Elon’s Crumbling Field of Dreams

Summary

TL;DR – As the first post on a short investment this is a bit longer and thorough than usual due to the required level of conviction. Key points of the thesis are as follows: (i) Tesla are not good operators and are not good at manufacturing cars which will hamper their ability to scale; (ii) Autonomous driving is a potential threat to all OEMs; (iii) Electric Vehicles have structural medium-term cost impediments; (iv) Tesla’s balance sheet is very weak; (v) Cash burn coupled with foreseeable capital intensity of growth means Tesla has high external funding needs for the foreseeable future; (vi) High rate of staff departures are a red flag; (vii) Tesla has corporate governance issues; and (viii) Tesla is overvalued.

1-ElonJesus
Elon as viewed by some Tesla / SpaceX acolytes

Elon Musk is saving the earth and Tesla is his gift that will redefine transportation, energy storage and render fossil fuels useless. While this hyperbole highlights Elon’s status as a cult figure, it is not too far from the actual adoration that has been thrown towards Elon Musk by fanboys as Tesla rockets upwards to a $50bn valuation. Tesla is a hype stock, driven by sentiment.

Tesla has completely changed the trajectory of the auto industry

Let’s take a moment to define Tesla as it stands today – Tesla is a car company. They take things, put them together to make an electric vehicle (“EV”) and then sell it. This is primarily how they generate revenue today and for the foreseeable future.

Tesla has yet to dominate or fundamentally change the auto industry as their volumes still represent a miniscule share of the global market, however they have permanently altered the trajectory of the industry by making every other original equipment manufacturer (“OEM”) sit up and take notice.

While Tesla acolytes will have you believe that OEMs are stodgy slow moving conglomerates that are not prepared for the rapid shift to EV, this is not completely true. OEMs are not standing still. Other car companies are well aware of the fundamental shift away from internal combustion engines (“ICE”) and inevitable rise of EVs and have been consistently funding EV R&D driving a stacked pipeline of electrified cars.

As the first mover and with a strong brand associated with innovation, Tesla will remain the electric vehicle company for quite a while. The market will eventually converge as EVs become more mainstream due to OEMs incorporating EVs into their lineup. This will slowly erode the unique EV factor currently attached to the Tesla brand and they too will come to be seen as just another car manufacturer. Yet it shouldn’t be denied that the more enduring factor driving Tesla loyalty, the “cool factor” may drive customer stickiness for a while. After all, drivers in San Francisco prefer the Alcantara seats of a Model X to the leather of an X5 for a reason.  In this way it is similar to another Silicon Valley darling, Apple. Despite being an innovative first mover it eventually saw the commoditisation of their core product, the smartphone.

2 - sidebyside
The hype and hysteria tied to the Model 3 launch was reminiscent of iPhone launches a few years ago

But a key question remains as to how this loyalty will fare over time? While people are willing to part with $1,000 on the latest iPhone (a device they spend many hours a day on) at a premium to comparable devices, will people be as price insensitive to spend $50,000+ on Tesla mass market cars?

Production / Ability to scale

Myth: Just as Tesla has leapt ahead of OEMs and reinvented the car, it is completely redefining the paradigm of automotive manufacturing.

Reality: Making cars isn’t easy, just ask Elon. Also, maybe Toyota (you know the guys who perfected just-in-time and popularized kaizen / six sigma manufacturing) and BMW/Mercedes know a little something about how to run a production line…

The Elon Musk approach is revolutionary for the world. Taking a blank sheet of paper to established rules allows him to constantly innovate better solutions. The problem with this is that sometimes the way things are being done (and being incrementally improved) currently is actually best practice. Tesla is striving to be a vertically integrated car scale manufacturer. The gargantuan nature of this complex task cannot be underestimated and auto execs have questioned the ability of the Tesla model to scale. To consider the two approaches (traditional vs. Tesla) we need to dive into what actually goes on at Fremont, Tesla’s sole manufacturing facility.

New United Motor Manufacturing Inc. (“NUMMI”) was an auto manufacturing JV between GM and Toyota located in Fremont that opened in 1984 and closed in 2010 due to the potential bankruptcy of GM post-GFC. The 5.5m sqft building which was carried on the books at $1.3bn was sold during the recession at the fire-sale price of $42m to Tesla. A shrewd piece of business similar to the sharp purchase by Tesla of a $50m stamping press for $6m that it’s VP of manufacturing Gilbert Passin described as “happily scavenging” from struggling companies.

NUMMI was initially intended to be a mutually beneficial temple of lean manufacturing. A way for Toyota to tailor its Toyota Production System to the US market (different regulations, suppliers and unions) while allowing GM a chance to learn from Toyota’s manufacturing methods. As Toyota implemented its optimized manufacturing principles, NUMMI grew and by 1997 it had 4,844 workers and produced 357,809 vehicles – c. 74 vehicles / worker per year. By contrast, Tesla has approximately 10,000 workers at its Fremont facility and a stated initial capacity of 250,000 (let’s not discuss the feasibility of this number yet), implying a much lower worker efficiency ratio than other OEMs. This may be a contributing factor to Tesla factory workers not even having a spot to park when they arrive to work, as illustrated by this amusing Instagram. However, this metric is not an apples to apples comparison because Tesla is fundamentally different to other OEMs in two meaningful ways:

  1. Tesla is much more vertically integrated
  2. Tesla’s manufacturing line attempts to have an extremely high degree of automation

Vertical integration: A lot of stuff in a Tesla, is made by Tesla. While this may sound obvious it isn’t how many OEMs operate as a typical vehicle has many component systems that are sourced by specialty componentry suppliers including the transmission, interior, electrical, suspension, steering, braking and the list goes on. While Tesla does utilize component suppliers, they engineer and develop a number of components in house that other OEMs do not, with everything from cup holders to seats and even electronic circuit boards made under the Tesla roof. This stands in stark contrast to other OEMs as they focus on their core competency of bringing together components, assembling and building cars (similar to how Boeing / Airbus make planes). The tradeoff of in-housing components to balance quality, margins and production scalability is up for debate but one thing that is objectively true is that the Tesla approach introduces a much higher degree of complexity.

Automation: In what seems like a logical move, Tesla even tried to “electrify” the production line by utilizing robots for multiple steps in the process from fabrication to assembly. For example, stamping body part plates/panels and utilizing a robot arm to shift them into place is standard practice within the industry, but a critical point where Tesla went further is assembly. Bernstein research notes that a typical carmaker will only have 3 to 5% of final assembly (refers to installing the power train, seals, interior fitting etc.) automated. This is not for lack of trying, as the primary customers (auto industry accounted for c. 35% of total industrial robot supply in 2016) of the big 4 robotics manufacturers (FANUC, ABB, Kuka, Yasukawa) are undoubtedly trying to automate where it makes sense and working closely with robotics suppliers. Robots are highly effective for repetitive high-volume tasks (e.g. welding) but are not as effective when it comes to adapting to a different task once designed / programmed for a specific action. Therein lies the primary tradeoff of high cost automation vs. retaining flexibility in a human dependent production line. Auto OEMs have clearly determined that as of today, final assembly is not a part of the manufacturing line that makes sense to automate as humans are more effective. Musk took his blank sheet approach and decided to automate as much as possible. Tesla even went to the lengths of purchasing German manufacturing design specialist Grohmann, a company that used to be utilized by OEMs to help assist in improving manufacturing processes. Tesla has also deployed a massive number of Kuka units (high quality and expensive robots) across their Fremont floor. The capital intensity of this approach is immense and has rapidly accelerated Tesla’s cash burn as shown below:

3 - Capexppegraph

Take a moment to consider the quantum of this capex. Over $3bn of capex in 2017 is high, but shouldn’t this be expected as they construct a new production line for the model 3 along with other Fremont plant improvements? Maybe, but this is not a greenfield project. Remember they paid $42m for the NUMMI plant and the above capex has been to build only the production line, with core infrastructure already in place. Despite this, the capex spend per unit of headline capacity is incredibly high compared to peers based on equity research estimates:

4 - Capexperunit

Even Elon stated on CBS when asked about whether the plant was over automated – “Yes, absolutely. We’ve got too many robots.” Here is the problem with course correcting from here… the exact reason managers carefully consider automation capex before deployment is that it can become stranded capital if ineffective. Automated lines and robots are difficult to redeploy and if ineffective and there could be two crippling consequences for Tesla: (i) Model 3’s rolling off the line that have defects due to robotics teething issues (that will either be sent to customers with issues causing headaches later or that will require time intensive quality control checking on the line); and (ii) The highly automated production line will be difficult to reconfigure for other models in pipeline (Model Y and others) hence payback on investment will be low.

Another principle that the Toyota way of NUMMI inculcated in its managers was the rule of no wastage as well as constant improvement. These were driving principles in establishing world class practices such as just-in-time (“JIT”) production. JIT saves a plant from having to store parts and reduces waste and inventory costs (leading to lower working capital and higher cash flow).

