- NQ Mobile is underinvested in expenses related to its product and customers.
- NQ Mobile is overinvested in ways that signal round tripping.
- NQ Mobile's auditor may have already viewed them as a high risk for fraud before even the Muddy Waters report.
Background and Highlights
*NQ Mobile (NYSE:NQ) paid nearly twice as much as similarly sized companies for its audit last year and had the third highest audit fee out of nearly 100 Chinese companies sampled. The only two companies that paid more had at least eight times the revenue of NQ Mobile. The high audit fees charged to NQ Mobile by Pricewaterhousecooper (NYSEARCA:PWC) are possibly because they already viewed them as a high risk of fraud even before the Muddy Waters report was published that provided much stronger evidence.
*Based on data from 97 software companies spanning 700 years, NQ Mobile looks exactly like what one would expect a round tripping business to appear.
*NQ Mobile is chronically underinvested in expenses involving its product and its customers (sales and marketing and product development expense, and computer equipment) relative to its revenue and revenue growth.
*NQ Mobile is chronically overinvested in expenses that involve generating and moving cash off its books (payments to third-parties); expenses that limit the actual cash they burn through on operating expenses (stock-based compensation); and activities dealing with accounting, investors, and their stock (general and administrative expense).
*Each of the 97 software companies surveyed shows its revenue grows only as it continually invests in employees, sales and marketing, product development, and computer equipment. NQ Mobile's revenue does not. In an updated and more comprehensive model that ignores $225 million in revenue, one can see that NQ Mobile's revenue grows as its payments to third-parties increases.
*This more comprehensive model assumes that the difference between the model's estimated cash balance and NQ Mobile's reported cash balance each quarter is the company's actual revenue. By combining the actual revenue with payments to third-parties, the model predicts that NQ Mobile's reported revenue would be $244 million since it has gone public. NQ Mobile's reported revenue during this time was really $225 million.
Conclusion: NQ Mobile's relationship with its auditor was already questionable even before strong evidence was provided in the Muddy Waters report that they were round tripping their revenue through Yidatong. In 2013, NQ Mobile's financials became even more egregious, and their CFO lined up a new job and resigned months before the annual report would be due. This story is similar to the one that played out with Ambow Education (NYSE:AMBO), which is the last PwC client that paid much higher audit fees than other companies. They were eventually delisted after PwC resigned as their auditor in March of 2013 because they were accused of round tripping cash paid in acquisitions nine months earlier.
Prepare for Blastoff
The fundamental basis of all businesses involves making repeatable investments that generate or support a greater return than the money the business invests. The types of investments may be different from business to business. How the return is generated may be different. At the end of the day, however, all businesses must follow this simple and universal rule.
In my last report, I gave a few examples:
*Chipotle's (NYSE:CMG) growth in revenue moved in tandem with its investments in leases for new restaurants to be in front of more hungry consumers.
*Groupon's (NASDAQ:GRPN) investments in employees was consistent with their revenue growth even as it grew from $94,000 to $2.3 billion in five years, a ridiculous 2.4 million percent.
*Even Microsoft (NASDAQ:MSFT) made repeatable investments in employees and computers as they monopolized the PC business in the 1990s.
I even included in the data set the example of Amazon (NASDAQ:AMZN) from 1995 to 1999. The reason why I included Amazon in the data set is not because I think their business model is like NQ Mobile's. I brought it up because it is an example of a company that at least superficially went through a massive transformation over a five-year period that still followed this rule.
In 1995 Amazon was a five-person company working out of Jeff Bezos' garage and generating $500,000 in revenue, and by 1999 they employed over 7,600 people and generated $1.6 billion in revenue.
Despite undergoing such tremendous growth, Amazon exhibited the same pattern that all small businesses that become large ones do. They created a product. They found that it met the needs of people. They suspected more people also needed their product. To capture that market they made similar, but much larger investments as before.
In Amazon's case, it turned out that the market was quite massive and their team was aggressive about capturing it. As a result, Amazon grew its revenue an astounding 320,000% from 1995 to 1999, which made them one of the fastest growing companies ever at the time.
In spite of all of this, however, Amazon's return on investment changed very little over these five years. For example, Amazon's revenue per computer increased by only 25% and their revenue per employee by only 111% from 1995 to 1999. That's because Amazon depended on computers and employees to generate each dollar of revenue as much in 1999 as they did in 1995.
Actually this is true of most software companies. After doing research on nearly 100 different public companies spanning more than 700 years collectively, it is plain to see that the main investments software businesses are making as they grow their revenue is in computer equipment and employees. No matter how fast their growth is, software companies are forced to make repeated investments in computers and employees because this is actually what is driving and supporting their revenue growth.
This is all very different from NQ Mobile, of course. Whereas Amazon is showing repeatable investments in computers and employees that scales with its revenue, NQ Mobile from 2008 to 2012 grew its revenue per employee and revenue per computer at 12 times the pace of Amazon, despite having just one-tenth the growth rate of Amazon. In fact, it has one of the highest revenue per employee and computer growth rates of any software company ever.
One of NQ Mobile's most diehard investors criticized comparing more than 500 data points across the software industry because the companies are not exactly alike, citing Amazon in particular as a bad comparison.
It's difficult to compare a Netflix or Amazon to NQ Mobile for a multitude of reasons (subsector, stage of growth, size/scale, location, go to market strategy, competitive advantages/weaknesses, prodigious use of operating leases vs capital leases, etc.)
-Trade Star, "Round Tripping: Further Adventures Down the Muddy Waters Rabbit Hole"
Meanwhile comparing Amazon to hundreds of other data points from other software companies yields this graph1:
The reason why Amazon resembles internet security company McAfee, Chinese companies Baidu (NASDAQ:BIDU), Tencent, NetEase (NASDAQ:NTES) and the hundreds of other data points despite not being exactly like them is because you have to invest to grow. The importance of this principle trumps any differences across business models, geographies, or eras.
