08 August, 2015

Uber (Part 2b) - A framework for analysis & discussion

A long-form "Uber" framework for analysis & discussion:

Below I offer a 5-point discussion framework in support of my view on Uber, developed while teaching a recent class on business strategy.  Hopefully it offers sufficient footholds for readers to challenge my facts, offer contrasting views on important points, and introduce ideas I haven't considered.  I'll continue to expand on several sub-bullet points in a series of follow-on posts over the upcoming weeks.

  1. Uber is large and growing quickly...
  2. ... and needs money to fund continued growth into "new" cities
  3. External fundraising will prove increasingly difficult because Uber's current valuation already prices in a "win" of much more than the existing global taxi market
  4. Internal funding by "mature" cities is likely difficult due to low current cash flow and possible margin erosion; the high capital costs of responding to any erosion increase the challenge further.  Counterintuitively, embracing drivers as employees may help
  5. Uber risks ending up more like Groupon than Amazon.  Uber may experience a value collapse if not outright failure; but in the process will fundamentally reshape local transit around the world

Apologies for the klunky list formatting, I'm still getting the hang of blogger and it doesn't seem to format nested bullets or alternate naming conventions (e.g. 1.a.iv) the way I'd prefer.  This is a deliberately long post so the entire discussion framework sits in one location.  As with the other posts in this series, this post is narrating public-record third-party reporting on Uber and should not be construed as investment (or any other kind of) advice

1.  Uber has raised a lot of money, is a large company, and is growing quickly

Raised a lot of money:

Large and growing quickly:
  • 1.g)... By the end of 2015 Uber expects to be at $10b annual *ride* revenue (run-rate, or $833m/month)
  • 1.h)... Spring 2015 monthly *ride* revenue was $415m/month (half of the end-year goal) and growing at a reported 300%/year.  This seems consistent with reports and expectations of prior growth.

2.  Uber must spend cash/lose money during growth

Uber's burn rate was a bit more than half a billion per month in Spring 2015 (calc's from #1):
  • 2.a)... Net monthly platform revenue in Spring 2015 was $83m/month ($415m/month in gross ride revenue at the reported platform margin to Uber of 20%)
  • 2.b)... Uber's monthly operating losses during Spring 2015 was reported as ($470m)
  • 2.c)... As a first-order approximation, Uber's operating cash flow was therefore 83-(-470)=-553
  • 2.d)... Actual negative cash flow is possibly higher, to the extent Uber is using cash to invest in long-term assets (such as their autonomous vehicle center [Update 8/12: R&D expenses are generally expensed under US GAAP, but capitalized under IFRS] , or acquisitions such as their rumored $3b bid for Nokia Oyj's HERE maps business)
  • 2.e)... Actual negative cash flow is possibly lower, to the extent that depreciation expenses are represented in the operating loss number.  This is unlikely to net negative of their asset investments given Uber's age and growth, so the actual negative cash flow likely exceeds ($553m)
  • 2.f)... Most  [8/12 Much] of the negative cash flow is likely due to "one-time" expenses on legal, lobbying, and driver/passenger incentives (explained more below) and is likely not representative of the underlying run-rate economics.  How much more "one-time" expenses remain, and whether the benefits accrue just to Uber or spillover to its competitors, are very real questions discussed at length below

Uber's continued need for external cash to finance growth depends in a large part on unpublished financial information:
  • 2.g)... A considerable portion of the operating costs is likely being used for one-time expenses to "open" new cities and grow Uber's business:  Incentives for drivers and riders.  Legal and lobbying expenses.  The duration  and magnitude of these expenses are very important.  My belief is Uber is absorbing most/all of the industry cost of introducing this new business model to riders, drivers, and governments.  To the extent their lobby is successful, Uber changes the rules for itself and all its competitors (unless it manages to re-regulate behind itself and lock out competition).  Uber currently experiences heavy legal/regulatory expenses at all levels of maturity: home markets like San Francisco and NYC, growing markets like France and Spain, and emerging markets like China.  (click through that last link to the FT's leaked China note from Kalanick to investors which lays out the tactical street-fighting nature of these expansion costs)
  • 2.h)... Steady-state direct operating costs are likely quite low; Uber is basically a double-sided market-making consumer-focused SaaS app.  App development and maintenance costs are likely trivial at any real scale.  Customer service costs exist, though Uber just announced their second center in Hyderabaad and it's "only" a $50m investment as part of a larger $1b commitment to India  (Note: lightweight operating costs are good for Uber, but also for any competitors riding their coattails of deregulation.  More on this later)