5 - Stacked doors pic
WIRED: “To maximise efficiency, the stamping centre produces a week’s worth of each part before retooling to make another piece. These side-bodies will remain in storage before being brought to the factory’s body centre, where they will be used in the assembly of the unibody chassis”

Whether the above Tesla approach is best practice for optimizing working capital is unknown, as well as whether it aligns with the Toyota just-in-time mantra.

Considering where Tesla is today we can surmise their production situation as follows:

  1. Vertical integration increases complexity and will make scaling difficult
  2. Automation has not been successful thus far, despite billions in capex. Tesla, which is on the currently entering the mass market segment, is at risk of: (i) Stranded capital; (ii) Production delays; and/or (iii) Production defects / product recalls
  3. Fremont / NUMMI is struggling to meet Tesla’s 2018 needs. It will not be sufficient to meet the Tesla 2025 vision. Greenfield sites are needed, with billions of capex needed to build them

Structural issues – Driverless cars on the horizon

Autonomous driving is on the horizon. While regulation and technological barriers will slow the inevitable likelihood of wide scale adoption – the structural shift to autonomous driving will impact Tesla and OEMs substantially. There are plenty of unknowns at this stage but two likely impacts of autonomous vehicles are:

  1. Less cars: The sharing economy increases utilization. Airbnb utilizes empty space like Uber utilizes dormant cars. In a future world where less people buy cars because they just use a ride hailing app, fewer cars will be needed to transport people from point A to point B as cars won’t be sitting in a garage unused for 90% of the week. Less cars = less sales for OEMs.
  2. Bigger customers: As autonomous driving is adopted, ride hailing apps won’t need to fund their main cost – drivers. They will have fleets of driverless cars to service customers. Companies like Toyota and Mercedes will then shift, instead of selling to millions of different individuals under a B2C model, they will suddenly be faced with large corporations as their primary clientele under a B2B model. Also, the margin contributors for many OEMs are the higher end models that they upsell (better engine, aesthetic body kit) to consumers however product mix would likely shift towards more basic models under a fleet purchasing scenario. In an already thin margin business with high capital intensity and cyclical demand, the auto industry faces another headwind – increased customer bargaining power.

Where are we today in autonomous driving? This is harder to ascertain given the multiple competing programs from manufacturers, some shrouded in secrecy while others are trumpeted to the public at trade shows. Misunderstanding of the landscape can be highlighted by the ridiculous headlines that were published a few months ago when many articles referenced a report by Navigant Research that stated Tesla was last among OEMs and startups within the autonomous driving space. You read that right… last. See below for a diagram from Navigant research

6 - Autonomous graphic

The methodology for this report is questionable when trying to rank the companies in the space who will be able to reach L5 capability but the authors clearly didn’t think to sense-check as some basic desktop research can undermine their “ranking”. It is absurd that Tesla ranks behind companies like JLR (who have next to nothing within the autonomous space) and Fiat Chrysler who beyond a partnership with Waymo have made absolutely no progress themselves. To define what the stages of autonomous development are please see the table below:

7 - Autonomouslevels

L5 is clearly the future that all players are striving to but products we have seen so far mostly are at L2 level, with Tesla/Waymo/GM making strides to achieve L3/L4 in the near term under controlled testing conditions. The ability of players to enhance their offering will likely be driven by three factors: (i) Hardware; (ii) Software; and (iii) Data. While OEMs along with GPU/Sensor suppliers are critical for the hardware part of the equation it is more likely that the software and data edge will determine the winner. Within software the players such as Google (Waymo) / Apple / Baidu are likely best placed given existing software development expertise and it appears Google have made a clear strategic decision to focus their efforts on developing an autonomous driving platform. Realising that there are plenty of OEMs and specialist manufacturers without software expertise, Waymo can be to auto OEMs what Android OS is to smartphone OEMs. Software will be the critical aspect that regulators will need to get comfortable with both from a safety and security aspect hence expertise and reputation will likely add some weight. The third factor is data – many players are constantly gathering data through trials however Tesla likely has a substantial edge in this area due to essentially having 100k+ beta units on the road constantly gathering real world data as each Tesla car rolling off the Fremont factory floor is connected over the cloud and feeds data back to the company.

Whether or not Tesla emerges as a key autonomous player is difficult to predict at this stage given the intensity and variety of competition but the overall shift is likely to have a material adverse impact for all OEMs if the outcome is a few large customers and less cars on the road. While this may be of lesser concern if we were analyzing Diamler or Toyota, this point is particularly pertinent to Tesla as most of the intrinsic value of Tesla is attributable to the long term (2025+) as this is when the Street expects them to be highly profitable and cash flow generative.

Structural Issues – EV Profitability / Attractiveness

A cost teardown of an average internal combustion engine (“ICE”) vehicle and an EV reveal a clear differentiation – EVs today cost materially more to produce than comparable ICE cars.

8 - ICEvsEV

As you can see above despite an EV not requiring certain components (transmission etc.) the incremental cost of only the electrified powertrain and EV infrastructure outweighs nearly the total componentry cost of an ICE vehicle. Battery cells which are the primary cost are steadily falling due to technological advancements however the crossover point for EV vs. ICE powertrain costs remains many years out as estimated below:

9 - EVinflectiongraph

Battery pack costs remain make or break for the feasibility of the EV industry over the next few years. Given how critical this is as the primary cost input for Tesla, let’s consider the drivers of battery costs and impact upon profitability.

The main costs of each battery cell are within the precursor metals of the cathode rather than the anode so it makes sense to focus on the chemistry of the cathode. For the sake of simplicity we won’t dive into the nuances of cathode chemistry but if interested I would highly recommend reading the following links to properly understand the COGS conundrum facing Tesla: here , here , here , here , and here. The key takeaway is that cathode powders are expensive. Tesla needs cathode metals to get drastically cheaper to reach their sky high promises on battery costs, gross margins and mileage but they have been getting exponentially more expensive. Two components that are especially sensitive to price volatility and have been rapidly driving up battery costs are lithium and cobalt.

10 - Cobalt price
Cobalt has been the hottest commodity of late, exploding towards $100k/ton
11 - Lithiumchart
Lithium has also seen staggering increases due to unexpected tightness in the market

The supply dynamics of these metals is not as attractive as say, graphite, the core anode base metal that is one of the most abundant resources on the world. Cobalt deposit finds are not common and currently c. 60% of global production comes from the Democratic Republic of Congo, a volatile and unstable country. Cobalt is primarily produced as a byproduct of copper and nickel production. Therefore due to limited growth in these metals, supply constraints for Cobalt are likely to persist. Lithium supply is not as precarious but has also faced constraints which have triggered sharp price hikes due to the production oligopoly (SQM, FMC, Albemarle) not expanding supply fast enough. Supply-demand imbalances will be exacerbated rapidly once major auto OEMs start producing EVs in meaningful volumes. Therefore many manufacturers have been jostling to secure off take agreements with limited success so far. Given the uncertainty around commodity prices it is difficult to see how Tesla is going to reach its gross margin targets in a rising input cost environment. While the much vaunted Gigafactory gives them a scale advantage in cell production, they are still beholden to the same resource risk as everyone else. A resource risk that is currently eroding their gross margins at the time they need cash flow the most.

Another headwind facing EV adoption has been the muted oil price environment since 2014. Many EV supporters have noted the upfront premium of EVs is recovered over time due to the lower operating costs compared to ICE vehicles however as oil prices have dropped significantly from >$100/bbl this “breakeven” point for the consumer has been pushed further out.

While EVs are clearly a fantastic product, Tesla itself is testament to the low barriers to entry in the auto OEM industry as a swathe of pure play EV startups have followed in Tesla’s footsteps. Today there are no credible threats but beyond the primary restriction of capital, the hurdles to step into Tesla’s turf appear to be manageable.

Demand

For a long period of time Tesla has been Elon’s very own Field of Dreams as for anything he has built, people have come. However pre-2018 Tesla is a different proposition to what investors hope for future Tesla as it transitions from the niche of high end performance vehicles to a mass market brand. Demand dynamics will be different moving forward.

Tesla generated immense hype when launching pre-sales for the Model 3, garnering nearly 400,000 deposits of $1,000 each in April 2016. This is even more impressive when taking into account that all customers had seen so far was a silhouette of the car and some loose details such as the $35,000 base price. The deposits are c. 500k now according to Elon and he must be applauded as this move provided the best source of working cap financing Tesla will ever receive, with potential future customers giving them a $500m interest free loan. However, while parting with $1,000 is a notable commitment, the conversion rate of deposits into actual purchases is unknown at this stage and it is worth noting that they are fully refundable. Three factors likely to impede conversion are: (i) Model 3’s aren’t really going to cost $35k as initially advertised; (ii) Fading US EV tax credits; and (iii) People are having to wait longer than they expected.

Model 3 ASP: Higher priced cars have fatter margins, and this is critically important for EVs to spread the high electrical powertrain costs over a higher base. Given Tesla’s volatile gross margins on the $100k+ Model S and X it is very difficult to see how they make money on a $35,000 Model 3. The company knows this and has so far been “skimming” Model 3 demand by upselling customers or focusing on those customers who are willing to liberally option their vehicles to improve short term cash flow. The problem with skimming is that it exhausts the most lucrative segment of demand at an early juncture. Also it is questionable whether Tesla will ever really sell many cars at the $35k advertised base price which that leads to the question, if customers realise their Model 3 will cost 20%+ more than initially expected, will they convert their deposits?