The Diehards' Guide to Comparing NQ Mobile
Purpose of the comparison
What are good comparisons
Figuring out what the company is worth
Figuring out if the company is exaggerating its revenue
The-sub-$50 million-in revenue-Chinese-mobile-only-security-that-prodigiously-use-operating-leases-highly-leveragable-platforms software industry
New or Improved Products and Sales & Marketing Are What Drives Revenue
How software companies allocate their investments in operating expense to drive their revenue growth also tends to be the same.
Here are some examples:
*Despite acquiring users for free virally and growing its revenue nearly 2,500%, Facebook's return on investment in sales and marketing increased only 85% from 2007 to 2011.
*From 2008 to 2012, Zynga (NASDAQ:ZNGA) grew its revenue from $19 million to $1.2 billion, a whopping 6,500%, but its return on investment on product development expense grew by just 24% during this time.
*In 1982 Microsoft generated just $24 million in revenue and by 1999 they generated $23 billion, or nearly 1,000 times as much as they did in 1982. Despite distributing its products largely through contracts with original equipment manufacturers, the most Microsoft sales and marketing return on investment changed from 1982 to 1999 was just 113%.
*The same is true of McAfee, the internet security company whose products were preinstalled with original equipment manufacturers, distributed by internet service providers, and sold through retailers. From 1995 to 2010, McAfee's revenue grew from $90 million to $2 billion, and yet its return on investment on its sales and marketing increased by less than 5 percent over these 15 years.
Even some companies that claim to obtain its users from viral marketing still show a strong, consistent relationship between its revenue growth and its sales and marketing expense. Take the example of AVG (NYSE:AVG), the consumer security company that, like NQ Mobile, credits viral marketing in its annual report for acquiring many of its users. As you can see below, even as AVG's revenue tripled from 2008 to 2012, their return on investment on sales and marketing expense was always roughly the same.
Looking at the graph above, one cannot help but ask: why does a company that acquires many users for free virally show that consistent investments in sales and marketing are driving its revenue?
Actually what AVG describes as viral marketing is not the same as the viral user acquisition that occurs with Facebook or Zynga, for example. Here is what AVG says about it:
The primary driver of growth of our active user base has been increasing user awareness … of the AVG brand, which we promote through online viral marketing, including word-of-mouth advertising, blogs and other media referrals.
If you notice, AVG subtly explains that it is actively pushing its brand's awareness by promoting itself through "word-of-mouth advertising, blogs, and other media referrals." In other words, AVG must directly promote its brand for a quite obvious reason - internet security products are not viral. That's because unlike Facebook, Zynga, and other viral products, security users do not need to invite other users in order to enjoy the product at its fullest, which is at the heart of what makes products viral.
What AVG does - "word-of-mouth advertising" - is simply a more effective type of marketing that involves, among many things, influencing the opinion of people by creating conversations about the product online and forming communities for potential customers to find information about it. In other words, what AVG does is much more like the marketing that appears is being done to push NQ Mobile's stock, not its security apps.
This type of marketing is natural, and it works. However, marketers are behind the scenes pushing it, which means it has a recurring cost, and the chart above shows that the cost has been roughly the same for each dollar of revenue AVG generated for the past five years.
Virality is Built Into the Products, It Does Not Happen Out of Thin Air
Number of monthly users
What it is
Why it was viral
Take and share photos with your friends on social networks
People took photos and then shared them with friends on social networks
Play games online with your friends
People invited their friends to play a game online
Share photos with friends online
People shared photos with friends online
75 million (at its peak)
Connect with friends online
People invited friends to connect online
Collect and share projects and other interests with friends online
People collected and shared projects and other interests with friends online
RenRen (the Facebook of China)
Connect with friends online
People invited friends to connect online
Share your location with friends on social networks
People shared their location with their friends on social networks
Create and share short videos
People created and shared short videos with friends
Protect phone against viruses
NQ Mobile says so*
Take and share videos and photos with friends
People took and shared videos and photos with friends
Listen to and share the music you are listening to with friends
People listened to and shared the music they were listening to with friends on social networks
*NQ Mobile claims 30% of its security users, which would mean 36 million of its active users, are acquired from referrals from existing users, and yet there is nothing about their product that would make people share it with new users, very few people in their largest market has heard of it, and 36 million users would put them on the short-list of the most viral apps around today.
Throughout time even the greatest software companies with the greatest products and greatest user acquisition strategies had to invest in sales and marketing, product development, or both to grow their revenue. In the same way that products are not "viral" without a very obvious reason behind it, a company's revenue does not grow without obvious actions being taken by the company that support it.
None of this is true for NQ Mobile, of course. Since 2009, NQ Mobile has grown its return on investment on sales and marketing by more than Zynga (350%) and their return on investment in product development very nearly more than Facebook (397%). This is in spite of the fact that most of Zynga's new users came largely from invitations from existing users, and Facebook was one of the most popular Internet products basically from the moment it launched.
I guess now it's nice to remind everyone that we are talking about a company that hardly anyone has heard of before in its largest market. I was in China in late December and January on a trip with my wife. Despite the fact they supposedly have 50% of the monthly active users of Twitter (NYSE:TWTR), not a single person I spoke to had ever heard of NQ Mobile or Netqin or had it on their smartphones. It was also nowhere to be found in any stores I went into or walked by.2
NQ Mobile has next to no brand recognition or visibility, no obvious reason why people would share its app with other users, and claims to have hundreds of millions of users. Yet its return on investment in sales and marketing is one of the highest of any software company, and its growth rivals that of some of the most viral applications ever.
The graph below shows NQ Mobile's sales and marketing and product development return on investment growth relative to hundreds of other data points from software companies. It includes 14 data points from security companies AVG, Qihoo, and McAfee, 21 data points of companies that had less than $20 million in revenue, and 67 data points from other Chinese software companies. It excludes companies that were overinvested at the beginning of the data set (less than $1 return on investment) or whose economics changed because of mergers/acquisitions (significantly more than NQ Mobile).
Yet again, NQ Mobile is an extreme outlier.