The basic growth model question of Uber's (current) business is to what degree can cash flow from Uber's "mature" cities fund their growth in "new" cities, and at what cost will they need to finance that growth externally through further debt and equity.  Important considerations include:
  • 2.i)... Free cash flow from "mature" cities
  • 2.j)... Investment & time horizon required to open each "new" city
  • 2.k)... Timeline of opening "new" cities
  • 2.l)... Ratio of "new" to "mature" cities
  • 2.m)... Investment in core and non-core business experiments (e.g. UberPool, autonomous vehicles)
  • 2.n)... Ability to raise external financing

3. To justify Uber's *current* valuation, they need to "win" roughly half the *current* global taxi market (or its equivalent in adjacent/"new" markets) while sustaining its *current* platform margins.  Future fundraising requirements may raise the valuation faster than Uber closes the gap.

Note: This quick-and-dirty revenue multiple valuation is a shortcut so we can discuss the strategic imperatives demanded by a $50b valuation.  It also lets us wave our hands and ignore momentarily both the unknown operating expenses discussed above in #2 as well as ignore any number of distracting things like revenue/pricing strategies, carpooling/virtual bus routes, etc  which are quite dynamic even within a given geography.  

More detail on origins/correctness of revenue multiple valuations of software companies will be provided in a later post.  For an expert's future-cash-flow valuation of Uber, see Aswath Damodaran here or more recently here.
  • 3.a)... Global taxi market is estimated by Prof. Damodaran as about $100b, growing to $180b by 2025
  • 3.b)... Software company valuations as a multiple of revenue decline as they mature, often to 4x-6x in high operating margin businesses (like SaaS) or as low as 1-2x for low-margin/high cost of service businesses.  On 8 Aug 2015 the EV/revenue ratio of Facebook is ~19x, LinkedIn is ~10x, Google is ~5x, and SAP is ~4.5x
  • 3.c)... If Uber "won" half the global taxi market before they matured, with 20% platform margins, and a 5x multiple on their platform revenue, they could end up worth $50b (100*50%*20%*5=50).  
  • 3.d)... $50b/year in ride revenue is about 20x [10x] Uber's reported size in Spring 2015; 10x [5x] Uber's planned size at the end of 2015
  • 3.e)... How much additional external funding will Uber require to earn into their current valuation?
  • 3.f)... How much farther will this raise the required valuation?  
  • 3.g)... Is the the ultimate market for Uber big enough that it can "win" before it runs out of market to grow into?
  • 3.h)... What type of downside protection will late-stage investors demand/have they already demanded? (i.e. in a cash crunch, how diluted will earlier investors/employees be?).  NY Times gives broad overview of this issue (here); Fenwick & West provide a more detailed study (here).  TechCrunch reports employees are asking cap table questions much more frequently (here).  Among my possible future posts is a breakdown of the implications of Uber's Spring 2015 convertible debt

4. Uber needs to "win" a big enough market while keeping their platform margin intact.  The local transportation market is definitely big; however, the dynamics of "winning" a city are unclear and competition among competing platforms is likely.  Uber's bigger risk is margin erosion, not market size.  To protect platform margins, Uber either needs to push re-regulation ahead of their competition and/or eliminate drivers.  It will be challenging to do this in a way that is not highly capital intensive.  Counterintuitively, embracing drivers as employees may help

