Do we DiDi?
Wow – we are halfway done with the year! Hope your portfolios held up nicely despite the volatility we saw in March and May. The last six months saw a lot of activity with new listings, including our topic of discussion today - DiDi.
Yes, the Chinese ride-hailing colossus went public this last week, and despite an early sell off, the stock price has rebounded in the second day of trading to about 17% above IPO price (this blog was written on Friday, 2nd July, Asia time). Since we had commented on Ant Financial before its failed IPO as well as given you our thoughts on Grab’s SPAC merger, we couldn’t help but also share our thoughts on Didi.
Now, as with anything with China, I will caveat. We don’t do any single-stock investing in China, as for now, we still find it out of our circle of competence. There are at least 1.4 billion people who know more about the country than we do. So, take our analysis below with a pinch of salt. Also, the basis of our analysis is DiDi’s S-1 filed with the SEC, so we are just going by what’s presented in that document.
Ok – let’s get to it.
In case you’ve been living under a rock, Xiaoju Kuaizhi, or better known as DiDi (“Beep Beep” in mandarin – like a car horn), is the largest ride-hailing operator in China, with a 90%+ market share in China. To be fair this figure is dated, and we haven’t seen any updated figures in the S1 (we’ll come back to why we think this is) but rest assured that it’s a monster in the ride-hailing space in China. It not only offers 4-wheel transport, but also offers shared bike services, food delivery (outside China), community group buying, financial service, etc. That said, its core-transport business makes up 95% of revenues, so any “other business” is more of a side show for now.
The Good:
Sheer Scale:
We’ll start the discussion about what we were really impressed by reading the S1. For one, it’s a very young business, started just 9 years ago, and despite that, it's worth over $70bn. It just proves how well the management team had to have executed to reach its dominant share during a time when DiDi was competing against other local startups and an aggressive Uber (here’s a great article about how they beat Uber). To give you the sheer scale of their size, let's look at some stats (just like Ant Financial, the hyperscale of some Chinese companies never ceases to amaze me).
As you can see from the above image, the company has had 493 million active users in the past 12 months to March 21. This is a huge number, and if we’re not wrong would be larger than the active users of Uber, Lyft, and Grab combined. This same user base of 493 million, spent RMB 341 billion in GMV during the last year. That implies that the average active users spent nearly 700 RMB/year with the company, that’s the equivalent of $9/month! This is essentially a subscription business 😊.
Improving Margins:
The one benefit of scale is that you can improve margins overtime. We can see that back in 2018, the company had very thin growth margins (note that Didi presents revenue as cash received before they payout drivers). The improvement in margins show that DiDi is enjoying some benefits of scale and have started to reduce the incentives they pay both to drivers and riders.
The Missing:
Ah, yes, you noticed. The last section was quite short. Well, that’s because there wasn’t much in the S-1. No details about market share, no talk about frequency of use, not discussion on future potential. Although on the last bit we did read online that management had said they would grow revenues comfortably at 20-30% over the next 20-30 years, which would at the extreme mean that they would have revenues of over US$57 trillion, which makes us a bit sceptical (hope the sarcasm translates well).
But compared to Grab’s filings earlier this year, which had several interesting slides and figures to digest, the DiDi’s filling was very vanilla. Some talk about a flywheel, some talk about the future of EV and self-driving vehicles, but other than that not much at all. Even the figures they presented only went back to 2018, which made us struggle to get a sense of growth because all we have to go on was 2019 vs 2018 (which was only a 14% growth in revenue) and then 2020, which was a covid year. Now the company has stated that international growth has been at a rapid pace (around 60% CAGR), but it is still a tiny part of the pie.
As stated previously, we were quite surprised that the company did not talk about market share. We think there could be two reasons for this. The obvious one is to not suffer the ire of the CCP who upon seeing boasts of an 80-90% market share, might force the company into a corporate struggle session for being far too dominant. This is not far fetched as we’ve seen what’s happened with Chinese companies lately who have gotten on the bad side of the CCP, most famously Alibaba.
