When we publish a negative report about a stock that appears to be under the influence of stock promoters, the accusations are always the same. We’re accused of being short when in fact we wouldn’t short a stock no matter how bad it looks. That’s because the market is not rational, and shorting a stock is akin to speculating. Nonetheless, plenty of “activist short sellers” out there make a living shorting stocks while publishing lengthy reports to support their positions. It’s a controversial occupation, and one of the world’s most notorious short sellers (so sayeth the FT) is Muddy Waters Capital.
Muddy Waters doesn’t use screens to find short candidates – there are too many false positives and false negatives. Rather their process is much more qualitative. Carson pays particular attention to stocks on a tear, to highly promotional management, to CEOs promising the moon. They like companies which are indebted and resort to trickery to preserve covenants. They will read call transcripts for several years to detect broken promises, management prone to using buzzwords, executives who don’t answer questions.
Credit: Behind the Balance Sheet
It takes about three months for Muddy Waters to produce a short report, and the one they released yesterday absolutely decimated a stock we just invested in – dLocal (DLO).
While we have found no pictures of its CEO wearing black turtlenecks, our research leads us to believe that DLO is likely a fraud.
Credit: Muddy Waters Capital
Loving dLocal Stock
In late September we published a piece titled dLocal Stock: Emerging Markets Payments which concluded that we liked dLocal far more than we expected to. Our accompanying video asked viewers to present whatever red flags they could think of and shoot holes in our bull thesis which was largely validated by the marquee customers dLocal has on board. Amazon, Microsoft, and Google are some of the largest companies in the world and likely to vet vendors extensively before adopting their solutions. Recently, we opened nearly half a position size in dLocal to complement our Adyen holding. Just yesterday, that position lost half its value when the Muddy Waters short report was released. Today, we want to walk through the report to gauge if there’s sufficient evidence to merit exiting our position.
Accounting Inconsistencies
A key premise behind the Muddy Waters report is that the presence of accounting irregularities points to a firm that’s cooking the books.
With DLO, we note material conflicts in reported TPVs, consolidated receivables, and subsidiary level receivables and payables.
Credit: Muddy Waters
The first discrepancy occurs between two charts that show the amount of total payment volume (TPV) coming in from new merchants for a particular year. Says Muddy Waters, “Despite such a massive change, DLO provided no explanation – it did not even highlight that it had made such a revision.” Perhaps these were just mistakes that were overlooked by whoever put these numbers together. The chart with conflicting data is horribly put together to begin with, so it’s no surprise the data is wrong.
This case, along with the others raised in this section of the report, could very well be mistakes made with no malicious intent. The example provided of the same table in a document having two different values (off by 11%) could be a case of the books not being closed while the reports were being produced. Muddy Waters argues that dLocal is having difficulties balancing all the lies they’re telling while another explanation could be the manual systems in place are susceptible to errors. dLocal only began operations in 2016 and went from zero to $2.7 billion in transaction volume in last quarter alone. Their SEC filings specifically call out their dependencies on manual processes as being potentially problematic.
Presently, several of our functions are performed using a number of different information systems that are not integrated. In part because of this, we rely on operations that are performed by individuals rather than automated systems and processes in the operation of all our IT, operations, and treasury-related activities. Accordingly, our treasury functions, for instance, require us to perform many manual reconciliations and other manual steps, which result in a high risk of errors….
Credit: dLocal
dLocal responded to the short report on the same day (always good to see) and said, “dLocal will rebut the allegations in the appropriate forum in due course.” They will likely point to the above disclosure as the reason behind the inconsistencies. Another example given by Muddy Waters relates to several dLocal entities not having matching numbers when comparing the receivables owed by one to the other. Again, this points to sloppy manual processes and is not necessarily indicative of fraud. The icing on the cake is a proposed $31.5 million loan to the President and the CEO who both needed the money to exercise options. Muddy Waters argues that the loan took place while the company states that it actually didn’t. In either case, it’s a related party transaction that dLocal spells out in their regulatory filing document that seems rather benign all things considered.
The High Take Rate
Perhaps the most intriguing critique presented by Muddy Waters surrounds the high take rate dLocal realizes. “Take rate” refers to the revenues dLocal skims off their payment volume in the form of commissions and fees which is expressed as a percentage. Below you can see how high dLocal’s take rate is compared to the competition.