Fading EV Tax credits: As Tesla continues to sell cars it moves closer to the 200,000 domestic unit cap – once it hits the cap the $7,500 per unit subsidy provided to customers by Uncle Sam starts to roll off within two quarters to $0. They will hit this cap in 2018 and then be in a position of relative weakness to both comparable ICE models as well as competing OEM EVs. Everyone else (except for General Motors who will also hit the cap sooner rather than later) will have a material pricing advantage and it is at this mass market price point where demand is most sensitive to price. The GOP tax bill tabled in 2017 proposed to repeal the incentive and if the bill is passed in this form then it would level the EV playing field but problematically place the whole industry at a disadvantage to ICE models as manufacturers will have to absorb higher EV powertrain costs or consumers will likely choose better value ICE models.

As a side note, even though the next few are completely focused on Model 3 deliveries it is worth keeping an eye on sequential numbers for Model X/S shipments. Recent numbers have not been trending well as they have fallen quarter-on-quarter. The company has stated Model 3 production is the focus and while this may be true, another explanation could be flat lining or fading demand for these higher priced models. If this is the case, product mix deterioration will be yet another hindrance to Tesla’s path to profitability.

All the above factors are short term however the real threat on the horizon is the eventual arrival of competitors on the EV scene. Nearly all OEMs have announced significant EV pipeline programs with targets for a specific number of models to be fully electrified by 2025. This will start to commoditise the value proposition of Tesla.

Recession: While it is impossible to forecast recessions it is worth acknowledging we are currently in the ninth year of the business cycle post-GFC. This likely means we are in the late stages of the economic cycle and risk of a recession is particular heightened in the next three years. The auto industry is somewhat cyclical in nature with a drop in US discretionary spending likely to hinder Tesla demand. Also, asset markets generally price a recession in earlier and fall 6-12 months ahead of an underlying recession, which triggers a tightening of capital markets. For these reasons, a US recession over the medium term represents a potential catalyst for the Tesla short thesis.

Balance sheet

Tesla’s balance sheet is a constant point of discussion as people debate if (when?) they will need to return to capital markets to keep the company going. Tesla has, to date, been reliant on open and accommodative capital markets as they have expanded their balance sheet in order to hit operational milestones. To state the obvious, capital is not free. Every time Tesla raises capital they need to generate a higher return than the cost of that capital to create value. They have not been doing this so far due to their malinvestment (remember those really pricey Kuka robots?) that has not yet resulted in corresponding cash flows, instead probably causing a likely low through cycle ROIC (assuming they do manage to get to operating profitability one day).

Every dollar of debt Tesla adds creates a servicing burden while every dollar of equity raised dilutes shareholders (however equity raising done at current lofty valuation makes this less painful) and therefore further raisings will likely be anathema to existing shareholders. However, the contrary may be much more painful… a scenario under which capital markets are not open. While it is unlikely capital markets will freeze, the loose financing conditions of the past 5 years are unlikely to continue given the cycle is now in a tightening phase. Herein lies a key risk and catalyst for the short thesis, if there is deterioration in financial markets and capital markets become costly or inaccessible to a junk-rated borrower such as Tesla it will wreak havoc on the valuation. Tesla in its current state is a black hole for cash completely reliant on external funding.

This is before considering the questionable accounting policies utilized by Tesla as they pro-forma their way to a better picture of financial health than may be true. Ignoring GAAP boundaries Tesla has utilized some creative accounting in a concerning way when reporting certain metrics. An example of this is ignoring obligations attached to lease car revenue – Tesla provides a resale value guarantee which would represent a real liability in the future if the resale market for second hand Tesla’s were to ever turn (as would likely occur in a recession or if Tesla’s became more mainstream). Also the company is carrying a rapidly growing deferred capex obligation that is equivalent to debt, only adding to the financing burden facing the company.

Valuation

Any valuation of Tesla will ultimately hinge on their ability to achieve strong profitability (multiples) or cash flow (intrinsic value). Tesla is in an early stage of its corporate life cycle hence benchmarking to comps will not be a useful exercise however it is still a valuable lens to understand required profitability and cash flows to support the current valuation.

As a car manufacturer to become a $100bn company Tesla will need to be a global mass market manufacturer across price points and models like VW or BMW.

graph1 - ASPperCar

When analyzing the above, remember that Tesla (due to the cost economics of EVs) will struggle to make any money at the gross margin level on their entry level base $35k Model 3. This is despite the most premium full range brands in the world (BMW/Diamler) having ASPs of c. $45k. To hit their valuation they not only need to move down market to the masses in a meaningful manner, but they need to improve per unit COGS enough so they are able to drive high gross margins to cover their overhead costs and deliver healthy EBIT margins.

graph2 - EBITmargin

As shown above best in class large scale operators Diamler and BMW post EBIT margins of 8 – 11%, much higher than other large OEMs such as Ford that operate at mid-single digits in the best case. This is primarily due to different gross margin profiles as Ford has a GM margin of 10% compared to the more premium operators who are able to achieve high teen GM margins. Tesla, has communicated its long term GM target to be 25%. This would suggest it can reach a similar EBIT margin of major players however this masks a few critical misinterpretations of Tesla’s GM. In the recent quarter Tesla was able to achieve a GM margin of “slightly above 25%” on Model S and Model X – this should be cause for concern for investors. While on the surface this appears to show it can reach best in class profitability similar to other OEMs once ramped, this GM profitability is on very expensive cars that will definitely be delivering higher margins than lower priced models that Tesla will be shipping in the future (Model 3/Y). Also Tesla gross margin benefits from their direct to consumer model where they capture the 10-15% markup that dealerships usually apply under a traditional OEM distribution model. However this markup obviously does not come without associated costs and Tesla’s service centers around the world generate real costs (remember Tesla has had some major recalls in its history) that are captured in SG&A that other OEMs do not have to consider. Also unlike many other OEMs, Tesla does not expense its R&D despite it being a recurring annual cost. If considering service and R&D expenditures within the gross margin for Tesla then it suddenly becomes much more difficult to see Tesla’s path to achieving an operating margin than can match/outstrip the BMWs of the world. It appears at current valuation levels the market is giving them credit for being able to reach a 10%+ operating margin not only on Model S/X but on the Model 3, Model Y, Semi and future models they are likely to roll out.

graph3 - Capex

Car manufacturers (and investors) also have to consider the capital intensity of the industry. In steady state, it appears that OEMs spend 5-8% of their sales on capex. Tesla’s revenue/PPE should scale over time to reach peer levels however that would assume they invest with similar capital discipline to their peers, by only incurring capex that will deliver sufficient ROIC. An issue of investing in Tesla today is that in an already capital intensive industry, they have a much higher capital requirement than their peers. Tesla’s requirement for greenfield facilities, more service centers (a capital contribution instead made by dealers under a traditional model), more superchargers (investments that do not deliver matching cash flows), and an inflated overhead cost structure mean that Tesla’s cash needs are unlike any other OEM. This combined with their high WC funding needs as they grow highlights the validity of concerns held by investors that Tesla will continue to dilute them through equity issuance’s.

graph4 - Cash

While Tesla is constantly navigating funding needs and solutions quarter by quarter it is worth noting that their field of competitors (who are on the precipice of entering the EV space) remain well capitalized and have deep cash reserves to invest. In a capital intensive business that is exposed to the consumer cycle, peers clearly see the benefit of retaining financial flexibility.

When considering Tesla’s valuation a more high level approach can be employed given the early stage of their corporate life and uncertainty around long term growth – we can consider a scenario whereby Tesla becomes a global mass market OEM with premium pricing and best in class margins. If Tesla is able to ship 2m units (BMW and Diamler shipped 2.4m and 2.2m respectively in 2016) at a $50k ASP then they should be able to reach $100bn in revenue within their auto business. If they achieve operating margins matching best in class operators of 10% then they should generate $10bn in EBIT. Assuming a 10x EV/EBIT then Tesla could feasibly be a $100bn enterprise in the future, justifying the current Enterprise Value of $60bn+. For the various reasons discussed previously, including the profitability profile and cash needs of Tesla, this optimistic scenario playing out is unlikely.

Other

Key man risk: Tesla’s greatest blessing of having a superstar leader in Elon Musk also has represents key man risk as Elon has been stunningly successful in various fields and has made no secret of his dream to colonise Mars. A scenario under which he one day leaves Tesla for SpaceX or another grand venture is not completely implausible

Staff departures:  Management departures are commonly cited by experienced short investors as a red flag. If that’s the case then Tesla is definitely worth a closer look as their C-Suite is a revolving door. It would take too long to detail all the departures but thankfully reputed short seller Jim Chanos has collated a list here. Senior personnel across many units (engineering / autopilot / finance) have consistently departed, many likely leaving behind lucrative stock packages which begs the question – why walk away from a bunch of Tesla stock or options as an insider?

