Why I bought more UBNT

I bought a few more March 55 calls earlier this week to take advantage of the temporary sale of UBNT shares. Prior to the Citron short report on Monday, I was struggling to understand why the price had gone down during the last few weeks from the mid 60s to the mid 50s despite excellent financial results and guidance. Now we know that Andrew Left and his friends were busy shorting (selling) shares knowing that they would issue a short report and drive the price down. Given their big short position, the short report gave them the volume they needed to cover (buy back) their positions. This is the opposite of a pump and dump and leaves the amateur short investors holding the bag while Left and his crew are on to their next target. The closing price of UBNT last Friday (the day before Citron’s report was published) was 54.95. The short report drove the price down as low as 47.78, but by the end of the week, the price bounced back to 54.18. So the net loss for the week in which a well-known short with a decent reputation called the company a complete fraud was only 1.4%. I hope some of the investors on this board who are more knowledgeable about the company took advantage to buy on this temporary dip. I believe Pera smartly took advantage and bought back a lot stock at a much cheaper price than would have been possible without the short report. As some of you may know, SEC Rule 10b-18 states that the company cannot buy back more than 25% of the average daily volume (https://en.wikipedia.org/wiki/Share_repurchase). By boosting the volume and bringing the price down, it allowed UBNT to buy back more share cheaply. We know that unlike other companies, when UBNT authorizes a buyback, they use it. When they authorized a $50M buyback on September 5, they also had $1.5M left on their previous authorization. We learned this week that between September 5 and September 18, they used up the remaining $1.5M of the previous authorization and used up $22.4M of the newly authorized $50M buyback. As such they bought back $23.9M of stock between that time. Let’s be conservative and say they bought back shares evenly over that time and had an average purchase price of 57. That would bring down the outstanding shares from 81.5M to 81.1M. With the authorization of another $100M buyback, the open authorization as of September 18 was $127.6M. I suspect that the company was able to buy a lot of shares cheaply given the higher average volume. Given the 25% volume restriction, I would assume they are not done buying yet, but are likely maxing out their allowed daily purchases. If they can buy back the $127.6M at an average of 53 per share, that would retire another 2.4M shares. That would bring the outstanding shares down to 78.7M shares. Pera has 56.3M shares, so after the buybacks, his ownership will go up to 71.5% (up from 69.1%). His ownership percentage would be higher than before he sold his 1M shares (57.3M shares over 81.5M shares = 70.3%).

The major benefit of having lower outstanding shares is that there are less shares to split the profits amongst. The previous guidance would have assumed that the share count remained steady (as they only had $1.5M in open buyback authorization at the time guidance was given). You can multiply the 2018 EPS guidance and multiply it by the outstanding shares and then divide it by the new outstanding shares. This would take the previous EPS guidance from a range of 3.7 to 4.3 (4 midpoint) to a new range of 3.83 to 4.45 (4.14 midpoint).

I won’t go over all the points that the Citron report goes over as there have been many good posts here that have refuted their claims, but will go over a few:

Why no CFO – see this tweet that shows a quote from Pera explaining this: https://twitter.com/emilio_gold/status/910337525306331141/

UBNT needs a CAO (Chief Accounting Officer) that handles the accounting and the financial controls. A CFO would be needed if the company were more interested in raising money, having investor presentations, dealing with wall street analysts, making acquisitions, investing cash, managing debt, etc. UBNT does not do a great job of dealing with analysts and between the CAO and Pera, they manage the other typical CFO tasks.

Why so few Board members – This has been discussed here: http://discussion.fool.com/1069/stock-comp-and-the-board-3271366…

Pera owns 70% of the company and wants to control it and not have an expensive CFO and Board members. There are pros and cons to this as it keeps costs and stock compensation down but decreases governance and gives Pera total control. Pera is very frugal with shareholder money. If he were to hire an experienced CFO and more Board members, they would make millions and get a lot stock comp each year. This would lead to dilution and increased SG&A costs. The company would be less profitable and would likely move more slowly. Did you notice how quickly the company approved the $100M buyback? Imagine how much slower this would have been if they needed to discuss this with 10 various Board members and a CFO. Instead, the 3 other Board members just rubber stamp Pera’s call to buyback more stock. With Pera being a 70% owner, this is fine with me. I am happy to invest in a company that reduces its share count rather than increases it with excessive management and Board stock options.

