that the ad agency business is picking back up after its post-COVID slump, while the media business remains sluggish. However, management expects an uptick in sales in Q4 from Q3 due to an ad-related tie-up with Eight, a business card management app by Sansan(4443)

Unipos sales increased 47% y/y to ¥106mn from ¥72mn LY. Unipos, a peer-to-peer bonus SaaS, now accounts for 23.3% of total sales.

UNIPOS sales grew at a CAGR of 62.8% over the last 2 yrs. Quarterly recurring revenue reached ¥106mn, meaning that ARR is around ¥424mn. Conservatively assuming a 3yr CAGR of 40%, ARR would reach ¥1.2bn by Q3 FY23.

Conservatively applying half the NTM EV/ARR multiple of health tech peer Kaonavi (¥2.8bn in ARR; trading at 13.5x NTM EV/ARR) would bring the value of UNIPOS alone to around ¥8.1bn (1.6x present EV, incl. ad business). The implied annualized return is 17% over the next three years for UNIPOS alone.

So, why has the share price fallen 70% since its IPO, even though Fringe81 appears to have such a promising HR tech SaaS product?

Products & services

Ad agency business: Fringe81 buys advertising spaces from advertising media (e.g. Google Display Network, Yahoo! Display Ad Network, Facebook Ads, and other DSP service operator and ad network operators).

It then sells these advertising spaces to advertisers and ad agencies with a certain margin added onto the purchase price. Hence, profits are based on selling price subtracted by purchase price of advertising spaces.

Fringe81 distinguishes its service by providing in-house ad technology solutions and analytic services.

Media growth business: Fringe81 operates ad networks and supports media ad businesses. It employs a revenue-sharing business model, whereby media ad operators share a portion of their revenues.

In February 2015, Fringe81 formed an alliance with D2C Inc. to develop and operate the “Docomo Ad Network,” which utilizes D2C’s cell phone subscriber information for targeted ads. Fringe81 receives a portion of ad network sales as part of a revenue-sharing agreement.

Fringe81 also provides advertising product planning and development to help media companies monetize their advertising business. The company’s largest client for this service is SmartNews, which it has been involved in since the launch of their ad business in December 2014.

UNIPOS: A “peer bonus” SaaS launched in May 2017. This is a service where employees can send each other “points,” an in-house currency system, whereby peers can send each other bonuses for their contributions.

Rates are applied to each point (¥1-5/point), which are aggregated at the end of a specified period and distributed to employees as bonus. To increase points, employees can either receive “direct transfers” of points from others or collect “claps,” which are similar to a pressing a Like button, which counts as 1 point.

The key point about this service is that the contribution of an employees is rewarded by peers, regardless of department or status. Hence, even small contributions that superiors might overlook in quarterly evaluation reviews are recognized and rewarded by peers, thereby raising employee engagement.

Market Environment

According to Dentsu’s study, “2020 Advertising Expenditures in Japan,” the TAM of the internet advertising market in Japan was ¥2.2trn (+5.9%) in 2020, a level matching the overall size of the traditional media market (see Fig). In particular, the Internet advertising market grew by 24.2% for e-commerce platforms.

Thus, Fringe81’s Internet advertising business is operating in a fairly promising market. Why has it then been so sluggish? Is it because of tough competition? Is there a chance that the advertising business would be a liability that drags down the promising SaaS business?

On the other hand, could it be that advertising revenues were depressed because of curtailed budgets during COVID. Will they recover going forward?

Competitors in the online ad agency industry include the following:

  • Full Speed (2159): SEO consulting, UI/UX consulting, social media consulting, affiliate advertising, ad technology services, etc.

  • Digital Holdings (2389): Ad agency services, web marketing support, HR support, etc.

  • Aun Consulting (2459): SEO, marketing consulting, PPC management, ad operation support on mobile/social media, etc.

  • Adways (2489): Marketing platform for mobile apps, ad management platform, native ad platform for smartphone, web affiliate services, etc.

  • UNITED (2497): Ad technology services including CHEERZ a service to help artists, TechAcademy, an online programming school, consulting, etc.

  • Septeni Holdings (4293): Digital marketing and media platform businesses, including cloud-based CRM and smartphone advertising services.

  • Cyberagent (4751): Media, online ad, and game businesses, including Abema for on-demand video viewing, Ameba for blogging, Tapple for online dating, AWA for music streaming, and WinTicket for online betting.

  • GMO Ad Partners (4784): Online advertising business, especially ad agency, web promotion, and marketing.

Japanese Internet Advertising Sector: Peer Comparison

Based on a quick look at the financials and valuation, I have two questions:

  • LY sales decline came from a drop in the Internet Advertising business. Was the sales decline transitory—e.g. investments postponed by clients due to COVID—or is there a weakness underlying the business?

  • Why is the equity ratio 0.9%? The company booked an impairment of ¥534mn LY and ¥1,290mn this FY, according to the balance sheet below. Furthermore, the current ratio is 0.82x. Total debt to equity is 2379%. Would it be able to withstand cash shortages should such tail risk transpire?

In short, the company has a promising SaaS product. This is fairly clear based on the steady increase in the number of employee accounts and client companies. However, whether to invest in the stock hinges on the recovery of the advertising business and its ability to deal effectively with the cash burn.

