|Shares Out. (in M):||300||P/E||0||0|
|Market Cap (in $M):||12,900||P/FCF||0||0|
|Net Debt (in $M):||-3,100||EBIT||0||0|
|Borrow Cost:||General Collateral|
I think Lyft is an interesting short. It's (obviously) hard to come up with precise estimates of LT GMV/revs/margins here but it feels as though there are sufficient datapoints and anecdotes about this company & space that all point in the direction of this being a worse business than the market continues to assume and downside to the current share price as the bulls' dream continues to fade away.
The datapoints / anecdotes are as follows -
1) Recent trends are completely inconsistent with anything like the TAMs that the company discusses.
2) Third-party in-city data also suggests that penetration is plateauing.
3) Cohort trends (from S1) also suggest that penetration is plateauting as well as a lack of network effects.
4) Google's insertion between company A & customer rarely (never?) ends well for company A.
5) Uber's European & LatAm experience highlights the instability of profitable market structures.
6) The magnitude of losses at present, combined with the current growth trends suggest that true (not Adjusted EBITDA) profitability is actually a long way off.
7) The fact that the co-founders of Lyft together own less than 1% of the company is instructive as to how capital/cost-intensive this model really is.
8) The fact that this company doesn't disclose bookings or rides is extremely worrying, especially given their ability to manipulate revenue.
9) Travis K. blasting out of his Uber stock as soon as the lock-up expired, down 35%+ from IPO is hardly a vote of confidence in these valuations / the business model (particuarly as everyone on the street seems to think that Uber is "cheap vs Lyft").
Unpacking each of these in turn -
#1 Last Q active riders grew by +2.3% Q/Q. In the year ago Q3, that number was +12.6%. Q4 tends to be seasonally very weak (given weather) - with an average sequential decceleration in active rider growth of ~ 8% over the last two years. I wouldn't be surprised if active riders shrink Q/Q this year.
If we look at the sell-side's best guess of GMV (let's assume they've had some steers from the company and the average of their numbers shouldn't be too far off) this is the Y/Y growth rate we've observed in GMV for the last eight quarters -
+136% / +102% / +88% / +69% / +58% / +60% / +52% / +45%
That's a pretty drastic deceleration and one that to me is highly inconsistent with the idea that there is a monster TAM out there waiting to "adopt" ridesharing. Remember, during this last two year period, Lyft has taken meaningful market share in the US. The underlying growth rate for the US rideshare market as a whole is far far below these numbers.
My estimates suggest that Lyft's pace of market share gains has steadily slowed over the last few quarters. They will probably continue to take share in future but likely at a decelerating pace and at the very least on a larger base (+10% on 30% is a meaningfully lower % growth uplift than +10% on 20%).
#2 The third party data suggests that customer adoption of these products is much deeper than the bulls / massive TAM estimates would have you believe - I'm yanking this chart from the Q&A for the prior Lyft thread as to me this is a great chart that shows just how stagnant rideshare now is in the bulk of the big US cities -
#3 Lyft put some cohort trends in their S1 which paint an ugly picture to me. Their 2015 cohort grew their rides +92% in 2016, then +24% in 2017 and then +12% in 2018. The 2016 cohort grew their rides +64% in 2017 and then +3%. The 2017 cohort grew their rides +44% in 2018. The Y+1 growth has gone from +92% to +64% to +44% over the last three years. I think it's Bill Gurley who is most often quoted as saying that if you have a busniess benefitting from network effects business, the clearest evidence is is that as you get more users, the product gets better so the new users are even more engaged at any given point in their experience than the earlier users were. That does not seem to be what's happening here, rather the oppposite.
It's also worth pointing out the fact that the 2015 cohort only grew +12% in 2018 and the 2016 cohort only grew +3% (!) in 2018 is another set of datapoints to support the contention in points #1 & #2 - that we are way more mature/penetrated with these products than these companies admit.
#4 Google (via Maps) is extremely well-placed to insert itself between riders & ridesharing networks - it helps the consumer compare prices and wait times far more seamlessly than by opening both apps at once. This feels like a nascent effort on Google's part which is only likely to accelerate over time as either they just want more $s or they want to start asserting a greater role for aggregation and search within ridesharing to better prepare consumers for a standalone Waymo offering at some point down the line.