FrontRow is garbage – Even if it is not a success, FrontRow is not built into the 2018 guidance. The R&D expenses have already been taken, so there is not much risk here and certainly not a reason to short the company. Besides, I have posted a few reviews on FrontRow and it has mostly been positive.

This positive review by Jonathan Morrison has over 584K reviews: (https://www.youtube.com/watch?v=7wGu96MzOzw&feature=yout…) That’s a lot of eyeballs seeing a very positive review of FrontRow. I am sure a few will even the product.

UBNT has continued to improve FrontRow. They recently upgraded the firmware with a stopwatch app and a translation app. See here: https://community.frontrow.com/topic/139/firmware-update-v1-…

The translation app is pretty cool: https://twitter.com/Latinvixen/status/908508885056618499

Why is inventory going up – This has been discussed multiple times and has been addressed by Pera on multiple earnings call. This Seeking Alpha article tries to go after the company on inventory levels: https://seekingalpha.com/article/4108802-ubiquitis-ever-chan…. I think the comments are actually a much better read than the article itself. Some resellers actually describe why inventory levels have gone up. In summary, the WISP community has complained for years that there is never enough inventory for them to buy. This leaves money on the table, so increasing inventory levels is a rational reaction. See some of the comments here:

“In late fiscal 2014 through fiscal 2016 it appeared the company was very satisfied with current inventory levels and appeared to be saying that the days of stock-outs at distributors were over. We have a difficult time understanding what the justification is for the company building inventory out to almost 15 weeks worth of sales.”
From what I’ve read, WISPs have never been satisfied over inventory levels. I still hear some complain, though this may not be from recent experience.
They’re also expecting decent revenue growth this year, which requires more inventory.
You missed some comments by Pera on investing in inventory as a way to make some productive use of their overseas cash horde. Larger production runs allow for volume savings, increasing margins.

Three of the contributing problems with inventory have been with new product cutovers and not enough inventory of the old product to keep the pipeline filled while waiting. This has been a huge issue when the new product has a defect and take way longer to release if it has to go back to manufacturing.
The second issue is the success of new products. Many time Ubiquiti has underestimated the demand and there is delays getting enough product to refill distributors. It also happens when Ubiquiti gets unexpected, massive orders that weren’t projected. Then the rest of us smaller quantity purchasers have to wait.
The third issue is that many WISPS are seasonal due to weather. That creates huge bumps and actually means Ubiquiti has to save for the cycles which might be 6 months.

As a Ubiquiti reseller we can say our biggest complain with UBNT has indeed been stock outs. Stock outs mean we have to accomplish the same project with different vendors. So I am happy UBNT is upping the safety stock even if it does show negatively on their balance sheets. Not every company can be JIT especially when all their products are made in China and shipped over on slow container ships.

As a Ubiquiti reseller as well I totally agree. I guess the only thing Ubiquiti is not good at is supply chain management.

UBNT recently opened up some Security Gateway XG | USG-XG-8 on the beta store for $999 and sold out in less than 10 minutes. See original tweet announcing it and subsequent tweet saying they were sold out already: https://twitter.com/ubnt/status/910614656388960256/

UBNT has a problem of undersupply and has been addressing it. Inventory levels increasing with expected increased sales should be expected and not a sign of fraud.

See https://twitter.com/ubnt/status/909446394267697152/ which shows a UniFi install in one of the largest hotels in Colorado. Plenty of demand out there.

Increasing DSO (Days sale outstanding) – Pera has explained this before in the last couple of earnings calls. This is due to moving to bigger distributors that can handle the increased sales and better distribute their product. Bigger distributors have less credit risk and can demand more generous payment terms. Increased DSO is to be expected when moving to larger distributors.

Fake community – Shorter Unemon (referenced in Citron’s short report) tweeted: @unemon1: “$UBNT … 3x Mikrotik Revenue + “EVANGELICAL” community … … yet … fewer people visiting http://UBNT.COM website? Hilarious!”