The total market is quite large, and Uber may well win a significant portion of it.
  • 4.b)... Note that many of these Uber rides seem unlikely to be people who stayed home before Uber (ie induced demand); many likely are shifted from other modes which exert some degree of countervailing market pressure - some taxis, some personal autos and some public transit.  Note: This bullet will be the subject of several future posts in this series.  One post is on the dynamics of (public) services which use profitable routes to subsidize unprofitable routes and the death-spiral economics of cherry-picking competitors.  Another post is on the asymptotic limits of "sharing" benefits: limits to the wider spreading of fixed costs and the dynamics of "time-" vs "mileage-" based depreciation.  Conventional wisdom in the "sharing economy" for autos likely over-estimates the savings potential directly attributable to underutilized autos; the savings will come mostly from reduced operating externality costs which only exist in sufficiently dense locations.  These externalities - such as parking, congestion, etc - are not being fully borne by drivers today in most big cities.  This will place limits on cost-shifting and profitability considerations as Uber is considered as a replacement for personal autos.  While "green" benefits also exist, they stem mainly from an assumption that ride-sharing autos will be "right-sized" (smaller cars); benefits from reduced vehicle mileage traveled is quite small.
  • 4.c)... As an aside, cities and other governments should be prepared for a decline in the usage of high-cost services which are used to subsidize other parts of their budgets (some parking, high-density bus routes, etc).  This is a similar dynamic as USPS/UPS/FedEx and public schools/charters.  Perhaps a future post on this topic.

Competitors are being launched frequently, and the cost of entry is quite low
  • 4.g)...Can Uber "win" a city and "keep" that win?  (i.e. What are the dynamics in an uncapped on-demand ride market?)  This may be the subject of future posts.  I believe that network effects/minimum density of drivers/passengers are low enough thresholds that large cities would likely support two or more platforms in equilibrium (or a monopoly provider prevented from earning rents under constant threat of limited market entry).  Those platforms will find it challenging to exert pricing power either on fares or on drivers.  Business models will converge to a city-specific mix of point-to-point, carpooling, and virtual busing depending on a variety of relative cost factors and the shape of local demand.  Interesting reading on the dynamics (pricing, structure, incentives) of different types of multi-sided markets here, here, herehere, and here.  In such a competitive environment, platforms could benefit from providing strategic misinformation to riders/drivers.  There's been at least one round of price wars in North America already, driving Hailo back to Europe; Uber and Lyft may  temporarily reduced their platform margins to lessen the impact of this price war on drivers
  • 4.h)... Suburban areas may not have enough rider density to support more than one winner, if any, but anyway this is not where the money is.  Parking is approximately free at both origin/destination, substantially reducing the cost/hassle savings of Uber.  Driving distances are longer, "empty" re-positioning trips are longer due to lower trip density, and wait times longer.  Commuters need to use their cars at the same time, though limited ride-sharing could help especially if the system encouraged dynamic transfers (multiple rides to get to work).  These all increase the direct cost/hassle of Uber and will limit its growth outside cities.

