A few weeks back I wrote on the NIO Auto’s IPO on the NYSE. Yesterday we saw another F-1 filed from Chinese scooter maker NIU as they aim to raise $150M to fuel their 2019 and 2020 expansion. Yes, that’s NIO Auto and NIU scooters. Confusing, yes.
Last March I wrote about the importance of India and China landing replacements for the trusty 50cc scooter. This is important because much of the world relies on 2 wheel transportation, and the mileage logged on 2 wheels around the world rivals that logged on 4 wheels in the United States. Except, instead of clean cars that adhere to some of the toughest emission laws in the world, scooters are much, much dirtier. The pollution that comes from 2 wheels in Asia likely exceeds the pollution that comes from 4 wheels in the US.
It’s a big problem.
Electric scooters will do a lot to mitigate this, especially in countries that are focused on improving the carbon intensity of their generation.
NIU’s business is sizable already, and their growth trajectory is strong, posting impressive year-over-year gains. They are scrappy, with few employees. They have started a fairly bold expansion into Europe with a dealer network that gives them a hundred or so stores across the EU. From the north reaches of Sweden to the southern tip of Spain, there is a shop that can sell and service your NIU scooter. When they first rolled into Europe, they predicted 10K scooters sold in 2018. This summer, they confirmed with Bloomberg they had sold 4,600 scooters in 2017 but the 10K annual figure will likely elude them another year.
Volumes and ASP
From the F-1, we learn on page 10 the following details on NIU:
2017 Revenue: $116M, 2018 Revenue: $160M+
2016: 85K units, 2017: 190K units, 2018: >250K units [Page 76]
The above data points let us calculate the ASP of their scooters. For 2017, this is $610 per scooter, and for 2018 it’s $640 per scooter. This would be their average selling price to the dealer. The dealer would put another 15-20% atop this, add destination and various VATs/taxes to arrive at the consumer price.
But roughly, their average scooter likely has an on-the-road price of $850 or so, including battery. And if your aim is to replace the trusty 50cc scooter that drives much of Asia, this is where you need to be. Note, too, the slight up-tick in ASP between 2017 and 2018: NIU has some high-powered scooters in the queue that were discussed in the previous article linked above. In that article, I estimated the “N-GTX” scooter at perhaps 6 KW continuous, and the Project X at 8 KW continuous. I’ve not seen updated figures on these models, but I assume we’ll hear more from EICMA next month.
But in any case, N-GTX and the Project X, will be expensive, lower volume bikes that will also help to nudge up the ASP depending on their success. What is important here is that NIU have solved the tough problem already: A low-cost & profitable bike that can replace the 50 cc. That is what Asia needs en masse. And then they can upsell you to something more if needed.
Looking at NIUs ramp to date shows they jumped from 85K to 190K units from 2016 to 2017—an indicator of both solid execution and demand. From their partial 2018 data, we can extrapolate that they’ll hit 250K units this year. Later in the F-1 (page 28 and page 135) they talk about 2019 volumes and 2020 volumes, and it seems they are aiming to expand manufacturing capacity to 700,000 units in 2019 and then to 1M units in 2020. It’s a bit curious that the jump from 2018 to 2019 will be 3X when the jump from 2017 to 2018 was around 1.3X. But that seems common in F-1s: A promise that they company is on the edge of something very, very big. They also don’t specifically mention volume they hope to get, but instead they mention capacity they hope to build. Two very different things that might need clarification. In any case, they are dreaming big.
Gross and Net Margins
NIU claims in the F-1 that they are hitting gross margins of 14.3% (page 84) on their current product mix. Not great, but consistent with a new, well-funded business looking to get a foothold. As volumes increase and their leverage with suppliers increases and smarter employees can be attracted to increase innovation, the margins can be further improved by reducing the COGS. More important, however, is the net margin. Currently, for the first 6 months of 2018, NIU is showing 56% loss (as % of sales). But they make an important point that much of this came from stock compensation that appears mostly one-time and related to the IPO. Absent the stock compensation, the net margin improves to about 8.3% loss (as % of sales).
Very reasonable at this stage.
Perhaps most interesting in learning about NIU is the employee breakdown:
348 Full time employees (page 143)
178 in sales and marketing
61 in R&D
109 in supply chain and G&A.
These numbers obvious don’t include manufacturing headcount. On page 27, they indicate that all manufacturing employees are provided by an outside labor service and the manufacturing headcount is then located in the cost of revenue line.
We know from their income statement on page 10 they spent $6M in R&D in 2017, and look to be on track to spend $8.5M in 1H2018. But remember, those figures include share-based compensation of $6M (page 11). If we back out 50% of that (assuming a big pre-IPO incentive to lock-in employees), that puts 1H2018 R&D around $5.5M, or $11M for all of 2018. Assuming the loaded cost of a R&D engineer is $35K/year in China, we can spitball that in 2018 NIU spent about $2M on R&D headcount and 9M on product development (prototypes, tooling, NRE, certifications, testing, field trials, etc). They probably had at least 3 programs underdevelopment in 2017 (Project X, N-GTX and a next-gen bread and butter platform), so you can estimate the all-up cost of developing a new platform (including headcount) costs NIU $3.5M to $7M since it might take 1-2 years to go from rough specs to shipping product.
That is the power of developing in China. That they are doing this with 61 R&D engineers points to them doing minimal in-house development and relying more on suppliers to deliver core technology. Which is fine as long as you can negotiate the supplier prices you need to be competitive. If you cannot, you must pull in house. But based on margins they seem in a reasonable place right now.