Governance: Tesla is a one man band and Elon makes all the calls. If you believe in him as a visionary / executive then this shouldn’t deter you, as many companies with high profile founders have gone as far as deploying dual class shareholding structures to retain control (FB/SNAP/GOOG). Of course this kind of reliance on an individual becomes a corporate governance concern when there are strategic missteps and outside investors are not able to have their interests protected. An example of questionable governance at Tesla was the $2.8bn acquisition of SolarCity, the solar energy company run by Elon’s first cousins Lyndon and Peter Rive. The multi-billion dollar takeout of a company with questionable strategic fit, in which Elon was the chairman and largest shareholder, should have been cause for concern and the market reacted as Tesla’s shares slid 11% on announcement. The strategic rationale of combining two entities with limited synergies (yes it creates a complete energy company from generation to use, but what exact revenue and cost synergies was Tesla planning to extract to justify the 25% premium?) was never that convincing and two years on we are yet to see the industrial logic of the acquisition play out even though it is still early days. The timing was worrisome too, as Tesla bought a company that was bleeding cash losses at a time when Tesla was cash flow negative itself and raising equity to fund further losses.

More recently, institutional investors have become increasingly concerned about the Board of Directors. Activist firm CtW has been publicly questioning the independence of non-executive directors as they completely lack manufacturing or relevant auto experience but are instead individuals with close ties to Elon. The board has barely changed since floating in 2010.

The bull case

Tesla didn’t reach its stratospheric valuation for no reason, as investors have offered various reasons to explain why Tesla will one day be a $100bn+ company. These generally revolve around one of four key competitive advantages:

  1. Energy dominance: Tesla isn’t a car company, they are an energy company. They are a renewable energy production (Solar panels), storage (Powerwall) and usage (Tesla cars) company. Admittedly, Tesla does all of these things. However none of their non-car businesses provide a meaningful contribution to their revenue today or for the foreseeable future. Also, does Tesla have a competitive advantage in the crowded field of electricity generation and storage?
  2. Battery technology: “No EV cars will be able to compete with Tesla because they have the best battery tech in the industry”. Tesla today has a meaningful edge in battery tech / chemistry however it is worth noting that batteries will eventually become a low cost commodity and that much of the discussed battery manufacturing IP resides with their JV partner for the gigafactory, Panasonic. A partner which is expected to partner with other OEMs such as Toyota in the near future.
  3. EV Infrastructure: Tesla’s network of superchargers will become the backbone of EV infrastructure across the country (potentially world) as it doesn’t make sense to duplicate and EV users will be drawn to the installed network. The vast majority of driving by many users is done within a single urban area that may easily be covered by the range of EVs so the utility to the masses of this advantage is questionable. To date Model X/S customers have used superchargers free of charge but this is set to change as Model 3 users will be charged per use. This investment again is likely to represent a very low rate of return on capital for the company while Tesla also runs the risk of stranded capital if there is a substantial shift in technology
  4. Autonomous driving: Tesla due to its data advantage and head start will be the dominant autonomous vehicle player. As discussed before, Tesla does have a data advantage but whether they will parlay this to become the leading autonomous player is debatable. At this stage it is difficult to predict however all OEMs likely face headwinds from the structural shift to autonomous. If Tesla does emerge as a dominant provider of autonomous technology itself then it may go on to justify an extremely high valuation.

Tesla makes fantastic cars. If they can retain a battery chemistry and technology edge then their cars may always offer better value for money and performance than comparable vehicles while their network infrastructure could become the backbone of EV transportation across the world. Manufacturing problems are temporary and the smart people at Tesla will eventually solve the difficulties of large scale auto manufacturing. Over time they will build a closed loop of Tesla energy as homes generates power from Tesla solar cells, store energy off grid in a Tesla Powerwall battery and are transported by a driverless Tesla car. This will entrench the Tesla brand halo and over time Tesla becomes the dominant energy and transportation company. While this bull thesis and its various parts are all plausible, the likelihood of this case is questionable and one investors need to weigh when considering to go long or short the stock.

Strategy

A common maxim of equity investing is “do not short based on valuation”. Tesla is overvalued however the market could easily drive Tesla’s valuation north of $100bn within the next few months if enough hype was generated as fundamentals are clearly not the driving force behind price action currently. At some point fundamentals will matter, but in the near term that does not appear to be the case. There are various catalysts to the downside however the following would likely see a substantial change in the Tesla narrative/price action: (i) Capital raises (whether it is equity or debt, it will spell bad news for existing equity holders; (ii) Model 3 failures – Production issues mean delivery targets fail to be met and cash burn issues are exacerbated; (iii) Model X/S fading demand – higher margin models sequential deliveries have started to fall further pressuring cash flow; (iv) Demand issues – Model 3 conversions from deposits do not meet expectations; and (v) US Recession (premium car demand would be hampered).

Tesla’s ability to operate as a going concern is reliant on capital markets being open. Under a scenario in which risks materialize, the company may realise that this is not always the case.

For the multitude of reasons discussed throughout this post, Tesla represents a compelling short. Despite having a fantastic product, the business model is not investable at a $60bn+ enterprise valuation with the share price currently at $319 having substantial room to fall. Shorting a high beta name such as Tesla should also act as an especially defensive holding through the next downturn.

Disclosure: Short Tesla (NASDAQ:TSLA)

Critical Mass

Company overview

Criteo is an advertising technology company that was a three man start-up in Paris in 2005 before launching its first product in 2008. It sits in a space (adtech) that is highly competitive, at the mercy of being sandwiched between giant customers / suppliers, and constantly evolving – don’t throw the whole segment in the bin yet though, as Criteo is a platform that has proven useful to customers with its data driven offering.

At its core Criteo is a retargeting company – this means that if you browse for flights on Expedia and then don’t end up buying them, ads for those flights you were looking at or similar flights that are predicted to be likely purchase items for you will be “re-targeted” to you on other platforms (e.g. while you browse Facebook or a news website). This is known as programmatic advertising whereby display ads are not random but rather programmed for you. Criteo is a market leader within this niche.

How does it work?

Criteo connects with an eCommerce retailer through first party data integration. We will continue with the Expedia example: Criteo will form a direct sales relationship with Expedia and establish a campaign based on which customers they seek to target and how they would like to reach them. Then Criteo will gather data on Expedia customers when they browse on the Expedia website as it will now be embedded with the Criteo cookie. This data is matched to an identifier (random character sequence, not your name to ensure privacy is protected) and then collated with the users browsing activity across all other platforms Criteo has access to. From these disparate data sources, Criteo’s purpose built engine is able to form a mosaic of each users browsing and purchasing patterns and able to craft personalized programmed targeted ads. Criteo’s engine utilizes this data to calculate the cost-benefit of placing an ad in front of a user live (in under 150 milliseconds) as they scroll Facebook, based on that users click/purchase probability and the traffic acquisition cost (“TAC”) of placing that ad in front of them. This allows Criteo to place real-time bids on ads on Facebook only when the engine predicts a profitable spread may be made by Criteo as it charges on a cost-per-click (“CPC”) basis. This is an advantage as display advertising is increasingly shifting to a real-time bidding model on websites as programmatic increases in popularity for advertisers. These ads are intended to deliver the right product placement at the right time in order to increase advertising ROI. This is the core of “programmatic” retargeting advertising and it is experiencing rapid growth transforming display advertising from its previously scattergun approach (where advertisers pay on vague cost per view metrics) to a targeted higher ROI advertising method that instead is charged on a CPC basis.

This method arguably creates a positive feedback cycle for Criteo’s data advantage. Through their existing partnerships they have access to >$550bn of annual online sales, or 1.2bn+ active monthly shoppers. They are then able to reach and re-target these shoppers through thousands of direct publishers (think the Washington Post, or ESPN). Due to their rich data set and personalized ads they are able to generate industry best ROIs for advertisers, generating >$27bn annual post-click sales for >15,000 retailers. This feeds back into the loop by increasing their access to commerce data and hence the Criteo engine should continuously be improving advertiser ROI as it accesses more and more user behavior and creates a more refined profile of each user as the platform reaches critical mass with its data sets.

In its current form, Criteo is the leading independent re-targeting platform however it can leverage its data set to garner more of the ad spend pie. A new vertical that Criteo mgmt. have focused on is Commerce Marketing, or inspiring people to buy things. Using their identity profile of users, they can work with retailers and brands to target specific demographics that are likely to engage with a brand and accordingly place ads for products in front of those users at the right time on the right website/app. This will likely increase in importance in the future as the trend toward programmatic ads play out in various forms for online display and Criteo shifts from being able to only target a single product purchase to instead acquiring customers.