Why is it that whenever data is compiled comparing investments in its product and customers to revenue this seems to happen?
Is it because of NQ Mobile's unique go-to-market strategies, the fact their revenue is partly international and domestic, the "prodigious use of operating leases," or whatever creative reason the diehards can generate in the face of overwhelming evidence that NQ Mobile is exaggerating its revenue?
Of course not.
As I explained in my earlier report, the one factor that distorts the data is whether the company was overinvested at the beginning of the data set. I mentioned several companies during the Dot-Com Bubble of the late 1990s that fit into this category.
NQ Mobile's investment levels in 2008 and 2009 were above average relative to their revenue, and certainly this impacts NQ Mobile's growth rates.
However, as you can see above, a big reason why NQ Mobile is the most extreme outlier when one compares its investments to its revenue growth against other software companies is because their return on investment is an extreme outlier.3
The truth is NQ Mobile's supporters are right. Comparing NQ Mobile to 700 data points from software companies is an apples to oranges comparison. The reason why that is true, however, is because NQ Mobile does not appear to be a software company.
Round tripping involves using a company's own cash to book fake revenue. The way this is done is by moving the cash off a business's balance sheet through fake expenses paid to businesses that are actually cooperating with them. Once the collaborator receives the cash, they hand it off to a fake customer that "purchases" the company's products or services with it. The more times this cycle is completed, the greater the revenue the company is able to report to its investors. The greater the revenue that is reported, the greater the value of the shares that they own.
One of the largest challenges a round tripper faces is finding credible expenses to move a lot of cash out of the company without the fake expenses being so egregious that the auditor notices. One way to make some of the expenses appear as if they are not egregious is to repay them.
For example, shortly before NQ Mobile's IPO, Yidatong repaid its $2.1 million advance. One may recall that this was cited specifically by Toro Investment Partners as a key reason why the short thesis was faulty in their response to the Muddy Waters Report. What this ignores, however, is that a few months prior NQ Mobile acquired a stake in Feiliu for $2.5 million in cash, part of which could easily have been diverted to Yidatong as its loan repayment. As the repayment of the Yidatong advance shows, this appears much like a ponzi scheme where each new expense inevitably replaces an old one even if it is repaid.
Therefore, as time goes on and NQ Mobile's revenue grows, one would expect the round tripping to become more obvious in their financial statements as more expense and asset categories get inflated since these are the sources of the companies' growing revenue and expense repayments.
I believe it is quite obvious that this is exactly what is happening with NQ Mobile. Let's take a look at how.
For most companies prepaid expense refers to minor corporate expenses such as insurance that is fully paid at the beginning of year, but is expensed throughout the entire year. For NQ Mobile, prepaid expense refers to minor expenses such as these too, but also highly questionable expenses such as advances to employees and interest-free bridge loans in connection with investments.
As prepaid expense is not a core business investment, the average amount of it showing on the balance sheet of the nearly 100 software companies surveyed is just 4% of the revenue they receive. NQ Mobile, on the other hand, shows five times the investment in prepaid expense relative to its revenue (20%) and has the highest such percentage of any company in 2012.
NQ Mobile's prepaid expense is so large, in fact, that it is higher than both its annual sales and marketing and research and development expense. Common sense tells us this is totally backwards and the data backs it up as well. Out of 510 data points from other software companies, this is one of just four instances when this has occurred (less than 1% of the total data set).4
Likewise only 5% of the companies in the data show greater prepaid expense than computer equipment, and the average company has about 50% less prepaid expense than computer equipment. Just one company (Marketo) has ever had more than three times as much prepaid expense as equipment, and this only occurred one year before they returned to the mean. NQ Mobile is on track to have more than six times as much prepaid expense as computer equipment for four straight years.
The bottom line is NQ Mobile has been consistently underinvested in computer equipment, sales & marketing, and product development, which does not make sense if they were a legitimate software company, but has been dramatically overinvested in prepaid expense, an asset category that is totally irrelevant for most businesses but makes a ton of sense to be overinvested in if your core business is actually round tripping.
Equity Investment in Associates & Acquisitions
Like prepaid expense, equity investments in associates is another asset category that jumps out in NQ Mobile's balance sheet. Equity investments in associates refers to investments NQ Mobile has made in private companies where they do not have a majority ownership, but they do yield some control, usually in the form of a board position. For most companies this is not a very important investment they are making, and in fact, out of more than 700 data points, only 26% reported any such investments in their annual report.5
Of the ones that do, the investment amounts are small and equal just 8% of the company's revenue on average. NQ Mobile, of course, has invested $41 million into associated companies, which is 22% of its projected revenue for 2013. NQ Mobile is on pace to have the third highest investment amount into private associates out of any software company in 2013 as a percentage of its revenue, even without considering any additional investments occurring in the fourth quarter (based on 2012's data).
Looking at NQ Mobile's balance sheet, one would deduce that they are a venture capitalist and not a software applications company. Out of 160 data points, only one other time has a company had higher investments in private associated companies than both its computer equipment and its combined sales and marketing and product development expense. That is less than half of one-percent of the data, and remember, 75% of the companies in the data set reported never having made any such investment before.
*There are actually 160 data points in this chart above. It looks like there are less because most of the data is near 20% or less and NQ Mobile in 2013 and another data point throws off the graph.
Out of more than 70 American companies in the data set, only Amazon, Microsoft, and Google (NASDAQ:GOOG) have exceeded the $50 million NQ Mobile is on track to have invested in private companies. Oh you know, just a few companies with at least 300 times the revenue of NQ Mobile.
The same story is also true of the acquisitions NQ Mobile has done. Based on data from roughly 500 acquisitions over 269 combined years as public companies, NQ Mobile is one of just five instances that has invested more than 40% of its revenue towards acquisitions in consecutive years. In 2013, NQ Mobile's acquisition total is already the 14th highest in the data set as a percentage of revenue. This is in spite of fact that it is based on three quarters of data.