The second reason, which is less obvious, is that it could be that the company is losing market share. There are a few data points that has led us to think this. For one, we have read a few quotes online, like this one from Zhang Xiang, an auto industry analyst with 21st Century Business Herald[1]:
"We have reached the second peak of the ride-hailing industry. The first peak was marked by a battle of subsidies, which DiDi [won], securing the market lead. But the second peak is marked by DiDi's market share rapidly falling and being taken away by competitors. New players like Meituan, AutoNavi, and internet giants and hardware companies are entering the industry.
Second, ShawSpring Partners’ Dennis Hong, posted this image on Twitter.
In this image you can see a user price-comparing several ride-hailing apps using an aggregator. This implies that the ‘moat’ for any one company is quite thin, and that there is little loyalty. Now we don’t know what percentage of users in China use this ‘metasearch’ route, but if the Chinese are as kiasu as us in Singapore, I would imagine it's quite large. A parallel to this is what happened to Online Travel Agents once the meta business evolved. This meta or “aggregator” business made it much harder for OTAs to acquire customers. Though, in our opinion this is much worse turn of events for the ride-hailing industry as there is not much of a commission buffer with which to play around with prices. This commission/incentive buffer is a tool that OTAs still use to acquire customer by offering cheaper prices than any other competitor (this is possible as the supply base for OTAs is not nearly has homogenous as the supply base of ride-hailing apps).
If we do some extrapolation, we would still guess that as of now the market share is still high. The 2019 and 2020 GTV numbers, which were around RMB 225 bn and 215 bn respectively, compare well to the market size numbers they presented later in the S-1 of 247bn and 223bn. There is a bit of a mismatch here as the GTV numbers include bike-sharing and international segments where as the market size numbers is just China shared mobility numbers, but still, it still seems like Didi has a super-dominant share. That said, none of that means much if the moat is thin. Because if the moat is thin, then the company will have to spend even more on marketing and operations to maintain their market share, which brings us to the next point.
The Bad:
While our thoughts on DiDi losing market share sound bad, they are just conjecture at this point. However, as we went through the numbers, something more glaring jumped out at us. Remember that gross margin chart we posted above? Well here’s the rest of it...
As you can see the operating losses are significant, and as a percentage of revenues, they are not coming down. This implies that marketing and other operating costs are not participating in any sort of operating leverage. Again, it's not the losses per say that are troubling, its that the losses are not coming down as a percentage of revenue. So for every $1 that DiDi earns, they are spending it, and then some. Now one could argue that this metric was improving in 2019, but then declined as covid hit in 2020. However if this is true, it is then unclear why Q12021 losses were so large considering that China had emerged from lockdown several months prior. Also it seems that Q12021 transactions fell 12% from the prior quarter (although there could be some seasonality there - but again, we don’t know because their data doesn’t go back far enough!).
Valuation:
When the DiDi IPO was being discussed, we saw a lot of lazy analysis on FinTwit which compared Price to Sales ratios of various ride-hailing companies and deemed DiDi a cheap buy. However, a quick read of the S-1 shows that DiDi presents revenue before they’ve paid out their drivers, which is a bit misleading. Thus, the best way to compare valuation would be using gross profits. Different companies present figures in different ways, but typically gross profits are after incentives and driver fees, credit card processing, and other such cost of revenue.
Do note that the above is just a quick calculation (done as of 1st July close prices), and there are few things to bear in mind. For one, we’ve just annualized Q1 numbers for each company to come up with the 2021 number, except for Grab as they had projected out their 2021 number. Grab's number is a ‘net revenue’ number which if we’re reading their presentation right should be equivalent to Gross Profit (but if we’re wrong then their multiple would also probably be in the 20s). On the surface, based on the above, the DiDi valuation is neither cheap nor expensive as compared to the group, its ok lah.
Conclusions
I think our reading of DiDi’s S-1 left us with more questions than answers. Why aren’t operating margins improving? Are they losing market share? Will their super-app strategy have any effect (95% of revenue is still transport focus)? Will the super-app strategy even work if their moat degrades?
Given all these questions, even if we did invest in China, I think we’d have to pass on this one until we received clarity on the above.
We hope this gave you a sliver of insight into the DiDi IPO. Thanks for reading, and as usual happy investing!
[1] https://www.protocol.com/china/didi-ipo-everything-to-know