Overall, the R&D headcount seems a bit light for 3-4 platforms, each with its own pack, motor, controller and chassis. Even if the motor and controller are coming from an outside vendor. But hey, they appear to be doing it.
Currently, NIU is generating more than $160M in sales from 348 full time employees. That’s about $460K per employee. That’s around where Tesla is today, but fair bit worse than Honda and Toyota, and a lot worse than Ford.
Customer Acquisition Cost
In 2017, NIU spent about $12M on the sales and marketing of 190K scooters with 178 employees. That’s about $63 per scooter, or roughly 10% of the scooter cost. Let’s swag the loaded cost of a marketing person is $25K/year in China. If true, about 1/3 of their marketing is on headcount and 2/3 is on ads, incentives, etc. Overall, this number seems pretty reasonable. In the digital age, hopefully the headcount grows more slowly while the remainder likely tracks volume.
The important point here is that they aren’t overpaying to get a customer.
Model Breadth and Depth
The F-1 indicated that their N-series (high end in the current portfolio) is about 41% of all sales, with the other two platforms (M and U, with U being their lowest end) split equally and making up the remainder.
They also have multiple tiers for each platform (mostly motor and battery options), as well as each country. They are doing a lot of product tiering with very few R&D employees. This suggests a thoughtful underlying architecture—SW and HW—is in place.
All stores in China are franchised. In China, there are 571 stores. And as noted above, in the EU, there appears to be another 100 or so dealers that are selling NIU scooters.
This is inline with their manufacturing strategy: Fewer employees on the books.
NIU sees global market opportunity of electric two-wheeled vehicles (for all manufacturers) shaping up as follows for 2020/E (page 110):
China will be 9.9M scooters and $4.9B revenue. This is $494 ASP
SE Asia will be 4.6M scooters and $1.5B revenue. This is $326 ASP.
India will be 152K units and and $74.1M revenue. This is $487 ASP.
EU will be 3.2M units and EU6.3B = $7.3B USD revenue. This is $2.3K ASP.
Their figure on India seems out of whack.
Note that outside of the EU, the world is needing ultra-cheap. At ASPs below $600 or so, these are closer to powered bicycles than scooters. Yes, the world demand for two-wheel is sky high. But don’t confuse the normal gas scooter you might see today in the US or EU with what the world can afford. They are two very, very different beasts.
NIU has done the right thing here and solved the harder problem first: They’ve started with a lower-power platform, added some digital bells-and-whistles that those platforms never enjoyed before, and they proved the offering at scale. And now, they will go build a higher end product.
The flip side is to build high-end and drive down. It’s usually easier to teach a company culture that began as “scrappy” to become sophisticated rather than vice versa. Look at the issues Tesla is having making a $35,000 car. It is nearly killing them.
Investments to Date
It looks like Series A-1 raised $16.1M, series A-2 raised $6M, series A-3 raised $10.4M, Series B raised $25.5M, All up, this is about $58M raised—the sum of what NIU calls the “total proceeds” for each round. Crunchbase indicates NIU has received $45.5M total in funding.
This is hard to tease out because there are a few things that aren’t clear to me and/or are omitted (left blank) in the F-1 at this stage (though the F-1 will be updated as these points become known).
NIU looks to have assigned a FMV of $2.62 (discounted as described on page 103) to their shares, and there looks to be 135M shares outstanding (page 161). This points to a current valuation of $350M plus the discount. Maybe this rises to $450M or so once discounts are removed.
But as a snapshot in time, you have a business today that is operating at scale and generating about $160M in revenue from a total raise of just $58M with a valuation they believe is maybe worth 3X of sales (a 3X price to sales or PS ratio). For comparison, consider the PS ratio of other businesses:
SYM Motorcycle: 0.53
Bombardier Recreation Products: 1.1
Harley Davidson: 1.75
The bet here is that you are buying the stock today at a premium because you believe it will rise to a point in the future where it’s steady-state business metrics will match those of similar businesses. For example, when Tesla was brand new, investors believed a 15X PS multiplier was reasonable because of the growth. But as Tesla has delivered that growth and slowed projections, the PS falls and will approach the PS ratio of established companies. Long-term, you need only look at Honda and GM to see how lousy their PS ratios become.
If NIU optimistically delivers 1M scooters in 2020 at an ASP of $900, the that’s $900M in revenue. If their PS is closer to Harley Davidson (rather than Tesla), then the valuation might be around $1.8B. That is, sell 1M scooters at $900 a pop and get a $1.8B valuation. That’s roughly a 4X pop over where NIU values the business today. There’s upside here for smart scooter makers for sure.
But it isn’t limitless. Remember, Mazda’s market cap is $7.4B for selling nearly 300,000 good cars per year. It’s tough. Long term, the market caps on these EV businesses will align with their petrol counterparts. It’s inevitable.
Update 10-Oct-2018: NIU has set the terms for the IPO, reported on here. They plan to raise $95M from an 8.3M share offering at a price range of $10.50 to $12.50. Assuming an $11.50 valuation, this would put the market cap at $900M.
Update 19-Oct-2018: Last night about 12 hours before IPO, NIU reduced the number of ADS shares down to 7M, and lowered the target price to $9. The stock was trading below the $9 open for much of the morning. the overall raise has been scaled back considerably: From $150M a few weeks back to $63M today. The final F-1 indicates each ADS represent two Class A shares, they are offering 7M ADS. Thus, at $9 offer, that is $4.50 per class A share, with 135M shares outstanding (see above). That’s a market cap of about $607M on 160M in revenue (3.8X PS).