Not only are NQ Mobile's acquisitions in 2012 and 2013 as a percentage of their revenue one of the highest over a two-year period, but they have used less cash and more of their stock while doing it.
In the data set, cash makes up 79% of a company's annual acquisition cost on average. NQ Mobile in both 2012 and 2013 paid roughly 11% of its total acquisition cost in cash and paid the other 89% using the company's stock. There are only eighteen examples in the data where a company paid for less than 15% of its total annual acquisition cost with cash (7% of the data). NQ Mobile has done this for two consecutive years.
Remember this was a company with zero debt and low capital expenditure that even before its convertible bond offer had 3 years of cash operating expense available in cash and term deposits, making them one of the most cash-rich software companies. Yet they consistently chose to disproportionately use stock instead of cash to pay for their acquisitions.6
Not only does NQ Mobile consistently acquire more companies using stock despite sitting on a mountain of cash relative to its cash operating expense, but the one acquisition that would have been a total no-brainer to make in 2013 if they were definitely legitimate - Yidatong - they refuse to do.
Why won't they do it? Probably because then Yidatong would be included in their audit.
Consider the lengths they are going to defend this relationship.
Cash spent defending Yidatong relationship
$3 million share repurchase
$2.9 million in revenue
$3 million "independent" investigation (est. based on last year's audit cost plus legal fees)
Less $1.7 million contributed by NQ Mobile (60%)
$6 million total
$1.2 million revenue excluding revenue portion that are expenses for NQ Mobile
This does not even include how the Yidatong relationship holds back the value of their shares, is a major reason why NQ Mobile pays some of the highest auditing fees in China, and how the "independent" investigation could easily end up costing much more than $3 million (Sino-Forest paid $50 million in theirs, for example). We know that NQ Mobile's contract with Yidatong expires in 2015 and they make up the majority of their revenue. Looking at the chart above, do you think NQ Mobile has the leverage to get this deal done at a price that makes sense for everyone involved?
At some point, faith must be placed in basic and obvious concepts such as the idea that people act in their rational self-interest in the marketplace. What NQ Mobile is telling us is that they think it is in their rational self-interest to spend a lot of money to defend their relationship with Yidatong but not to pay the possibly lesser amount to acquire them.
Why do you think this is?
It certainly is not because they are afraid of acquiring companies. They also have a strange habit where they prepay for five times as many expenses as the average company. Yet they pass on the ultimate NQ Mobile transaction - an all-stock acquisition of Yidatong to prepay their payment processing expenses.
Purchased Computer Software
Another expense that sticks out as possibly being round tripped is NQ Mobile's purchased software in 2012. In the 227 data points where purchased software and computer hardware are separately listed in a company's financial statements, 96% show that the value of their computer hardware is greater. In fact, computer hardware is three times greater than software on average.7
NQ Mobile, of course, is one of the four percent of the data points that shows greater purchased software than hardware on its balance sheet. This alone puts them opposite 96% of the companies in the data set, but how NQ Mobile's software was acquired also makes them an anomaly. From 2008 to 2011 NQ Mobile showed a fairly consistent, low level of investment in software, and then suddenly in 2012 decided that purchased software was one of the most important investments they could make, acquiring nearly 9 times as much software as they owned in 2011 while spending very little on hardware.8
In fact, the change in the ratio of computer software NQ Mobile owned relative to hardware was the second largest change out of 180 data points.
So not only is it unusual that NQ Mobile owns more purchased computer software than hardware, but the rapid way this occurred is also highly unusual. To put it simply, this expense is highly improbable, and much like the company's acquisition of NQ.com, it appears like another intangible asset that was possibly acquired specifically because it is difficult for an auditor to determine the fair value.
It is also part of a consistent pattern where NQ Mobile decides that something that is for the most part is immaterial to other software companies is suddenly a more important investment than computer equipment. In 2012, purchased software suddenly became the more important investment. In 2011, it was their domain name, NQ.com, and from 2008 to 2010, the most important investment was in their payment processor.
Purchases of Equipment & Intangibles
More important investment
$720,000 (excludes domain)
Licensing domain name, NQ.com
Keep in mind that although Yidatong "repaid" NQ Mobile (as mentioned earlier, I suspect the repayment was round tripped from money paid for the Feiliu "acquisition"), the cash software companies put into buying computers is also repaid when the investment returns as revenue. In fact, the roughly 100 software companies surveyed returned 11 times the cash they paid in computer equipment in revenue per year. If computer equipment becomes useless after three years that means they generate $33 for each dollar they invest in computers.
The only software company that seems to treat investments in computers as "lost money" is NQ Mobile. Instead they think a better use of their capital is to lend $4.8 million at zero percent interest to their payment processor.
Let's remember that NQ Mobile claims they lent this to Yidatong because payment processing is not a core function for them. Yet the $4.8 million is greater than the total amount of computer equipment and software NQ Mobile purchased in the last 5 years.
In light of this, let's use our common sense for a moment. What is more important to NQ Mobile's business - payment processing or creating and maintaining software applications?
Advances to Yidatong
Purchases of property, equipment, intangible assets (excluding domain)
Putting It All Together
So let's run it all back.
Where software companies are invested
Where software companies are not invested
Equity investments in associates
Purchased intangible assets
Sales and marketing
Not very important to NQ Mobile
Very important to NQ Mobile
The reason why prepaid expense, equity investment in associates, purchased intangible assets, and all-stock acquisitions are of incredible importance to NQ Mobile but not other software companies is because these appear to be the investments that directly generate revenue for NQ Mobile through round tripping.
To show this, consider the chart below. This chart shows the change in cash revenue NQ Mobile has received (excludes accounts receivable, but includes deferred revenue) relative to the change in NQ Mobile's prepaid expense, equity investments, purchased intangibles, and acquisitions. The change in cash revenue in each quarter is relative to the cash revenue it generated in its first quarter as a public company (second quarter 2011). During this time, NQ Mobile's cash revenue increased from $7.5 million to $64 million.