Protecting margins is likely more important than growth
  • 4.i)... If cities are single "winner takes all" markets, then it's likely that Uber would be able to exert pricing power on drivers through higher platform margins.  The first open question under "winner takes all" would be how much/long would Uber need to spend to "win" in a given city and whether it could internally/externally raise that money quickly enough.  There's an absolute limit to the platform margin that depends on something like minimum wage, cost of owning/maintaining an auto, and utilization/ride search - I'd be surprised if the platform margin could exceed 40% except in the case of dynamic bus routes
  • 4.j)... A second open question under "winner takes all" would be keeping that win - at 20% Uber's platform revenue in San Francisco is already $100m annually.  This is of course reduced by any driver incentives.  Will Uber be able to wean its drivers off incentives?  How long before an enterprising group of independent drivers launches an autonomous collective which mirrors Uber's pricing and attempts to distribute the gains of that profit pool between drivers and passengers?  How about a well-funded competitor (or a state entity on a jobs-for-benefits kick)?  Lyft co-exists with Uber in many cities, including San Francisco.  Some drivers stay logged into both apps and take the first available fare.  
  • 4.k)... Perhaps Uber could co-exist with one or several other competitors.  If network effects give Uber a sustainable driver utilization advantage, it could translate into higher platform margins.  UberX platform margin is currently 20%.  Uber may be attempting to raise the margin to 25% for new drivers.  If platform margins increased as far as 40%, Uber would need only to win $25b in global rides to earn their current valuation - two and a half times their expected end-of-2015 run-rate.  Higher margins would also substantially increase internal cash flow, better funding the expansion
  • 4.l)... A sustainable driver utilization advantage seems difficult as long as drivers can "wait" on multiple apps: the effective driver density for competing apps becomes equivalent.  As long as there's a chance a customer might use "EZ Ride app" at acceptable economics, a driver could stay logged into both.  Direct attempts to prevent drivers from waiting on multiple apps would weaken Uber's independent contractor argument; indirect attempts (such as capping the number of active drivers in an area, setting minimum "availability" lengths and booting drivers who cycle their availability too frequently or don't accept ride requests within a certain time, etc) might be possible without reclassifying drivers
  • 4.m)... So long as competing apps need to (mostly) match ride fare rates, and driver density is equivalent, it seems like platforms will eventually compete explicitly on rates (directly or through continued "incentive" payments) If platform margins eroded as far as 5%, Uber would need to win $200b in global rides (a quarter of our "expanded" market) to earn their *current* valuation.  Uber would also need to finance more expansion costs instead of relying on cash flow from "mature" cities (worsening the issues discussed in #3 above).
  • 4.n)... The tactical reality is obviously much more complicated than 4(i) through 4(m) above.  Even without the one-off driver incentives frequently in use today there could be (for example) time-varying margins where Uber uses surge pricing profits to entice extra drivers for a short period of time; a "successful" balance wouldn't be equalizing supply and demand at regular fare prices as much as it would mean something like maximizing the revenue until the marginal rider's fare goes entirely to enticing the marginal driver to "log in."  (Not paid directly to the driver of course; things get pretty lumpy and averages are only useful to aggregate the financial performance of the overall system)
  • 4.o)... Uber's biggest risk in their current business is arguably not the need to find large enough markets, but rather successfully defend against erosion of the platform margin.  Efforts such as ride pooling can be viewed as ways to raise/consolidate demand (through lower individual ride prices) by combining revenue at more than one individual ride and less than two separate rides.  To the extent the riders are indifferent (modulo the price/time tradeoff), this slows revenue growth while protecting driver income and boosting margins to Uber