Off a cliff

SP CRTO

The share price has been on a steady descent since May as shown above as it has been increasingly subject to negative sentiment due to the announced introduction of Intelligent Tracking Prevention (“ITP”) by Apple to Safari. What this technology intended to do was cut Criteo off at the source by not allowing their cookies to track user activity on Safari hence Criteo could not re-target ads to users as it is unable to view their browsing. Criteo management initially had a technological workaround that was going to mitigate c. 50% of the impact from the protocol change, yet the stock continued to be pressured. The negative sentiment was eventually justified when last week the stock dropped 25% in a day as management announced the iOS update blocked their workaround and therefore instead of the revenue headwind from ITP being 9-13% as initially guided, it would likely represent a 22% drag on 2018 revenue as they were cut off from access to Safari. The harsh reaction of the market to this news likely prices in the worst case scenario for ITP with Safari sourced revenues lost for perpetuity. This may be an overreaction by the market or it may be the market punishing mgmt. for inadequately communicating the Apple policy shift and misguiding the impact initially.

What am I investing in?

After speaking to a friend who looked at Criteo to invest in recently, he told me that after some diligence he realized ultimately it wasn’t a scalable tech story but instead an ad agency which sells digital content. It is worth noting Criteo has a direct relationship with 87% of its clients so this assertion appears valid especially when considering the required sales network to support revenue growth. To evaluate it as an ad agency, we must instead look at the unit economics of the business to assess profitability, sales and marketing productivity (customer acquisition cost), and retention rates / churn.

CRTO Financials

From the above historical financials (2013 -16 was a period of high growth) we can see what Criteo’s cost structure looks like with R&D running at 14% to maintain product development, S&O is 35 – 40% in order to acquire and service customers, while G&A slowly sees the benefit of operational leverage. Given the hit from ITP, it is likely EBITDA margins will contract substantially from the c. 30% LTM due to inability to nimbly right size the sales force.

Given Criteo’s business model appears consistently profitable with incremental S&M driving top line growth, the critical operational metric is churn which will define unit economics. This gauge of customer satisfaction should be the key warning sign to assess in each quarterly update from the company. Criteo has maintained c. 90% retention rates for the past five years along with strong net client adds each quarter while 78% of their LTM revenue ex-TAC came from uncapped client budgets. Given their performance thus far it appears investing in Criteo’s ability to continue to build strong client relationships and maintain access to client ad budgets is a worthwhile proposition.

Risks

Fraud: As an ad agency, client satisfaction is critical and any dent to Criteo’s reputation could be fatal to the business. Gotham City Research have released numerous reports that claim Criteo garners clicks from fraudulent websites and clickbots that misstate effectiveness of campaigns. Such claims may be true given the difficulty in validating adtech data and nascent shift to online ad spend for brands with many still working out optimal digital strategy, however it is difficult to qualify this claim. Criteo’s historical retention rate and same client spend growth suggests clients have been overwhelmingly supportive of their platform thus far.

Privacy: Increasing concern about use of cookies is a constant risk to adtech players. It is this concern that led Apple to prevent tracking by implementing ITP on Safari. There is a risk that this may have a spillover effect to the rest of the Criteo ecosystem as customers may perceive Criteo’s targeting as limited. European regulatory changes to be implemented should have a limited impact on Criteo, while Chrome is unlikely to mirror Apple’s browser change as it would adversely impact advertisers and hence the Google advertising platform. Also Apple is unlikely to extend ITP to AppStore applications as this would impact those apps attractiveness to advertisers and publishers would migrate to Android, harming the iOS ecosystem

Market power: Criteo is the largest independent in a space dominated by two giants, Google and Facebook. They control the majority of online ad spend and offer advertisers comprehensive advertising solutions, therefore they may look to further pressure adtech players by replicating their business models at scale. Criteo does offer an alternative to publishers and brands otherwise being forced to pass on all their sensitive customer information to Google and Facebook, which they may be averse to

Valuation

Criteo has an attractive business model due to its high margins and growth which is relatively capital light (R&D at c. 14%). This lends itself to an attractive cash flow generation profile which will consistently de-risk the stock.

On a trailing basis (2017) the company trades on a P/E multiple of c. 9.5x and EV / EBITDA multiple of c. 4x. These attractive low multiples don’t look so rosy on a forward basis as consensus 2018E EPS forecasts a 50% drop in EPS, even though revenue is roughly flat year-on-year. Non-Safari business growth is expected to offset the 22% revenue hole left by the loss of Safari revenue. The significant EPS fall is due to the significant deleveraging as Criteo will have an inflated cost structure (Sales personnel) as they did not plan for ITP revenue contraction.

Despite the temporary earnings pressure posed by ITP in 2018, the company will continue to be cash flow generative and by end of 2017 it will have 25% of its current market cap in cash.

View

Criteo is a risky investment. Remain cautiously optimistic due to valuation, cash flow generation and entry point. Ultimately the risk of permanent capital impairment is evident as Criteo may be threatened by regulatory changes to protect consumer privacy, fraud risk attached to opaque advertising practices, or changes in tact by either Google/Facebook to apply more pressure to independents. Despite these risks, at current levels it is worth investing a smaller stake in Criteo as the risk/reward appears to offer reasonable upside over the medium term. Long term growth remains attractive due to the rise of programmatic advertising, and buying it today on a trailing EV/EBITDA of 4x (a fair indication of Criteo’s earning power) offers upside.

Disclosure: I am long Criteo (NASDAQ: CRTO)

Viva la revolución

Latin America is finally being liberated to the age of eCommerce and the company best placed to capitalise on this is MercadoLibre.

noragrets
#noragrets

It is never healthy to dwell on what could have been, especially in the game of investing, but many people regret not riding the decade long run up in Apple, Google, and [insert tech name here]. Similarly many are amazed at the wealth created for Amazon shareholders as it recently blew through a 500bn market cap, yet I always saw it as trading far beyond reasonable value for a company that couldn’t even turn a profit. What my value-orientated perspective missed was that Amazon & Jeff Bezos had a singular focus – a moat. Bereft of profits it had cash flow, to continue to plough into new ventures and innovations as investors gave them credit to continue to execute with the vision that one day they should be the last one standing in a winner-takes-all landscape.

Second bite at the apple

Hit rewind a couple of years on the Amazon corporate timeline and we have a similar opportunity today, MercadoLibre (“ML”). ML is often dubbed the Amazon or eBay of Latin America (“latam”). When looking at the ecosystem of ML however it instead reminds me more of an Alibaba. At the core lies Marketplace, a place for third party buyers and sellers to transact either through an auction or fixed price format similar to eBay where a fee is skimmed on each dollar of Gross Merchandise Value (“GMV”) sold through the platform. Once a transaction is completed, payment is encouraged to be made via MercadoPago a payments solution in the mould of a Paypal where fees are taken as a % of Total Payment Value (“TPV”). Financing is also bundled into MercadoPago whereby users may elect to pay via instalments, and then ML can offload credit risk to underwriters and take a revshare. MercadoEnvios is their third party logistics (“3PL”) solution that merchants can utilize to take care of the logistics of transactions (just like Cianiao for Alibaba). MercadoShops offers an end-to-end solution for offline retailers to create their eCommerce store for a fixed fee, similar to Shopify. They also have a sizeable classifieds business that works off a fixed fee business model rather than a take rate on transaction value. Finally, there is also MercadoClics which offers advertisers real estate on ML websites and charges on a cost-per-click basis similar to other eCommerce websites such as Amazon and Alibaba. ML is the dominant eCommerce player in Latin America today, and as you can see it could easily be viewed as the Amazon (B2C), eBay (C2C), Craigslist (C2C), Shopify and PayPal of latam. Despite the attractiveness of such a description it is also worth noting that global behemoths such as Amazon and Alibaba have already built similar ecosystems in their home markets and there is nothing stopping them from stepping on MLs toes.

Multi-decade story

The Total Addressable Market (“TAM”) is huge. Latin America has over 600m people, that’s nearly as much as the US and EU, combined. Yet due to the lagging telco infrastructure in the region, lower disposable income and lack of adoption, internet penetration in the region severely lags other markets.

Yugetrump
The TAM is yuuuuuuuuuuuge

Emerging markets across latam offer secular GDP/capita growth as the middle class grows in the region and the macro growth will bring a host of benefits for ML. Growth in internet penetration to catch up with the developed world will see the world of e-commerce open up to more of the 600m consumers in end markets.

Internet penetration (% of population)

2010 2017 Penetration increase
Latin America / Caribbean 36% 62% 26%
Asia 23% 47% 24%
Europe 58% 80% 22%
North America 79% 88% 9%

It is still very early in the innings for eCommerce and widespread adoption of smartphones will further support the strong positive momentum for penetration increases region wide. There’s not much value in trying to perfectly predict growth but the basic message for regional internet and smartphone penetration is this – it’s going up. The trajectory of adoption will of course depend on infrastructure spending and discretionary income and therefore overall economic health is likely to be the underpinning factor.

eCommerce as a % of total retail sales

2015 2016 2016 YoY growth in eComm sales
Latam 2.2% 2.6% 22%
Brazil 3.1% 3.6% 14%
Mexico 1.4% 1.7% 27%

eCommerce catching up – Consumers in the region haven’t appeared to widely and completely embrace eCommerce as a medium yet, as shown by ecommerce penetration as % of retail above. This is a key lever for growth as latam consumers get more accustomed to using the internet to transact. The blue print is clear as developed markets have high single digit eComm penetration and China the truly digital society at c. 15%!

ecomm penetration
ML markets (yellow) significantly lag developed nations / China

However, it is worth noting there are challenges specific to the region including mistrust of online payments, and logistical challenges (unionized state postal service) meaning delivery times are often relatively slow. Smartphone adoption (mCommerce share of eCommerce is growing rapidly) and familiarity with the internet will eventually help latam eCommerce growth inflect.