(1) Net loss and impairment losses

Based on the Japanese disclosure of FY19 results (here), net losses stemmed largely from ¥609mn in growth investments for UNIPOS. On the other hand, the company booked ¥295mn in impairment losses on software related to UNIPOS.

The disclosure states as follows: We booked impairment losses on software based on a conservative estimate of the recoverability of software investments during the amortization period, in accordance with accounting guidelines. However, based on UNIPOS’ churn rate, the expected usage period will exceed the depreciation period. As such, we view the investments to be recoverable, albeit over a longer period than the amortization period.

For the same reason, the company booked impairment losses of ¥179mn on UNIPOS software during cumulative Q3 FY20 (disclosure).

In short, it appears that Fringe81 has made heavy investments in UNIPOS software that cannot be recouped during the amortization period. But they believe the investments are recoverable over a longer duration.

(2) Decline in sales

The Japanese disclosure of Q1 FY20 results (here) explains that the company adopted a new accounting standard for revenue recognition. As such, Q1 sales appear to have dropped 60% y/y at first sight, to ¥508mn from ¥1,267mn LY.

But the disclosure states as follows: “When retrogressively adjusting revenues for the new recognition standards, adjusted revenues were up 18.9% y/y.”

Still, even based on adjusted revenues, cumulative Q3 sales (Apr-Dec) were ¥1,437mn (-9.6% y/y) due to sluggish sales in the media business, which the company attributes to reduced ad spend caused by covid. Q3 results (Oct-Dec) were particularly weak with sales coming in at ¥454mn (-25.6% y/y).

(3) Cash burn

Management has taken several steps to address the cash burn [see page 4 and 5 of the Q3 deck here in Japanese].

  • Fringe81 consolidated three domestic subsidiaries as of October 1, 2020 to keep a lid on growing administrative costs.

  • In addition, it privately allotted new shares to Sansan (4443) and raised roughly ¥160mn in December.

  • In conjunction, it issued subscription rights to new shares to be executed from end-December 2020 on a rolling basis, meaning that capital will start to increase in 2021 and lift the extremely low equity ratio.

  • It also started a tie-up with Sansan, whereby Sansan would incorporate its data analysis, sales, and product development into UNIPOS, and Fringe81 would support Sansan with renewing its Eight Business ad delivery system.

  • Management decided on reducing office space to cut rent expenses starting from April 2021, which is estimated to boost profits by roughly ¥100mn annually. It also dissolved its German office to focus on UNIPOS (saving of roughly ¥110mn). In the advertising business, the company withdrew from loss-making operations (savings of roughly ¥100mn).

Key takeaways

  • Fringe81 is an ad tech and HR tech company with a struggling advertising business, which it attributes to reduced cash spend caused by COVID-19. The HR tech business consists of SaaS called UNIPOS, which is an internal currency system for employees to give small bonuses to each other.

  • Employees give other employees “points” either through direct transfer or “claps,” which are analogous to pushing the “Like” button on posts made by others. Points are converted to cash and later distributed as bonuses. This system allows employees to be recognized for their contributions, which may otherwise be overlooked by superiors during performance evaluations

  • Fringe81 now has 470 client companies and 60,000 accounts using this internal currency system. The number of accounts has increased by 186% over the last two years, demonstrating robust growth. 3Q revenue reached ¥106mn (ARR of ¥424mn) with sub-1% churn rate.

  • Still, the company’s shares have dived 80% since IPO-ing in June 2017, from ¥1218 to ¥355 (as of Mar. 26 close). This is because management is grappling with several issues: namely, (1) ongoing net losses (& impairment losses on software), (2) falling sales in advertising business, and (3) cash burn.

  • To address these issues, management has taken a flurry of measures such as consolidating subsidiaries, dissolving a business, reducing office space, issuing new shares, and forming a tie-up with Sansan. Should these steps prove effective, Fringe81’s stock could be poised for a re-rating.

Score: 4/5 [Reason: Compelling value thesis based on longer-term return potential from SaaS product that has taken off. However, I refrain from assigning a 5 due to competitors with internal currency systems, e.g. Disco’s “Will”,’s “OOIRI”, and Link & Motivation’s “LIMO”, meaning that the market could grow crowded down the road. See here (in Japanese)]


Income statement

Major shareholders

Relevant links


You can support my 200-day challenge by buying me coffee here.

Feel free to email me:

Archives (Past 3 Days)

Day 0: The Creator Economy: “200 day challenge”

Day 3: Nihon Jyoho Create (4054): Category: Real estate tech, Mkt cap: ¥27.7bn, NTM EV/Sales: 9.85x, Gross margin: 73.4%, 3yr Sales CAGR 9.3%, Founded: 1994

Day 4: Property Data Bank (4389): Category: Real estate tech, Mkt cap: ¥11.2bn, NTM EV/Sales: 4.32x, Gross margin: 49%, 5yr Sales CAGR 18.5%, Founded: 2000

Day 5: AZoom (3496): Category: Real estate tech, Mkt cap: ¥12.5bn, NTM EV/Sales: 2.43x, Gross margin: 41%, 3yr Sales CAGR 37.5%, Founded: 2009