Uber will eventually need to protect margin and capture driver revenue in a way not permitted by open competition with its rival on-demand firms.  Re-regulation and "acquiring" the drivers seem to be the two thematic options.  Both of these would likely result in capital intensive businesses, exacerbating the main valuation question
  • 4.p)... Uber essentially launched through a regulatory "loophole" present in most major cities: there is no cap on limousine operating licenses, but each limo must pass safety inspection and each ride must be contracted individually ahead of time.  Only licensed cabs (and illegal jitneys) can troll the streets looking for riders to flag them down real-time; cost/efficiency limitations of manned centralized dispatching kept limo companies consigned to higher price points.  The dynamics of this result in highly fragmented and relationship-based black car companies
  • 4.q)... The core problem Uber addresses is one of allocating costs and surplus from short-term mismatched supply/demand created by time-varying demand and a relatively fixed number of drivers. In a traditional taxi system, during periods of peak demand (including locations at the fringe of the network) consumers don't have to pay more but artificial shortages make it hard to catch a cab.  At other times, cabs are easy to find but artificially costly.  Uber and its competitors allow supply and pricing to float to varying degrees.  Some drivers report that driving for Uber is only profitable during price surges.  Independent drivers help, but aren't required for this model to work as Uber could easily arrange to schedule a mix of full-time drivers and paid temps.  A demand spike allocates the surplus to drivers (and Uber) by raising prices temporarily.  If this spike persists, additional drivers come on line.  In theory, prices then adjust back down.  In practice (today), that doesn't happen immediately in part due to the challenges of getting drivers to the right place.  Uber is actively trying to shape the locations of drivers to balance rides particularly during surges.  
  • Ops note: in practice, each additional driver should shift the optimal location distribution of existing drivers.  To be efficient, this requires centralized coordination and drivers to be more clearly employees.  In the extreme, drivers should follow a route (including waiting in place) dictated by central optimization when empty.  By leaving the waiting location up to individual drivers, Uber externalizes this cost/risk onto drivers.  Uber apparently tries to guide drivers, but the process is subject to strategic behavior by drivers.  Depending on the number of drivers coming on-line during the spike and the duration/density of the spike, Uber's own financial interests may be aligned towards maintaining the surge price.  Perhaps more on this in a future post
  • 4.r)... At launch, Uber recognized the ubiquity of smartphones could add a fig leaf of legal cover to dominate black cars; artificially high (and undynamic) taxi rates later enticed Uber down-market to reduce costs and compete directly with taxis.  Continuing down-market will place Uber in competition with privately owned autos and, in certain circumstances, public transit. Re-regulation is an interesting question: what type of regulatory structure could help Uber, be "good" for riders, yet still deter Uber's competitors?  Perhaps a future post
  • 4.s)... Uber has two broad strategic options to protect platform margins: restrict competition (re-regulate) to inhibit follow-on competitors, or replace drivers to capture their margin (either as "employees" of Uber taxi co, or as autonomous vehicles).  
  • 4.t)... It's not clear what structure would re-regulate the market in a way which inhibited competition other than a cap on active drivers.  While Uber could self-impose a (time-varying) cap, it seems unstable to assume its competitors would maintain similar discipline in a given city and disastrous if Uber drivers aren't kept from other apps.  A (time-varying) regulatory cap would re-create the taxi industry, albeit in a new form with different medallions.  At this point, Uber could take a larger platform percentage explicitly (functioning as medallion rental cost) or implicitly (making drivers employees).  
  • 4.u)... One wild idea is Uber could employ their drivers.  While many variations on this idea are possible, the basic concept is Uber could control exactly how many drivers are on the street, at which times, and in which neighborhoods.  If the core problem of the price/platform margin wars is that Uber's driver density can be piggybacked by other services, this would solve it.  Perhaps there are regulatory barriers to this, and I haven't thought through the cash flow implications.  It's unlikely to solve Uber's long-term valuation problem though it could put a band-aid on margin erosion.  It seems like if Uber flipped a switch and offered its existing drivers guaranteed income (hourly+mileage) at a schedule that worked for both of them, probably varying with time/location, competing services would have difficulty maintaining their own driver density without following Uber's massive cash investments.  Uber already has the raw data (in 300+ cities) to make this scheduling efficient
  • 4.v)... Uber could also capture the driver revenue through use of autonomous vehicles.  What type of business would Uber end up being... auto manufacturer?  Train manufacturer?  Taxi company?  Transit system?  Manufacturing is very capital intensive.  So is transit, and the flexibility of locating capital assets generally means any profits accrue to vehicle lessors rather than local operators.  Historically neither the auto industry nor public transit carriers return their cost of capital.  
  • 4.w)... Any transition into a capital-intensive business by Uber will exacerbate the dynamics of the main valuation question outlined in #3 (above)

5.  Uber is on a treadmill it can't afford to slow; even winning may not be enough for existing employees/investors if the high valuation can't be maintained/increased.  Will Uber prove to be more like Groupon or more like Amazon?  Implications vary depending on where you sit

Winning may prove disastrous for some shareholders if the high valuation can't be maintained.  This is an increasingly common theme in high-growth tech companies
  • 5.a)... At the right price, (almost) any investment is a good one.  New investors and lenders to any business will only do so if they believe it will earn a sufficient risk-adjusted return.  Existing investors (and lenders) will balance the business' need for cash with the cost/strings attached.  
  • 5.b)... When the need for cash is low, costs/strings are much more about agreeing on the allocation of gains from faster growth.  When the need for cash is high, cost/strings is much more about protecting investors from downside risk while leaving enough on the table to keep key people incentivized
  • 5.c)... High growth companies also rely on steadily increasing valuations for marketing reasons, including attracting/retaining top talent and "unicorn" status.  Negative "optics" of down- or sideways investment rounds have popularized complicated ratchet provisions and other potential dilutions to kick-the-can down the road.  If the company clears the higher valuation hurdle, everyone wins.  If the company fails to clear the higher hurdle, the newer investors are often protected by owning a larger share of a smaller company, diluting earlier investors and employees.  This is probably fair and appropriate if sufficiently transparent
  • 5.d)... You could learn a lot by working for a high-growth company trying to reshape the world, just make sure your cash comp is exciting on its own.  Eric Schmidt famously told Sheryl Sandberg  "If you’re offered a seat on a rocket ship, don’t ask what seat. Just get on."  The utility of that advice of course depends on your assessment of the risk of the rocket exploding and the quality of the contingency evac/recovery plans.
  • 5.e)... If asked to invest in Uber - including selling your company in exchange for a large amount of Uber stock - these valuation considerations should be top of mind.  Perhaps these dynamics contributed to Uber's failure to win its $3b bid for Nokia's mapping business.  Given the fundraising and cashflow considerations outlined in #1 and #2, it's likely that Uber's bid was heavily stock-based not cash