The chart above shows that as time goes on and NQ Mobile's revenue has grown dramatically, the investments they are making in prepaid expense, equity investments, purchased intangibles, and stock in acquisitions is typically very near to the amount that their revenue increased. As NQ Mobile's revenue grows and gets further from reality, they must also constantly introduce new investments to fuel their revenue growth. In 2011, prepaying for expenses was of incredible importance to the company. In 2012, the new most important investment was in associated companies, and in 2013 NQ Mobile began heavily acquiring private companies using their stock.
If they are not round tripping, it is quite the coincidence that all the areas they are overinvested in relative to other software companies have this kind of direct relationship to their revenue growth.
Let me show you what I mean. I created a model that completely ignores $225 million in cash revenue that NQ Mobile has generated as a public company, and instead assumes that the cash NQ Mobile pays to third-parties are the source of its revenue. The cash NQ Mobile pays to third-parties is usually acquisitions, equity investments, prepaid expenses, fees paid to payment processors and other cost of revenues and intangible assets.
This model is much more comprehensive than the one from my previous report and includes very nearly every transaction that has occurred in the company's income statement and cash flow statement for the past 10 quarters. I also updated my previous model to correct for some mistakes, such as:
- Not including costs that are probably real such as issuance costs for raising capital and inventory
- Not including a sales tax, which is 3 to 5% in NQ Mobile's case according to their annual report (not 20% like one author alleged)
- Not adjusting for some non-cash expenses such as accounts payable
- Double-counting foreign exchange
- Using an inaccurate performance earn-out vesting schedule (The model never intended to assume NQ Mobile used unvested shares to generate cash. The vesting schedule was simply inaccurate because I based it on a press release, not the actual contracts.)
In spite of the fact that the model ignores $225 million in cash revenue, it is typically within 90% of predicting NQ Mobile's ending cash balance for each quarter. If I assume that the difference between the estimated cash balance and the reported one is NQ Mobile's actual revenue, then I can combine it with NQ Mobile's payments to third-parties in order to predict NQ Mobile's cash revenue since they have gone public within 87% each quarter. The model predicts NQ Mobile's revenue since they have been public was $244 million, while NQ Mobile's actual reported cash revenue during this time was $225 million.
If the model is accurate, then NQ Mobile's security software revenue since they have gone public is actually $90 million, or 42% of the $211 million they reported (everything other than Nationsky). You can view the model yourself here. (Also note: what I am doing here is why Dropbox is another viral product).
I have no illusion that the model is precisely how NQ Mobile is round tripping. However, there is little doubt given the consistency of the model that the company's largest purchases such as equity investments, bridge loans, cash paid in acquisitions and purchased intangible assets correlate to the company's revenue for ten consecutive quarters. This kind of a direct relationship to a company's revenue should be expected of cost of revenues ($65 million of estimated revenues), but not for equity investments, prepaid expenses, bridge loans, purchased intangible assets, and the like ($95 million of estimated revenue, excluding stock in acquisitions).
Given that these are the exact expenses NQ Mobile is grossly overinvested in relative to 97 software companies, that most of NQ Mobile's users cannot be located in surveys, that NQ Mobile's revenue growth appears exaggerated relative to its investments, and their relationship with Yidatong has been questioned because of precisely the risk of round tripping, this is a really big deal.
In 2013 the model continues to correlate to these expenses and also scales along with stock awarded to Chinese parties. If I do not include cash generated from stock awarded in acquisitions, the model only begins to deteriorate in the first quarter of 2013, which is probably when the first shares vested from NQ Mobile's acquisition spree.
As I argued in my last report, I believe that NQ Mobile is generating much of this missing cash by using stock awarded in these acquisitions either through a margin loan or by selling the stock. Contrary to what one author alleged, doing this would not be discovered by their auditor in "annual impairment tests," American executives would not need to be complicit and neither would the company's partners or vendors.
This same author criticized the notion of using stock to round trip revenue in this way, arguing that it was "terribly inefficient and costly."
"As it stands, using stock to round-trip is a terribly inefficient and costly way for the company to defraud investors."
-Trade Star, "Round Tripping: Further Adventures Down the Muddy Waters Rabbit Hole"
You'll have to excuse him, everyone. He has been spending too much time looking at the fake business model showing on NQ Mobile's financial statements where you do not have to invest any capital to grow your revenue endlessly.
In the proposed round tripping business model, I assume they invested roughly $30 million out of the $116 million awarded in stock-acquisitions in the last 12 months, which increased their revenue and contributed to increasing the value of the founders' shares by $234 million. The decrease in days sales outstanding, increase in cash revenue, and increase in their cash position also helped to convince people they were generating cash like a regular business so there was actually demand for their $170 million convertible bond. This bond is legally the founders' thanks to the fact the company's VIE structure cannot be enforced in China, and even the $30 million is ultimately still the founders because it gets processed as revenue into the company bank account, which they control. This $30 million processed into the account would also offset the $30 million increase in NQ Mobile's cash operating expense relative to the first three quarters of 2012, helping the founders preserve the capital raised in the IPO.
Only in the world of a diehard NQ Mobile supporter is earning up to $404 million by depositing $30 million that is not even yours into a bank account that you legally control totally unacceptable. Why is it unacceptable? Because it involves giving people that helped a cut, paying taxes and possibly some interest, and doing some extra legwork beyond what simply stealing the $30 million outright would require.
The diehards, ladies and gentlemen.
General & Administrative Expense
The previous section dealt with the expenses that NQ Mobile is overinvested in that appear to be sources of round tripped revenue. There are other expense categories that NQ Mobile is overinvested in that may not be sources of round tripped revenue, but still speaks volumes about their business model.
Take general and administrative expense, for example.
Imagine looking at a basketball box score and seeing that the head coach, or the person operating in the background to support the activities of the players, was actually the team's leading scorer.
That is a lot like looking at NQ Mobile's general and administrative expense.