Macro double-edged sword – Key risk but reasons for optimism

The macro story is a double edged sword as the long history of political instability suggests. The headline risk is political turmoil that may trigger GDP slowdowns, spiraling inflation and FX deterioration which would severely hamper ML fundamentals. It is US-listed hence reports in USD$ and latam currency depreciation as experienced in the past few years substantially reduces reported revenues/earnings.

Maduro
Dic(k)tator Maduro loves a hi-five

Brazil (54% of 2016 revenue) – The BOVESPA has rallied significantly in recent times from its lows as investors begin to see the light at the end of the tunnel as the country slowly emerges from one of its worst ever recessions. While the outcome of the next election is unclear at this point it is difficult to see a far left (Lula) or far right (Bolsonaro) candidate winning. Recent data supports the uptick in sentiment as real growth in wages and economic activity appears to be returning.

Argentina (31% of 2016 revenue) looks set to pick up over the next few years as the political environment remains stable as recent polling showing President Macri’s coalition Cambienos being favored which would suggest economic reforms will continue. Political certainty and healthy demand for government bonds should lend support to the currency. There is also very low leverage in the private sector (20% of GDP) which allows for a multi-year credit cycle to fuel growth. The central bank also have acknowledged they are committed to fighting inflation which has been a drag in recent periods.

Others: Mexico (6% of revenue) has shown relative strength in the region and with stable political leadership looks poised for steady growth (c. 2%) in the near term. Venezuela used to be a core market contributing 10% of 2014 revenues before President (Dictator) Maduro plunged the country deep into turmoil and essentially turned it into a failed state (Venezuela contributed 4% of 2016 revenues).

Payments Platform

As with many emerging markets, latam has a significant unbanked population and payments offers a huge runway for growth to accompany internet adoption. Cash is still the prevalent form of payment however MercadoPago TPV growth has outpaced GMV growth as the TPV/GMV ratio tracks a very similar path to that of Paypal TPV growth vis-à-vis eBay GMV growth as it grows to be a platform beyond ML. The option value embedded in MercadoPago is extremely high as a standalone fintech enterprise that will be able to resolve a key friction point of eCommerce via payments but also offer a multitude of services beyond that of an online payment gateway (e.g. Consumer lending, Vendor financing, Off-platform payments processing etc.).

GDPbubble
MercadoPago has a pathway to become the dominant payments platform in the region

Competitor risk

Is first mover advantage enough? Network effects built into a marketplace model give MercadoLibre a big edge. Longer term a capable management team with local knowledge improves understanding of end markets and users better than international players. Ultimately it becomes a question of business model. Alibaba operates a similar capital-light model to MELI and moved first in China, when eBay subsequently tried to enter China they failed and now Alibaba has a dominant position. JD continues to be a threat due to the advantages to the consumer of a capital heavy vertically integrated model which allows for a better customer experience in certain instances (similar to AMZN logistics network).

Bezos
The squeeze is about to come…

Initially when reports of Amazon entering Brazil were published two months ago ML stock plunged 10% as clearly the market acknowledged the threat of the Bezos behemoth. ML mgmt. have faced this threat head on by playing an aggressive share game to reduce the risk posed by the AMZN model – their solution was to subsidise free shipping across their Brazil website. This is an expensive move as the previously healthy margins take a huge hit to support growth and protect share. I do not view this as temporary. eCommerce customers expect free shipping and I don’t see a future whereby ML will be able to revert back to having healthy margins and not having to subsidise shipping costs. This is a fantastic move by ML as it means management are taking a long term view to being the dominant platform in latam, just as they have done since 1999. This singular focus is similar to the approach taken by Amazon in North America. The resultant impact is likely going to be no or very low profitability for the foreseeable future (c. 5 years) as ML seeks to ride the wave up, but LT EBITDA margins should ultimately fall in-line with targeted margins for similar players at c. 10 – 15%. The short term depression of earnings means share price is likely to be volatile over the next few years, creating stress for holders but opportunities for those in it for the long haul.

Management

Management are critical as minor strategic missteps in an industry that moves as fast as eCommerce can be fatal. Marcos Galperin the founder and CEO has almost all his net worth ($1bn) tied up in stock. However it isn’t his financially aligned interest in ML stock that gives me comfort – it’s his story. Long story short he was the heir to SADESA (his family business which is one of the largest leather companies in the world) yet instead founded ML in his mid-20’s while doing his MBA with a vision to build the eBay of Argentina. His vision has grown along the years and it is clear today that his aspirations were always greater than financial reward as he seeks to build a truly fantastic company. CEO’s with this tunnel vision eventually create tremendous value for shareholders in the long term. Still incredibly young at 45 years old he has a long road ahead to build ML into something completely different to today. The rest of the senior management team are also tied to long term retention plans with cash vesting over rolling 8 year windows.

Valuation

It is difficult to value the ecosystem today however each piece (payments / financing / logistics / advertising) helps compound the value of each $ of GMV.

Without wasting time attempting to build a DCF, lets just focus on the five drivers that matter: (i) TAM; (ii) GMV; (iii) Take rate; (iv) EBITDA margin; and (v) Multiple.

By far the most important variable for value is the GMV, hence the long term regional macro story is the key driver for TAM while the ability of management to fend off AMZN (and others) will determine market share to arrive at GMV. We are also relying on management to continue to effectively monetize the platform to increase take rates hence a high degree of confidence in their capability is required. EBITDA margins are likely to trend towards steady state for vertically integrated eComm players which based on targeted margins for similar players may fall within the 10 – 15% range. The multiple is difficult to predict in steady state as it will likely be years if not decades before these types of companies trade based on traditional metrics (feel like we will be waiting a while for the day a P/E ratio will matter for Amazon). Regardless of when that day will arrive, the market will ascribe significant value to a dominant, profitable and cash flow generative eCommerce franchise.

From a comparative perspective MLs EV/GMV of 0.74x appears in-line with peers approximately: Alibaba 0.57x; eBay 0.48x; Amazon ex-AWS c. 0.7x

Arguably the growth runway ahead for ML is more attractive and for the purposes of an analog comparison to consider the addressable opportunity let’s compare China to Latam. China’s GDP is approximately 3x Latin America GDP. However Alibaba and JD.com combined market cap is $550bn while MercadoLibre’s market cap is $12bn (46x larger). On a relative basis looking far out into the future of MercadoLibre potentially as the leader of latam eComm the opportunity to invest today is attractive.

At $266/share, for the reasons discussed above MercadoLibre represents a compelling investment opportunity today. Don’t be mistaken though, to buy this for the reasons above is to make a long term commitment. Not a long term commitment like Kim Kardashian getting married for 72 days, a real commitment to a story that may take a decade to play out. Expect volatility in the next few years however the payoff will be immense if MercadoLibre ends up being a key player in Latin America’s eCommerce landscape.

Disclosure: I am long MercadoLibre (NASDAQ: MELI)

Sleeping Beauty

BWX SP.JPG

Up over 275% since its ASX debut… is BWX expensive? This sleeping beauty could be about to wake up as a tenbagger. 

Image result for sleeping beauty
Step 1: Good skincare regime; Step 2: Profit

Background

BWX is an Australian based owner, producer and distributor of the following personal care brands: Sukin, DermaSukin, Uspa, Edward Beale, Renew Skincare, and Mineral Fusion. It originally started out as a contract manufacturer to the skin care industry but over time transitioned to being a vertically integrated skin care company by purchasing brands. As it stands today, the third party manufacturing side of the business has shrunk as the focus of the company is solely focused on its flagship brand, Sukin. It represented 83% of revenues in FY16 and is set to grow, while it is already the No. 1 selling skin care brand in Australian pharmacies.

Brand positioning

Sukin is positioned as a “masstige” product with pricing being accessible to the mass market but offering luxury characteristics. Strong category growth of the natural skin care market has been supported by a shift in consumer preferences towards: cleaner and greener products, increased health and safety concerns, environmental concerns as well as awareness of the potential threats posed by synthetic chemicals commonly found in skincare products. This trend is here to stay as millennials drive the eco-conscious consumer mindset, but most current natural products command a material pricing premium to chemical based counterparts. Sukin sets itself apart by offering a vegan, carbon neutral, chemical free, paraben free natural product at almost the same price as standard products. This ethos is aptly captured by their motto printed on each of their bottles (which is recyclable, of course) – “Skincare that doesn’t cost the earth

Sukin range

BWX’s natural skin care focus represents advantageous category exposure. FMCG itself is seen as a defensive sector as people still brush their teeth and wash their hair in an economic downturn (one would hope), but home and personal care (“PC”) specifically is seen as an attractive segment due to higher margins. As a result PC focused stocks (Procter & Gamble / Beiersdorf) command a valuation premium to Food FMCG stocks (Nestle / Danone). Within PC, the largest market is skincare, which is expected to post the strongest market growth from CY14 – 19 (per Euromonitor estimates). Drilling down deeper, the natural segment of the market is expected to grow faster than the overall category. Sukin is well positioned to benefit.