Is Uber more like Groupon or Amazon?

Groupon is a company that over-estimated the value in rolling up small business marketing spend.  This somehow persisted past their IPO, but not much longer.  It's likely that many investors and employees got much less than they bargained for, though some probably came out ahead.  It's likely that the Groupon story is basically complete.
  • 5.f)... Groupon had a 2010 revenue of $312m; 2011 revenue was $1.3b.  GRPN is a story of aggregation of local marketing; business model/strategy not necessarily flawed but had some basic limitations to scale which were overlooked (or ignored) by executives and investors.  Lots of hype, crazy growth, multiple competitors, crash in value.  It didn't help that many local businesses using the services found the cost of discounted customers so high (and difficult to convert) that it's original concept was of limited use except in overcapacity/commodity products (hotel rooms, group yoga sessions, restricted spending dollars, etc.).  Wikipedia has a good rundown of press coverage at the time, much of which speculated on these weaknesses.
  • 5.g)... Groupon rose from a $1.35b valuation in April 2010 to a successful $13b IPO in June 2011 .  By late 2012 it became clear that its original business model and growth were not sustainable; the market value had crashed to under $3 billion despite $2b in 2012 revenue.  By mid-2013 the SEC was investigating material weaknesses in their accounting.  The founder was fired; new management took over the business.
  • 5.h)... New management has grown the business by 50% in the past two years both organically and through acquisitions... and Groupon's market cap remains about $3b
  • 5.i)... Is this winning or losing? Depends on perspective.  IPO investors lost, though part of the pop was due to low float (ie price established by a handful of most bullish buyers).  Despite hundreds of smaller competitors, Groupon's nearest competition today (LivingSocial) is maybe 10% of their size.  Andrew Mason (founder/early CEO) reportedly walked away with $200 million.  VCs probably cashed out at least part of their investment during IPO at a strong win relative to their investment.  
  • 5.j)... Groupon revenue and market cap this morning were still both about ~$3.2b.  Is 1x revenue a good multiple?  Is growing the business by 50% in two years, but barely adding any value to the market cap, a success?  Once you're overvalued, the incentive structure is to double-down and somehow grow your way into the overvaluation. 
  • 5.k)... Disclaimer: Groupon's founder/CEO Andrew Mason is an alumnus of Carnegie Mellon (as is this author)

Amazon is a company that used a low-margin, minimal-profit (at scale) business to branch into related businesses.  They are still writing their future, and have large physical (and IP) assets underlying their valuation
  • 5.l)... Amazon.com launched in the mid-90's as an online bookstore, quickly extending into CDs.  It then leveraged its expertise and assets in e-commerce "storefronts" as well as back-end warehousing/fulfillment to grow not only its own business but often that of other retailers such as Toys R Us.
  • 5.m)... In the early years, Amazon was relatively agnostic on the need to operate their own fulfillment - one team brought in third parties like Toys R Us to help accelerate economies of scale; a second team focused on making the fulfillment centers efficient including expansion as required; and a third team focused on identifying items which were better fulfilled by specialized third-party fulfillment.  (disclaimer: this author interned on the second team during 2003 and declined a full-time position including a signing bonus of 4,500 RSU at ~$17/share)
  • 5.n)... The seasonal nature of retail meant that Amazon had massive amounts of unused capacity during the non-holiday parts of the year.  Especially computing capacity.  What started as an experiment in renting out spare cycles in the early 2000's had turned into the formal business of AWS by 2006.  This sparked an entire industry of on-demand/cloud computing.
  • 5.o)... Other innovations and synergies followed, each using the "platform" of Amazon's underlying infrastructure and customer base to extend the business in interesting ways.  Prime.  Kindle.  Fire TV.  Music/video streaming
  • 5.p)... Amazon's market cap peaked in 1999 at about $24b, fell below $5b during the dot com bust, and didn't exceed it's 1999 peak again until 2007 (after Prime and AWS both started paying off).  Today it's worth about $240b (and while cash flow positive, still posts minimal earnings as it reinvests profits in the core businesses)

Thinking out loud...