For most software companies, general and administrative expenses are not more important investments than the operating expenses that deal directly with their customers - sales and marketing and product development. To understand this, consider the chart below, which plots the difference in the percentage general and administration makes up of a company's operating expense versus sales and marketing and product development.
As you can see above, there are hundreds of data points from companies located in the third quadrant (lower left).10 Companies in the third quadrant (lower left) show that both their product development and sales and marketing expense is greater than their general and administrative expense. In fact, a software company is 14 times more likely to have higher sales and marketing expense than general and administrative and two and half times as likely to have higher product development expense.
To put it simply, sales and marketing and product development are what matters for most software companies. That's because creating a product people need (product development) and then pushing it to customers (sales and marketing) is another fundamental basis of business.
NQ Mobile, of course, defies the odds and is part of the less than 2% of instances where a software company shows higher general and administrative expense in its income statement than both its sales and marketing and product development expense. This is true even without stock-based compensation included. In fact, NQ Mobile is the most extreme outlier not once, but twice, and very nearly three times, and this expense category has gotten so large that today it is now nearly twice as large as product development and sales and marketing combined.
Not only is general and administrative much greater than these other operating expenses, but unlike sales and marketing and product development expense, general and administrative appears to be the recurring investment that grows with NQ Mobile's revenue. As you can see below, NQ Mobile's revenue has consistently been between two to three times its general and administration expense, while its return on investment on sales and marketing and product development has more than quadrupled over the past six years.
For basically every software company other than NQ Mobile, general and administrative expense has a symbiotic relationship with sales and marketing and product development expense, not a completely independent one.
This makes sense for software companies. As they increase their investments in sales and marketing and product development, they will likely have higher general overhead costs. As their revenue grows with their investments, the scope of their business increases, which requires more managers, accountants, lawyers, and people dealing with an increasing number of investors.
The data supports this fact as well. As you can see below, the fact that NQ Mobile continually invests in general and administrative as its revenue grows, but not product development and sales and marketing once again makes them the most extreme outlier.11
Before I continue, let me pause to address the concerns that the diehards will bombard the comment section with.
1. It is not just a function NQ Mobile's high stock-based compensation.
Below is a chart that excludes stock-based compensation from the operating expense categories.
2. It is not because NQ Mobile hired executives in 2012 and 2013.
Here is the same chart excluding the years 2012 and 2013.
3. It is not because NQ Mobile IPO'd in 2011. Here is the same chart excluding IPO years.
So why does NQ Mobile's general and administrative expense seem like it is the recurring investment that drives revenue for the company? Probably because it is.
Here's a list of the general and administrative expenses that appear in this category that matter much more to round tripping companies than legitimate software companies:
- Accounting salaries and expenses - round tripped transactions are much more deliberate and labor intensive and require more employees as the number of transactions increase.
- Salaries of employees that interface with investors - as the amount of revenue that is round tripped increases so does scrutiny, and round tripping companies would need more and better executives to deal with it.
- SEC reporting related expenses - auditors view suspected round trippers as a higher risk and charge them more than other companies; round trippers are willing to pay any price to their auditor because it is the main thing that validates their business (independent research and their own financials do not).
- Lawyers and legal fees - round tripping companies try to access capital markets and acquire more companies than regular companies because they lack the cash flow they claim, which increases their legal fees.
Knowing that these are the expenses in general and administrative, are you surprised to know that no other software company consistently earns less for each dollar it invests in general and administrative than NQ Mobile, and none show that it is a consistent investment that grows with its revenue completely independent of both sales and marketing and product development?
All of this makes complete sense when you remind yourself that NQ Mobile's business model appears to be centered on creating and maintaining the expectations of investors, not software applications. When everything is based on creating and maintaining expectations, expenses related to your financials, your stock, and the employees that interface with investors are where you push your investments because these have a much more direct impact on the value you create than expenses related to actual customers (computer equipment, sales and marketing, and product development).
Similarly, NQ Mobile's stock-based compensation signals they are a round tripping business. One of the main problems of being a round tripper is that although you show high revenue levels, you are not generating much new cash, and therefore, you cannot continue operating without burning through your capital on expenses that you must pay in order to conduct your round tripping (general and administrative) or expenses that are necessary to maintain the façade that you are legitimate (i.e. payroll for other employees). Burning through your capital is not good because that alerts investors to the fact you are not really generating any cash, and in NQ Mobile's case, the cash held in China is literally the founders' because of their legal structure (variable interest entity), which means they especially would want to protect it.
For NQ Mobile, the main way to mitigate this problem appears to be to pay for much of their operating expense with stock. Through three quarters in 2013, nearly 49% of NQ Mobile's operating expense was stock-based compensation. Since they have gone public, stock-based compensation has averaged 38% of their total operating expense. This is twice as much as the next highest company from nearly 40 software companies in their 2011 IPO cohort and nearly four times the average.12
It is even more ridiculous when you incorporate the data from the hundreds of data points from the 97 software companies in the data set. For example, there are only eight instances out of more than 500 data points (1.5%) where a company's stock-based compensation was more than 20% of their operating expense when they were publicly listed. In every case but one, their stock-based compensation declined the next year. NQ Mobile's stock-based compensation has been above 38% of its operating expense for the past two years as a public company (2011 and 2012) and has managed to climb to 49% of its operating expense in 2013. This is the highest from hundreds of data points and more than twice the next highest percentage when you exclude companies in the year they IPO'd.
In the same way that NQ Mobile must continue to make its expense categories more inflated as time goes on in order to grow its revenue, they must also pay for more expenses with stock as they increase their operating expense. The never-ending expansion in NQ Mobile's financials is the inherent nature of their business model. In a lot of ways it reminds me of a supernova that constantly expands until the weight caused by its expansion causes it to collapse into a black star. In NQ Mobile's case, the exaggerated nature of their financials is ultimately a weight that will cause them to collapse once it is time for gravity to rule the day.
The Auditor - AKA Gravity
Round tripping is difficult for auditors to detect because they do not have access to the records of the cooperating vendors and payment processors, and therefore, as long as the documentation (i.e. invoices) matches the bank records of the audited company and the expenses are not egregious, the auditor is in the unenviable position of making a judgment call on whether the transactions are real or fake.