Geographic expansion

In FY16, only 22% of Sukin sales were from overseas. So far, the story has been the domestic success however, over time we the real upside to be derived from RoW sales currently targeted at New Zealand, China, Singapore, Malaysia, UK, US and Canada. The ability for Sukin to replicate its success in these markets represents a vast opportunity with many of the addressable market sizes outstripping local potential (e.g. UK skin care market size is c. 2x Australia).

Management takes a measured approach to new market entry via independent distributors, and subsequently once established they set up a direct distribution office on the ground to service larger clients. This blueprint has been followed in the UK where previously they sold via independent distribution but in mid-2016 pivoted to direct distribution through the establishment of a sales office and warehousing after they secured shelf space with the largest health and wellness retailers in the country. While the success will ultimately be determined by the “pull” from the customer, the driver of earnings in the near term will be the ability of BWX to establish distribution partnerships to “push” the product to as many shelves as possible.

We see promising signs thus far with their entry into the UK as key retailers (e.g. Boots) have increased the number of stores stocking Sukin, with further runway available. While the market opportunity is substantial in the UK, China is a whale of a market for skincare products. Sukin’s reputation is boosted in the eyes of the Asian consumer by the positive connotations attached to the natural healthy Australian image. Just look at Chinese affinity for Swisse pills or A2 Milk to gauge the type of premium attached to Australia’s food, health or natural exports. Initial steps taken by management to establish stores on TMall and JD.com are ideal first steps to go direct to the Chinese consumer.

Strategy

On July 3rd the company announced the acquisition of Mineral Fusion (“MF”), the No. 1 natural cosmetic brand in the US. The deal was fully debt funded for a consideration of US$38.4m, and a $4.6m earn out. The market did not seem to react to the announcement yet this is one of those few times where an acquisition may actually generate a very high IRR for the acquirer, despite the premium transferred to the target. Given the guided EBITDA range for 2017 is US$3 – 4m, this deal looks pricey at first glance given the implied acquisition multiple of 11 – 14x EV / EBITDA however it is worth considering the strategic potential.

Firstly, the factors that make Sukin attractive in Australia make MF appealing in the States. This is clear as its exposure to the natural cosmetics category has helped drive a 3-year sales CAGR of 20%. Secondly, the complementary nature of the acquisition is clear as 77% of 2016 sales for MF were from cosmetics and nail care. MF core sales represent new segments for BWX that will add width and diversification to the BWX stable, allowing BWX to drive more volume through its established distribution in ANZ and abroad. Thirdly, while the US is a massive market with consumer preferences similar to other western countries due to fierce competition and difficulty with distribution it is notoriously tough to build a brand there without burning cash. The MF platform represents a ready-made US distribution arm for BWX to enter the US market, something that may prove invaluable over the long term. Finally, as the acquisition is debt funded it will increase leverage to 2x net debt / EBITDA post integration which, given the cash generation of BWX, will not stress the balance sheet. Rather the utilisation of debt headroom will drive shareholder gains through earnings accretion. Despite the nauseatingly common parroting by corporate press releases of a “match made in heaven”… this could be one of those times.

Advantages

Marketing strategy: The key differentiator of BWX to other branded PC products is its marketing spend. While maintaining similar or lower gross margins, the reason for industry leading EBITDA margins is due to its spend on sales and marketing of 10% of sales comparing favourably to international competitors at 25 – 30%. This competitive advantage is derived from the word of mouth endorsement of Sukin by women who love the product, which has and always will be the best form of marketing out there. BWX choose to utilise digital media and influencers to promote their brand rather than traditional media, a far more effective return on investment due to the increased likelihood of a young woman trying out a new product recommended by a blogger they follow. Bloggers are more likely to be perceived as a trusted source compared to just another billboard or commercial. Beauty bloggers are all powerful in this space as lead influencers, and if you want proof ask any young woman what beauty blogs they read and you will soon unearth a whole new side of the internet. Blogs upon blogs with high traffic, devoted to discussing Napoleon Perdis vs. MAC (pretend like you know) as a much more worthy match-up than Mayweather vs. McGregor. Another valuable bit of market research is to google “Sukin blog review” and read some of the bloggers reviews of the products, or give their Instagram page a visit to see the engagement they nurture with their core demographic. A far more valuable feedback tool for consumer investing than sell-side research.

Sukingoogle
Blue = “Sukin”; Red = “Mineral Fusion”

Interesting trend in search queries – 100% growth of interest for both Sukin and Mineral Fusion over the past 3 years

 

Capital light growth: Even though it is vertically integrated, BWX should be able to support further growth as it has in the past with minimal capex (1 – 2% of sales)

Management: The management team and board have extensive FMCG experience and have been around the company for a long time overseeing substantial growth thus far. Moving forward this group who have to date made all the right moves are incentivised to keep growing the brand since key senior personnel have the majority of their net worth derived from paper fortunes in BWX stock (CEO John Humble holds 10% of shares outstanding)

Risks

Customer concentration: A concerning characteristic of the Australian pharmaceutical market is the relative concentration of a few key players such as Priceline, Terry White, and Chemist Warehouse. BWX derived 50% of its FY16 sales from three customers and it is plausible that these customers may attempt to exercise market power to squeeze BWX margins or demand volume based promotional discounting as Sukin grows. This however seems unlikely while the Sukin brand remains attractive and customer loyalty is high

Competition: Currently competing with MNC and other nature focused company sub brands such as Nivea, Olay, Jurlique, Trilogy, Natio, Burts Bees. The revolution of natural products has not gone unnoticed as FMCG majors will increasingly seek to enter and dominate the sector, not ceding share to smaller names. As such, it is likely competition will increase in the “masstige” segment of the market. A natural mitigating factor to this is the personal importance of skin care to most women and the high customer retention rates once a product is established as a part of a skincare regime

Execution: The downside for BWX is that overseas consumers tastes deviate from Australian consumers and overseas growth does not pan out. Under such a scenario, domestic sales are likely to hit a ceiling over the medium term and the multiple is likely to de-rate as it could go ex-growth

Brand deterioration: While highly unlikely, a scandal attached to the Sukin brand that adversely affects its natural image or results in a recall would be a disaster for the company. Bloggers may also take a negative view of the brand as Sukin is currently not “organic certified” which some posts online have already noted

Poor execution: This risk is particularly relevant within the Chinese market. It is uncertain what proportion of domestic sales are derived from informal “diagou” sales channels, however estimates have put it at 10 – 15%. The risk of entering China too quickly via e-commerce channels may be that diagou margins are compressed and the diagou promote other Australian skin care brands over Sukin to their Chinese clientele. Another risk to the Chinese market is regulatory changes that are currently delayed indefinitely by the government are introduced whereby Sukin will be effectively locked out of direct sales into China due to requirements for products to be animal tested

Brand building expenses: New markets take increased discounting to grow distribution and higher brand building costs, compressing margins

Valuation

In terms of valuation, on first glance at the share price chart it feels as though we have missed the boat on BWX due to the stunning run up in price since the $1.50 (!!!) issuance in November 2015. This is due to the company soundly beating its prospectus forecast through outperformance of Sukin. Share price has continued to appreciate due to a raft of earnings upgrades and improving analyst commentary, but as we consider the investment today is there significant upside to the lofty ambitions the market has given the company credit for already?

Based on management guidance and consensus estimates (Annual results are due in a few weeks) on a trailing basis for FY17, BWX trades at c. 22x EV/EBITDA (assuming $50m of acquisition debt related to MF) and at 31x P/E. While the pricing on face value seems stretched and beyond the grasp of a traditional value investor, this fails to recognise the obvious growth potential and momentum of the business. On a forward basis utilising FY18 consensus, BWX trades at c. 15x EV/NTM EBITDA and 23x P/E NTM. The upside from these levels is significant as the growth levers attached to geographic expansion, exposure to underlying consumer trend shift to natural products, operating leverage and cross selling between Mineral Fusion and Sukin means earnings can continue to grow substantially off a relatively low basis for many years to come.

Traditionally, due to the mature nature and similar capital structure of many FMCG businesses most stocks within the sector trade on a P/E basis, with multiple premium reliant upon profitability and growth trajectory. Looking at the valuations of a PC peerset (JNJ; PG; OR; ULVR; CL; EL; BEI) shows trading levels of 20 – 28x forward P/E for lower growth than BWX (albeit all these conglomerates hold a much higher quality stable of diversified brands). Increasingly stretched valuations over the past few years in a post-QE world may be inflating this but considering comparables we can expect BWX to trade at an EPS multiple in the future of 20x given the higher growth profile. From a cash flow perspective, the FCF yield on the business may seem weak over the near term however; cash conversion from EBITDA to FCF is not reduced due to capex but rather NWC investment to support growth. Upside from buying at these levels is clear when you consider the very real possibility of EPS growing threefold from current levels.