A knowledgeable UBNT investor posted on the Seeking Alpha boards the following response:

Look at the size difference (in terms of number of posts) in the communities:
UBNT has 2.7x the number of posts. And no, you won’t see them all without beta and alpha access.
The reddit UBNT community has 3.7x the number of subscribers:

Left asked how can there be an evangelical community over home wireless routers. UBNT has not even gone after the consumer market until recently with AmpliFi. Making this statement totally ignores the WISP market, which loves UBNT products. A wireless network engineer posted an article discussing Left’s report. It’s a good read that addresses the community concern. See here: https://medium.com/@Wirelessnerd/an-insiders-view-on-ubiquit…

Also, you can see that UBNT.com traffic has been increasing here: https://www.similarweb.com/website/ubnt.com#overview

Notice that UBNT.com traffic has been increasing dramatically from 5.7M visits in April to 7.9 visits in August. Note that the August number was before the short report and I would expect the number of visitors to be much higher now with investors researching the community to see if they are real.

For me, what gives me the most confidence that the company is not a fraud is from the following disclosures from the recently issued 10K, which states:

We have in the past and may in the future fail to maintain adequate internal controls. For example, as reported in the Annual Reports on Form 10-K for the years ended June 30, 2015 and 2016, management of the Company determined that the Company did not maintain an effective control environment, which contributed to three material weaknesses in internal control over financial reporting.

Regarding the lack of sufficient, competent personnel necessary for effective financial reporting, we recruited and transitioned the leadership team within the finance organization by adding additional qualified and experienced personnel. We have updated our accounting policies and procedures documentation together with key process level documents. In conjunction with the recruitment efforts, we have clarified roles and responsibilities within the finance organization.

Regarding the ineffectively designed and maintained controls required for safeguarding of the Company’s funds and timely detection of improper transactions in the general ledger, we have implemented new policies and controls surrounding disbursement authorities and journal entry creation, including training on required supporting documentation.

Regarding the ineffectively designed and maintained controls over user access and transaction privileges to modify and post entries to the general ledger and subsidiary ledgers, the Company has reviewed and updated user access and transaction privileges to the general ledger and interfacing subsidiary systems through the implementation of automated and manual controls. Additionally, we have performed a segregation of duties analysis over general ledger access. Management has also developed policies to ensure that future system changes consider impacts to the general ledger, including transaction privileges and segregation of duties.

We have completed our testing of the controls discussed above and conclude that our previously reported material weaknesses in our internal controls over financial reporting have been satisfactorily remediated as of June 30, 2017.

We all know that UBNT did not have the strongest financial controls. This was an example of Pera being too frugal and not spending enough on a good accounting team. The $46.7M wire fraud opened up Pera’s eyes to the importance of having good accountants and immediately hired FTI (http://www.fticonsulting.com) to help implement the right controls and build out a good accounting team. This was expensive, but led to a stronger financial control environment. The company hired Kevin Radigan as the Chief Accounting Officer and he took charge of the accounting and SOX controls. He was the CFO for Tunstall Americas from January 2012 to March 2016, so he has good experience and has resolved all the material weaknesses as they were all remediated in the 2017 SOX audit. KPMG assessed that all the material weaknesses were remediated. Any company that had previous material weaknesses will have their controls and account balances scrubbed very closely. The last thing the external auditor wants to do is have the PCAOB (part of the SEC) review their audit and find that it did not satisfactorily test the SOX controls. This is an area I know very well as started out in external audit for 2 years and have worked in internal audit the last 16 years. During these 16 years, I have worked with 3 of the Big 4 auditors closely to test SOX controls, so I am very familiar with their testing procedures. With SOX, the control testing has gotten more intense and the evidence of the controls need to be well documented. When SOX first came out in 2004, you could initially get away with just signing off on a report as evidence of review, but now the accounting staff and reviewers need to be able to explain the inputs to the reports used to perform a control, verify the accuracy and completeness of all the inputs (IPE), explain how the control operates, describe how the review process works and document evidence of review which goes beyond just signing off (evidence of following up with the preparer which includes email, meeting minutes, notations within the detail of the report, etc). The external auditor will then test the IPE, review the control to verify it was performed focusing in on evidence of review and reperform the control. When there is an area of risk (based on potential for a material weakness), the auditor will make the company implement multiple controls to give them comfort that there is no risk of a material misstatement. I work for a smaller company with a market cap of $500M and we have over 300 key controls that our external auditor will test in great detail. UBNT is a much larger company that paid their external auditor (PwC) $1.8M in fiscal 2015 and $2.1M in fiscal 2016. For a Big 4 auditor, a $200 per hour blended rate is typical, so for fiscal 2016, PwC would have incurred around 10,500 hours for the SOX and the financial audit. Also, the external auditor has full access to all the accounting records, Board minutes, accounting personnel, management, etc. For audit engagements, Big 4 firms don’t sell anything other than their seal of approval and they are audited every year by the PCAOB, so they need to make sure they audit extensively and document all their testing because they will be audited sooner or later and they don’t want too many black marks from the PCAOB. See https://pcaobus.org/Standards/Auditing/Pages/AS1215.aspx for the audit documentation standards that the auditors are held to. This isn’t the pre-SOX days where you could get away with not fully understanding the business and controls like what happened with Enron. In the SOX world, the PCAOB will review the external auditor’s workpapers to verify they fully documented according to the standards. Also, the PCAOB performs their audits based on risk, so you better believe that KPMG (the fiscal 2017 external auditor) was fully expecting a PCAOB audit and dotted every i and crossed every t when it came to the UBNT audit.