#5 I think there's a lot to be learned from Uber's European experience. If you go back a couple of years, Uber was a monopolist in many major European cities. The product was good (cheap prices, low wait times, clean cars, reasonably friendly drivers etc etc) and it was the only option. Today none of that is true. There are multiple start-ups seemingly having a real impact on Uber's market share in many major European cities. In addition (maybe partly as a result) the quality of the Uber product has deteriorated meaningfully - the prices have risen, the wait times have extended, the driver cancellation frequency has risen, the cars look awful, the drivers are miserable.
The reality is that it's just not that hard to raise $20m today to buy some off-the-shelf mapping/routing software, run a bunch of billboard ads, fund a referral program and pay drivers well-enough that you can bring some over to your platform. To give a concrete example - Bolt launched in London in June 2019 and now claims to have ~ 40% of Uber's drivers on their platform and ~ 15% market share of rides - all within a few months.
#6 It's worth thinking about just how much the top-line needs to grow to cover Lyft's cost base -
If you look at Lyft's adjusted OpEx bill last quarter, it was $700m. Then they guide to $250m of run-rate SBC which wasn't included in that number, so the real number is $950m per Q / $3.8b per year.
Last quarter Lyft generated annualized gross profit of just under $2b (doing 50% gross margins).
If we give them credit for GMs getting up to 60% over time (they've said they're targetting 70% LT, so I've split the difference between that and where they are now) that implies that - given the current cost base - they need bookings to rise +60% vs last quarter's run-rate just to breakeven. Remember, (a) last quarter active riders grew by +2% and this quarter may well shrink and (b) in 2018, if you aggregate the 2015 & 2016 cohorts of riders, their rides grew by just 6% YoY in 2018.
If they're not able to generate any of that promised GM expansion and they remain at the current ~ 50% level, then they would need bookings to rise +90% vs last quarter's run-rate just to breakeven at the current OpEx base.
If there's any OpEx inflation over time then those required GMV growth numbers will only go higher.
#7 It's not dispositive in it's own right but the fact that Zimmer & Green together own less than 1% of this company is quite amazing to me. It's just a very real sign of just how much capital they've had to burn (and dilution they've had to accept) to reach this point in time.
#8 Again, not dispositive and I'm sure they'll hide behind the curtain of "competitive threats" but the fact that they stopped disclosing core KPIs in their first quarter out of the gate does not inspire confidence in either the business or how management plans to treat shareholders.
#9 Again, not dispositive, but I do think it's pretty interesting that Travis is blasting out of Uber as soon as his lock-up ended.
If you put it all together, I think the future looks fairly bleak for this company. Below I try to set out some VERY hypothetical calculations around potential end-state profitability for the company to help frame the question of the extent to which the market is already discounting it -
Booking.com converts ~ 5.5% of GMV into EBIT. That's a pheonomenally run company with a hugely dominant position in Europe where the marketplace benefits from heterogeneity of product (as opposed to rides which are basically all the same). Expedia's core OTA looks to me to be converting ~ 1.1% of GMV into EBIT.
If we take consensus GMV for five years out (so FY23) it's $23b. That's roughly a 2x vs the FY19 consensus number. That seems aggressive to me given all the above points I've made around signs of market saturation/maturity, but it's only a ~ 18% CAGR which doesn't sound totally crazy either so let's stick with it for now.
If we assume they make Booking.com's 5.5% EBIT margins on that, that would equate to $1.25b or ~ $1b of NOPAT. If they make Expedia's 1.1% margins on that, it would be more like $250m of EBIT or ~ $200m of NOPAT.
If we assume their losses gradually moderate between now and then, they will probably burn $2.5b to $3b to get to that point. They have just over $3b of cash today (and no debt) so I think it's reasonable to assume they use most of that up over the course of the next 3-4 years.
Let's say the stock is worth 18x our FY23 NOPAT number (Booking.com today is on just over 16x FY20 P/E and grew room nights 11% Y/Y last quarter).
That multiple would translate to an EV of $18b in the Booking.com case or an EV of $4.5b in the Expedia case. If you discount those back to today at 10%, you get $13.3b or $2.6. Lyft's market cap today is just under $13b which suggests that despite what feels like an increasing lack of faith in the LT UEs of these companies, the stock is still fairly optimistically valued today and so the risk/reward skew to being short remains favourable.
I'm obviously framing this fairly narrowly with my Booking/Expedia comparisons and am aware that the true range of outcomes here is probably far far wider. I'm sure people will point out myriad differences in the business models & UEs of ridesharing vs OTAs. That all said, I still think it's a useful exercise.
Catalysts are overrated.