Recommendation

Buy. Expensive at $5.65, but cheap when considering the growth opportunity for BWX. While expansion carries risk, the reward on offer gives us confidence to be a buyer at current levels. Will likely require patience as a multi-year expansion plays out and there will be bumps along the way.

Disclosure: I am long BWX (ASX: BWX)

iSentia, putting the story back together.

Background

Isentia Group Limited (ASX: ISD) is a media intelligence software-as-a-service provider positioned to service a range of corporate and government clients. Established in 1982 as a traditional press-clippings business, iSentia today holds a dominant position in the ANZ media monitoring market.

iSentia has its traditional standard media monitoring offerings as well as value added services (VAS) which includes social media monitoring and analysis/insight reports and other services.

Mainstream media monitoring comprises the bulk of the firm’s revenues (circa two thirds) and is generally paid for via subscription (some legacy clients pay via the volume of articles read, however this is being phased out).

For clients, media monitoring services start out with a basic ‘mainstream media alerts’ service that sends out relevant notifications via email/text once every 24 hours for a recurring fee. Customers can upgrade to online news notifications for an additional recurring fee and to the Media Portal platform (which provides live updates and is also the platform for media analysis) for another recurring fee. A range of other premiumised services is layered on top.

Analysis

Buyer beware

Do you know who weaves a great story? Roadshow bankers. Now take those bankers, have them pushed by private equity firms, and they will be able to sell oil to the Saudis. I am generally sceptical of an IPO since the reason for floating is usually not in the buyers favour (taking money off the table, couldn’t get a trade sale away, etc.), but if there is one thing to always be sceptical of, it is an IPO from a private equity vendor. Now that does not mean every PE exit is a dog, just that there is a very high likelihood the major upside has been harvested… with just enough meat on the bone to survive a lock-up period.

Special IPO gift bestowed on the market from our PE pals.

 

The story – Did we land on Mayfair?

The narrative on the back of a hot offering (offered at $2.03 with an opening pop of c. 20%) was picked up by the market feverishly. iSentia was a critical media aggregator for corporates and governments alike, with a 90% market share in Australia representing a quasi-monopoly that would levy price increases upon its customers into perpetuity. As it delivered strong ANZ earnings growth, the narrative strengthened and the market bid up the price to a market peak of $4.85 as momentum led to analysts extrapolating past growth far into the future.

Image result for monopoly man cartoon

Keep printing money.

 

Narrative break – Do not collect $200 for passing Go.

Benjamin Graham described Mr.Market as a manic-depressive who is emotional, euphoric and moody. The flip side of such euphoria is punishing when companies are priced for perfect execution. The narrative of iSentia fell apart quickly when people realised that this may not be the cash cow monopoly they initially expected. A confluence of factors hit as: (i) Meltwater, a global player with strong social media capabilities made a concerted push into Australia pressuring iSentia’s pricing; (ii) Traditional media outlets put up paywalls and extraction of value from content came into focus – leading to increased copyright costs; and (iii) King Content, the $48m acquisition into an adjacent service (content marketing) went sour with integration issues and momentum loss causing a projected EBITDA contribution to swing to a loss.

Image result for monopoly do not pass go cartoon

Monopoly man falls on hard times.

 

Where are we now?

While the outlook for iSentia is vastly different today from 2 years ago, the core of the investment case remains the same. The compelling reasons that were initially there to buy it remain:

  • High quality top line stemming from durable recurring domestic software as a service (“SaaS“) revenue
  • Diversified customer base
  • Pricing growth from upselling

Monitoring requirements for organisations in a more complex and disparate media environment are growing rather than shrinking, entrenching the product need to end users. iSentia continues its pivot to add social media capability while the continuous investment into R&D is critical, allowing for the roll out of new products that enable pricing increases and ensuring a product edge ahead of competitors.

Durability

While its cloud capable platform analysing a multitude of media sources is difficult to replicate, global competitors have similar offerings that may be portable to APAC clientele. The threat to iSentia’s moat moving forward is unlikely to materialise from a superior product offering from current competitors but instead from increased churn of larger multinational clients to a holistic solution from a global player. MNCs may prefer a single provider for worldwide media coverage, rather than the regional offering currently promoted by iSentia. Given iSentia’s average revenue per user (“ARPU”) the ability to move down-market to mid-market and smaller firms is limited due to prohibitive costs for end users, hence defence of their core large cap clients is critical. Short-term churn will likely drive price action as the market is sensitive to the threat of Meltwater eating into ISD’s client base, but it will prove difficult for Meltwater to meaningfully alter the long term market structure given iSentia’s 90% penetration. Pricing differential alone is unlikely to induce switching; only offering an advantage when pitching to end users without any current external media analytics. Rather a product edge is required – Meltwater’s core competency is social analytics, while iSentia is (perhaps late to the game) substantively beefing up their social analytics. This has been fleshed out via the addition of sentiment analysis within their core Mediaportal platform as well as completely new products focused on social such as Storyboard. A full product suite serves as the best ongoing defense to their moat, mitigating the risk of ANZ churn. The market appears to be pricing in a negative forward view of ANZ churn, as the Macquarie Australia Conference presentation catalysed a 15% jump in the stock price. The important new information to come out in that deck? Not much… apart from a line on p. 18 that “Q3 client churn returned to historic norms”. A short term stabilising of customers was enough to drive a 15% recovery in SP, hence once the overhang of Meltwater’s initial impact clears and iSentia’s market share retention is clearer to the market we can potentially expect a further recovery. Barring a multi period large scale hemorrhaging of customers, it is fair to ascribe a value to the ANZ business on a standalone basis similar to current trading levels as management continue to drive steady ARPU growth via upselling.

Growth option

The true upside to iSentia and potential for share price appreciation lies within the option value of new markets. The growth strategy communicated by management has been focused on markets with no established player and initial roll out was within Philippines, Malaysia, Thailand, Singapore and Vietnam. iSentia is well placed to capture a leading share in these nascent markets, with no incumbent as a roadblock to establishing client relationships in the region. Management estimate they only serve c. 25% of their target clients in S.E Asia, therefore there is substantial runway left for growth.

Localised players such as Wisers and Hottolink that operate in China and Japan respectively make entry more difficult in other Asian markets; however, this can be treated as a free option with even small penetration into such large addressable markets likely to move the needle. Asia is a key plank of the investment case moving forward, and iSentia have made promising early strides as well as placing the right people in the driving seat with the hire of David Liu (long time media exec, with regional experience as Aegis AsiaPac head) indicative of their appetite for growth.

What could the future look like?

To invest into iSentia means you need to take a view on the following operational drivers:

ANZ customer churn: The critical short-term data that will influence sentiment and market consensus on whether iSentia can restore its former glory. I personally model net churn to peak in FY18 at 5% with a slow deterioration of clients thereafter, as Meltwater steadily chips away

ANZ SaaS pricing: There are two choices for iSentia to battle Meltwater, retain pricing power and risk churn or retain customers and have pricing deteriorate. The latter would be the wise choice. As such, the pricing growth into perpetuity premise is broken and a hit to pricing for basic SaaS products is likely on the horizon

ANZ upselling: Consistent product innovation allows ARPU to grow as clients utilise more products, with opportunity for value added services to increase penetration within existing base. Value added services penetration within existing customer base increased c. 600bps over the two years to FY16, but it is likely this rapid growth will slow, as the largest clients with the biggest budgets are exhausted

Asia customer net adds: Difficult to gauge what the runway will be like but based on the total addressable end markets they are targeting it is reasonable that total customers for rest of the world will increase by 50% by FY25 – this doesn’t translate into strong penetration in target growth markets but is representative of promising client growth

King Content: The turnaround is underway with a new unit head for King Content and it may not turnaround, but the acquisition has a scrap value if jettisoned. Content Marketing does not have the attractive defensive qualities of the core business therefore it is difficult to forecast, however a conservative case is projected with revenues falling until FY18 and stabilising thereafter.

Content acquisition costs: The trend of content for traditional media being paywalled and becoming more expensive is likely to continue and as such, a spike in copyright costs is forecast for FY17 as a new equilibrium is found

Image result for business turnaround cartoon

King Content’s new CEO has parachuted in just in time.

 

Recommendation

Buy.

(Un)fortunately the SP has been grinding higher since I started writing this post, with $1.50 at the outset morphing into $1.90 at the time of publishing. Remains attractive at current levels – consider adding to your portfolio at any price below $2 as the risk/reward skews to the upside with my valuation implying a 25% margin of safety at $2. Entry point offers sufficient downside protection via the core ANZ SaaS revenues, which are relatively sticky. The market will react in the short term to domestic churn, but as churn and pricing impact from the heightened Meltwater competition stabilises the share price will recover. Upside to be driven by a re-rating in the stock to trade in line with SaaS peers as well as meaningful upside within the Asian growth story – an option we are happy to pick up on the cheap at current levels.

Join up to the email subscriber list (see pane to the right) if you would like to get the valuation model for iSentia to understand the valuation drivers, as well as updates for future posts.

Disclosure: I am long iSentia (ASX: ISD)