In NQ Mobile's case, there are strong indications from 2012 that Pricewaterhousecooper already had doubts about their financials. For starters, out of nearly 100 companies sampled, NQ Mobile not only paid more for its 2012 audit than any other company in China other than two companies, but its audit fee nearly doubled and was nearly $1 million more expensive from the year prior. The two companies that have higher audit fees than NQ Mobile had more than eight times the revenue of NQ Mobile, which is the most obvious driver of auditing fees.
To understand this, consider the chart below, which details the auditing fees by revenue of Chinese companies sampled with revenues less than $300 million.13
*Ambow Education's earnings were restated in 2011. PwC resigned a month before their 2012 annual report was due.
PwC itself also cites risk as an important driver of pricing for audits in its report, "Considering an IPO." Here is a chart of just the PwC clients and their fees for an audit.14
*Sohu had an obvious reason why its audit fees were higher. Ambow did as well. Why does NQ Mobile pay twice as much for its audit as similarly sized companies?
Under the audit risk standard mandated by the Public Company Accounting Oversight Board (PCAOB), auditors are required to gather more evidence, ultimately charging more in fees, for companies that they perceive to be a higher risk of fraud.
Higher risk to auditors usually comes down to weaker corporate governance, lower quality and competence of management, and a poor attitude toward fair and transparent financial reporting. Knowing that Chinese executives will not go jail for defrauding foreign investors, is one surprised to know that the average audit fees of Chinese companies with revenues of less than $100 million is $522,000, while their American counterparts pay just 35% of that amount (see the aforementioned PwC report)?
Of course you're not surprised. Chinese companies, particularly smaller ones with fewer controls, are a much higher risk for auditors, and therefore they are charged substantially more. Basic stuff, right?
The Diehards' Guide to Important Questions about NQ Mobile
The Obvious Answer
The Diehard Answer
Why does NQ Mobile pay such high audit fees relative to its revenue?
PwC perceives NQ Mobile as a higher risk.
NQ Mobile's billing and operations are far more complex and decentralized than other Chinese companies.
See comments under my first article.
Why do surveys show NQ Mobile has a fraction of the users they claim?
NQ Mobile is exaggerating its number of users.
Access to and willingness of people to take a survey in NQ Mobile's main geographies "falls off a cliff" in relation to people in other parts of China.
Why was there a crash in NQ Mobile's stock last Friday?
A lot of shares were sold.
A trader accidentally sold NQ Mobile when he meant to sell the Nasdaq e-mini 100.
Knowing that risk is one of the most important drivers of audit fees other than revenue, it should be no surprise that one of the outliers in the chart above, Ambow Education, was eventually delisted when its audit fee not only was the second highest in China for 2011 but doubled from 2010.15
Despite the fact that its annual report was filed late with accounting restatements, Ambow still had longs willing to ignore it all. In a Barron's article from June 23, 2012, the author wrote, "San Francisco-based hedge fund Toro Investment Partners thinks that Ambow shares should be worth at least twice as much."
The Diehards' Guide to Important Questions about Ambow Education
The Obvious Answer
The Diehard Answer
Can I trust a Chinese company whose management already lied to shareholders about its financials?
No, foreign shareholders have no rights in China and management can do whatever they want.
Does not matter, 40% of the company is owned by (foreign) private equity companies that are "hell-bent on restoring the stock to glory" (author's words)
Less than 10 days after Toro's bullish comments to Barron's, Ambow's CFO resigned after allegations from a former employee were revealed to its audit committee that Ambow used bogus acquisitions four years earlier where "the cash purchase price for certain schools was secretly being returned to the company in form of fake software sales that inflated its revenue and income."
If what Ambow did to exaggerate its revenue sounds familiar, it is probably because it is exactly what I have proposed NQ Mobile is doing that has been derided by the diehards as a "conspiracy theory" and a "Further Adventure Down the Muddy Waters' Rabbit Hole," according to the title of one piece.
Immediately after the allegations were raised, PwC still allowed Ambow to release its next quarterly results, a share repurchase program was announced, and the company conducted an "independent" investigation of the allegations. The "independent" investigation was never released, and instead they strung along their investors by announcing that the company received a go-private offer from Baring Private Equity.
Guess what? None of it mattered. Days after the "go private offer" Ambow was halted and then delisted because PwC resigned. As I already mentioned, Ambow Education's annual report was filed late with accounting revisions the year PwC charged them such a high fee, unlike NQ Mobile. The important difference between Ambow in 2011 and NQ Mobile in 2012 is PwC was able to find strong evidence in Ambow's case that warranted an earnings restatement.
While this is certainly positive for NQ Mobile relative to Ambow, having an auditor possibly view you as a higher risk of fraud is still a precarious position for NQ Mobile to be in, particularly since Muddy Waters in 2013 exposed things that PwC unlikely knew before. For example, surveys of 800 consumers and surprise visits to Yidatong's office and pings to its email servers all substantiate the egregious financials that were already showing in NQ Mobile's financial statements in 2012.
Then in 2013 these financials became even more egregious, and the person most well-acquainted with them and NQ Mobile's relationship with PwC, the ex-CFO, decided that it was a good idea to line up a new job and resign before the annual report would be due and his reputation would possibly be tarnished.
Using the Ambow experience as a template, there is only one question in my mind:
Is the evidence against NQ Mobile strong enough to get them delisted or will there simply be a restatement of past earnings?
The story shown by independent surveys, the Muddy Waters report, and others is quite clear. Strong evidence has been provided that:
*NQ Mobile's user base is greatly exaggerated.
*Yidatong is not a real business and is actually controlled by NQ Mobile.
*NQ Mobile's management lied to investors repeatedly (about its two largest customers and about their relationship with Yidatong for example, see article on this here)
*The money from NQ Mobile's IPO could not have entered China the way it is described in the company's annual report given the country's currency controls.