The report goes into detail analyzing why the foreign exchange contributions to take rate are suspicious when compared to other providers that offer a similar product/service. It’s a valid criticism, but we’re inclined to believe that sophisticated institutional investors would have probed this right off the bat prior to investing in the business. General Atlantic, one of the world’s largest private equity firms with over $80 billion in assets under management, has been an investor in dLocal since 2019 and is currently the largest shareholder with around a 19% position and two board seats. They presumably invested in dLocal because of their ability to operate with such high take rates relative to the competition, and the reasons for that should have been thoroughly vetted.
Muddy Waters argues that General Atlantic already recouped their original investment and is playing with “the house’s money” so their holding doesn’t qualify as an endorsement. But what about Tiger Global, a notable investment management firm with $125 billion in AUM, that increased their position last quarter by 138% making them the sixth largest institutional investor in dLocal? Again, one would hope they fully understand the reasons why dLocal is able to operate with such a high take rate relative to the competition.
Other Concerns Raised
The report goes on to list several additional observations that could suggest fraudulent intent. The manual processes in place would particularly suit a company that was cooking the books, and Muddy Waters questions the choice of auditors used by the company. The structure of the company is unnecessarily complex when compared to their competitor Adyen, and the UK subsidiaries appear to be sidestepping UK regulatory oversight which is somewhat understandable. Less regulation means less busywork. dLocal’s two offerings sold $1.4 billion worth of stock of which only a small fraction of the proceeds – $86.5 million – went to the company. The rest “almost entirely benefitted selling shareholders (primarily managers and directors).” Cashing in on your shares when a startup exits is what every founder dreams of as they plough through 80 hour weeks trying to scale a platform. It’s why there are usually lock up periods established so that everyone can’t run for the exits at once. What’s probably more concerning is the turnover they’re seeing with key roles such as the CTO and CCO who didn’t last more than two years.
Not as bad as Editas, but still concerning.
Some Thoughts
Despite the obvious conflict of interest inherent to every short report, they usually contain some valid criticisms and reason for concern. The Muddy Waters report is no different. Given how young dLocal is, and given that they’re operating in emerging markets and across multiple entities, it’s no surprise that there are inconsistencies in accounting numbers, especially when the company all but says to expect that because of a lack of automation. As Muddy Waters points out, why weren’t these problems rectified before the IPO took place? In the best case scenario, what we’re seeing here is the inability for management to identify and prioritize key risks.
After reading through the report and considering the evidence presented, it’s hard to imagine that the entire firm is a house of cards built solely for the purpose of fleecing large institutional investors. The report talks about Google being a key client for dLocal and perhaps their largest client. We would hope that customers like Google, Amazon, and Microsoft would understand the platform well enough to identify fraudulent activities if they’re as blatant as Muddy Waters claims.
dLocal is now having to do damage control for two groups of stakeholders – customers and investors. If they lose one or more key customers, it’s likely the grave accusations in the report have merit. If institutional investors use this opportunity to add to their positions, that’s a vote of confidence that suggests otherwise. In either case, we should see validation of this report in the coming months aside from the initial share price drop.
The reason Theranos shocked the investment community is because nobody saw it coming. Plenty of institutional investors ponied up cash without performing sufficient due diligence and demanding to see a proof of concept. It’s similar to how some “AI companies” have been found to use expensive manual processes to do work with the attention of adding the AI automation stuff later on. If dLocal hasn’t built out their technology stack properly, then the platform will likely run into problems as it scales. If they can’t do basic accounting and reporting without making a bunch of mistakes, then perhaps their platform has similar deficiencies? dLocal needs to address each inconsistency with a root cause analysis and report back to investors how they plan to prevent similar problems from happening going forward. General Atlantic has two seats on dLocal’s Board of Directors so they should be capable of finding out whether there’s actually fraud happening as Muddy Waters claims.
Conclusion
We’ve already established half a position size in dLocal and will not take any action based on what we’ve discussed today. We don’t believe that sufficient evidence has been presented that merits exiting our position, so we’ll see how the company and stakeholders respond to these accusations. If we had capital left to allocate, we wouldn’t be purchasing any shares until we see customers and investors have confidence in the company. Inaction is a sufficient vote of confidence, but any large customers exiting, or key investors dumping large quantities of shares, means we’re also calling it quits on our dLocal position.
Tech investing is extremely risky. Minimize your risk with our stock research, investment tools, and portfolios, and find out which tech stocks you should avoid. Become a Nanalyze Premium member and find out today!