For that matter, Google is a company that has built a ~$500b (market cap) company directly from a search tool that Excite wouldn't pay $1m for in 1999.  In 2003 Google's fifth acquisition was a $100m purchase of Applied Semantics (AdSense).  Today, Google still earns 90% of its revenue from ads; their popular "experiments" are largely ignored in their market valuation and can arguably be seen as perks afforded to the management by the controlling shareholders.  (Who may effectively be the same people.  And yes, I think most of the experiments are pretty awesome)

Thinking out loud, can Uber build a ubiquitous fleet of (autonomous?) vehicles to tackle adjacent point-to-point transportation markets?  The obvious markets (fast foot delivery?) don't seem nearly large or lucrative enough, but that's a vision they're pushing and have been for several years.  Uber would need quite a few hops to pull an Amazon, and their current internal free cash flows don't seem to support it the way Amazon's did.  Amazon's early investments were in physical plant which continued to pay dividends; Ubers are (mostly) in ethereal driver/passenger incentives and legal/lobbying which benefit the whole industry if successful.

It's not clear there's an AdSense-like Google pairing either, where a fourth party would be willing to bear the costs of helping Uber by subsidizing bringing drivers and passengers together.  Can Uber sell its detailed knowledge of passengers to show exceptionally valuable ads in vehicles?  Conduct political polling?  Lots of crazy ideas in here, but hard to imagine a mass market where impressions are worth $5 or more (what percent of Uber rides are to the same riders taking multiple rides per week?  Per day?)

Another possibility is that Uber's software platform is likely to be robust enough that it could provide market-as-a-service to local governments and/or other entities who want to efficiently and transparently administer a local taxi and bus fleets.  Many of these cities/competitors are small enough they are unlikely to be able to support the development/maintenance of a single-city app.  And they're bureaucratic enough that even the larger ones are likely unable to properly administer the fair and timely development (Exhibit A: Healthcare.gov or any open data city portal).  And (modulo local maps and pricing/demand dynamics), every city is going to work more or less the same way.  

A platform-as-a-service would shift the capital requirements (and political/regulatory/legal burdens) of Uber's expansion back onto existing/new entrants.  Would Yellow Cab contract with Uber if Uber had an option which supported legal taxi rates?  Could local Taxi/Limousine Commissions require all on-demand rides be transacted through a single marketplace (at prices set by individual participants)?  Perhaps.  A variant of this is that individuals/organizations who own self-driving cars could sublet them out to Uber when not in use by the principal owner.  With value pricing - Uber capturing a portion of the revenue savings from disintermediating existing overhead - this could possibly work in the short term.  Would be hard to work in the long term unless Uber sold 50-year leases a la Macquarie and roads/bridges/plane terminals.

In this model, Uber would function more like a stock exchange.  What are the dynamics around the persistence of multiple regional/global stock exchanges?  And what are the dynamics that valued the NY Stock Exchange at $8b in total when it was purchased in 2012, despite being the platform on which ~$20 trillion of annual transactions took place?  Seems relevant.

There are many interesting questions which I haven't asked, and the answers will have significant impact on various stakeholders beyond those outlined above.  

TL;DR: None of the above ideas seem likely to justify Uber's current valuation, nor generate sufficient cash flow to fund the path forward.  I believe Uber is a great idea being executed in an average way.  It's helping to unravel decades of rent-seeking in legacy transportation costs.  I hope they (or their intellectual successors) are successful.  I would not be keen on investing, unless I were offered significant downside protection as well as protection from future dilution.

Thanks for reading,

No comments:

Post a Comment

Focused, topical critiques are welcome though my responses may be limited. Off-topic comments, trolling, spam, or anything not pulling its weight will be deleted.