*Several of NQ Mobile's invoices and term deposit slips appear to be fake.
The strong evidence provided by Muddy Waters and others is also substantiated by NQ Mobile's financials, which resemble exactly what one would expect a round tripping business to look like.
In light of the above, if you are a long, what you are betting on essentially is that no matter how illogical and obvious NQ Mobile's business appears, it will make no difference because PwC cannot know for sure since no one confessed or tattled, and the cash is (or was) still there.
That is a pretty low standard considering all it requires is (a) executives to not steal the IPO money right away and be able to raise a convertible bond later; (b) executives to continue receiving generous compensation while not signaling that they had direct knowledge of wrongdoing by resigning before the auditor does.
Maybe PwC will shirk its responsibility, but I would not count on it, particularly in light of the recent pressure from the SEC on the Big 4 Auditors in China. So far PwC can say they have done their job. They reported NQ Mobile's financials so transparently that you can very nearly reverse engineer what they appear to be doing. Now that direct evidence of round tripping has been provided, which they probably lacked in 2012, I expect it to play out like Ambow Education did.
Ambow March 2012
Ambow March 2013
Institutional ownership reported
First CFO resigned five months before annual report was due but remained as a consultant according to company press release
Second CFO resigned after specific allegations surface from an insider
Annual report filing
Filing is delayed in late April 2012 and a number of accounting adjustments are made
Board members resigned and PwC resigned in March 2013
Stock is worth 50% less in 2 months and 80% less in a year
Halted, delisted, bankrupt; Founders kept everything because a VIE is not enforceable in China
If you're long, your best hope is that this never comes down to a decision by PwC at the end of April. I could be wrong. My tournament bracket tells me this can happen when I try to predict the future.
Only one thing is for sure. If you're with the diehards, then the best idea is to dump as much as possible into the stock like Toro Investment Partners, Jim "there is no greater vote of confidence than placing a very large bet" Oberweis, and the latest anonymous Internet tough guy harassing NQ Mobile critics 24-7, Punitive Damages. If and when this goes belly up, it won't be the first time supremely overconfident people were wrong either. One of the companies in the data set, LifeLock (NYSE:LOCK), were so confident in their identity theft product that they put their CEO's social security number on billboards and drove them around busy intersections.
Ultimately, the product was not as good as LifeLock thought, and the CEO's identity was stolen 13 times and they were even sued by the FTC and 35 attorney generals for deceptive marketing.
When you put your social security number on a billboard or title your Seeking Alpha articles about a stock accused of fraud like these below (4 of the last articles about NQ Mobile):
All it really means is you are a diehard. If the LifeLock example is any guide, people that are diehards about money matters are either incredibly naive or they just think that other people are.
No matter how you cut it, it is time to start thinking for yourself, longs. Because as far as the data is concerned, the "blastoff" you've been promised is not going to be from the explosion in NQ Mobile's stock price. It is going to be from the land mine you're about to step on.
1 Excludes examples where business model was clearly unsustainable because the company was over-invested. Overinvestment is defined as having a return on investment of less than one for computer equipment and a return on investment for employees of less than twice the minimum wage of the company's country.
2 I would have visited the exact stores where they distribute their products, but NQ Mobile's investor relations never responded to my email requesting this or any of my emails or calls (other than one where I requested the bank balances).
3 Excludes data from companies that were excluded in the previous chart because they were overinvested (less than $1 return on investment in sales and marketing or product development).
4 This excluded VeriSign from 2008 to 2010, which had large prepaid expense resulting from Lehman's bankruptcy, and Netease from 2009 to 2012, which has large royalty prepayments made to Blizzard and other video game developers that make up the bulk of their prepaid expenses.
5 Equity investment in associates is private companies only. I excluded spin-offs or investments in other public companies, which were both rare. Both equity investments and acquisitions are reported differently depending on the company, and I tried my best to properly categorize the data, but like purchased software, it requires I used my judgment. I only compiled data on companies while they were public.
6 Cash position is the company's cash, term deposits, and short-term investments. Excludes companies whose stock acquisition was more than the cash on their balance sheet that had to pay for the acquisition with stock. Also Sohu in 2000 was excluded because it throws off the scale of the graph since their acquisition was 5 times as large as their revenue.
7 Purchased software is any software not internally-developed, for internal use, or acquired in an acquisition. Some software companies do not separate software internally-used/developed/acquired from purchased. I used my best judgment, but this is a far less black and white categorization than computer equipment and other line items I included in this report.
8 The software is not from the acquisitions of Nationsky or Feiliu because the intangible assets acquired from each are detailed in NQ Mobile's annual report and neither one includes software.
9 Refers to change in gross carrying amount since it is unclear from cash flow statements how much was paid to acquire intangible assets. 2012 excludes purchases of intangible assets.
10 I excluded data from pre-IPO because the acceleration of stock vesting for companies as they approached their IPO distorted the data. The results were not consistent and said very little about the actual business model of the company.
11 This chart plots the change in product development and sales and marketing ROI relative to a company's change in general and administrative ROI. NQ Mobile invests in general and administrative continually but not sales and marketing and product development.
12 Excludes the data from 2011 when a company IPO'd after July and company's shares did not unlock until 2012.
13 Excludes revenue greater than $300 million in revenue and EDU who paid $10 million in auditing fees when they IPO'd in 2006 because it threw off the scale of the graph.
14 Audit fees correlate to a company's revenue and NQ Mobile's are significantly higher without an obvious reason other than their auditor already viewed them as a higher risk of fraud even before the Muddy Waters report. It is unlikely that NQ Mobile's billing is more complex and decentralized, and plenty of companies in the data set have done secondary offerings and convertible bonds. Excludes companies with greater than $1 billion in revenue.
15 NQ Mobile's financials became even more egregious in 2012 from 2011, and PwC's fee for its audit doubled. Excludes EDU in 2007 when its audit fee decreased by $9 million because it throws off the scale.
Editor's Note: This article covers a stock trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.