I am pretty sure that due to the material weaknesses found in previous years, PwC tried to jack up the audit fee for fiscal 2017 knowing that they would have to increase their audit hours in preparation of getting scrutinized by the PCAOB. This is likely the reason they switched to KPMG. KPMG isn’t cheap either so I assume that the 2017 audit fee will be at least as much is the 2016 $2.1M fee. We will find out the 2017 audit fee once UBNT files their proxy in December. Knowing that KPMG gave UBNT a clean audit and that the previous material misstatements were remediated in 2017, I have no doubt that there are no material misstatements or material fraud. I know this because I work closely with my company’s external auditor and my team performs SOX control testing in order for management to sign their 302 SOX certification (http://www.soxlaw.com/s302.htm). Pera and the CAO sign the 302 certification as well making them personally liable if a material fraud is later found out. So who do you believe: KPMG who had over 10,000 hours of trained auditors go over UBNT’s accounting records or Andrew Left who has no audit background and no access to company records in addition to having done very little research into UBNT (based on some of the easily refutable claims he made)? I’ll go with the external auditor charging $200 per hours with full access to company personnel and records. No external auditor wants to be the next Arthur Andersen and KPMG knows that their UBNT audit workpapers will likely be closely dissected by the PCAOB. Left does not need to prove anything. He just shorts the stock, posts a report that has many glaring holes, goes on CNBC crying fraud and makes millions with no external government agency scrutinizing his claims.

I am very comfortable with my UBNT investment. I know there is no material fraud or material accounting misstatement. I know the company has been aggressively buying back shares on the cheap and recently raised the lower end of their Q1 revenue guidance (from $230M - $250M to $240M - $250M). Assuming the company completes their buyback authorization at the current prices, their blended 2018 EPS will be around $4.14 and at 54, the forward PE is 13 and their earnings are growing nicely. In an expensive market, I’ll take that kind of deal any time. My advice is to tune out the noise (Citron) and hold your shares while the company proves that they can meet their guidance.


Thanks Wooter, for the great summary.

Pera’s engineering skills, drive, and vision are not in question.
As far as a stock, it’s a bit of a crapshoot. Obviously, Wallstreet is going to impede its progress, and the professional traders are trying to exploit market swings. So, what’s an investor to do? … Invest, but be nimble.

I re-entered with this Citron BS, via long calls, shares purchased, and short puts at a lower strike to exploit. Thank you, Mr. Left, for being such an AH. I hope that innocent shareholders were not effected.



1 Like

Thanks Wouter. That was a terrific and comprehensive write-up and must have taken a lot of work. We appreciate it very much.