WeWork-ed On A Few Numbers…
Ahh, WeWork. The hard-charging, rapidly expanding, commercial property company has finally released its IPO prospectus (vis-a-vis the ‘The We Company’, but I’ll refer to WeWork for nostalgia’s sake) as it gears up to take on the public markets. And as expected, its causing quite a controversy, given its slated $47bn private valuation. So I couldn’t resist taking a peep. But first, the usual disclaimer ;)
Disclaimer: Nothing in this article represents any form of investment advice, and are my personal opinions. I don’t hold any shares in WeWork (or any financial positions) either. This article is purely for academic purposes.
If you want to skip to the conclusions and the TL;DR, skip to the bottom. Otherwise, let’s get cracking.
There’s been a lot of attention on WeWork’s incredible losses as they match the rapid revenue growth, without a clear path to profitability. Or their umbrella corporation structure. But I don’t want to focus on that. I’m really intrigued by their unit economics, and how they sell it.
For the un-initiated, WeWork’s core business model is simple. They take out long-term leases from landlords, do up the property to be suitable for renting out to companies (you know, beer taps and all), and then rent out the space. Startups, freelancers, large enterprises all can purchase WeWork memberships that give them access to their constituent properties (including shared spaces, desks) in addition to further ancillary services. As long as you have enough occupancy in a property, and the combined revenue from your members exceeds your operating costs, you should be in the black for that asset. Here’s a pretty diagram:
WeWork is very keen to emphasise their strong unit economics:
“Our strong unit economics, together with the increasing cost efficiency with which we open new locations, gives us the conviction to continue to invest in finding, building and filling locations in order to drive long-term value creation” (p.g. 11)
Keep the parts in bold in the back of your mind. We’ll need them later. Now have a look at what I think is the most important diagram in the whole prospectus:
WeWork’s metric of choice for unit economics is contribution margin (p.g. 72). This in short is revenue from memberships and associated services less the operating costs of running the assets (i.e. leases and other operating costs). Think of it as gross profit for a property, or location (as it doesn’t take into account sales, marketing, or general admin costs). You’ll observe that the diagram above seems to deviate from this slightly (hint: it doesn’t), so let’s walk through what is going on here.
Breaking into the contribution margin
According to US GAAP accounting rules, leasing costs need to be treated on a ‘straight-line basis’. Usually, when taking out a commercial lease in the way WeWork does, WeWork does not pay a fixed rental amount from the get-go of signing the contract. They usually get a ‘rent free’ period where they pay no rent (usually while they do up the asset), which is usually an average of 9 months or so (p.g. 74). In return, towards the latter end of the lease period (which can last 15 years or more, p.g. 73), WeWork will pay ever increasing amounts of rent, termed as ‘rent escalation’.
For accounting purposes regarding costs however, WeWork is obliged to cumulate all the payments over the lease term, and then report the average payment over the period (i.e. ‘straight line basis’). So at a given point of time, there may be a discrepancy between the cash payment of the lease and the reported expense for that asset. If an asset has rent-free and rent-escalation features in the lease, which the large majority of WeWork’s properties do (p.g. F-16), then this applies. In the early years of the lease, the cash payment is lower than the average expense due to the rent-free period. The differential is termed as a positive ‘non-cash’ expense. During the latter years where rent escalation occurs, the cash payment is higher than the reported expense, and so there is a negative ‘non cash’ expense. In other words, the non-cash differentials between the cash payments and the reported lease payments net to zero over the course of the lease (p.g. 73).
Hence in the contribution margin diagram for the 6 months ending June 30 2019, WeWork defines the ‘adjusted lease cost’ as the cash payment, and the ‘non-cash GAAP straight-line lease cost’ as the differential. So if I take away the cash payment of the lease from revenues, as well as any other location operating costs, I should get the gross profit in ‘cash’ terms for that property. This is the $340mm (and corresponding 25% margin) that WeWork is referring to as the ‘contribution margin excluding non-cash GAAP straight-line lease costs’ (jeez, what a mouthful). If you’re wondering where the differential is recorded in financial statements, it increments the deferred rent liability on the balance sheet (p.g. F-16).
For my sanity and yours, I’ll refer to ‘contribution margin excluding non-cash GAAP straight-line lease costs’ as ‘contribution margin excluding non-cash lease costs’. And ‘contribution margin including non-cash GAAP straight-line lease costs’ as ‘contribution margin including non-cash lease costs’.
WeWork also reports the contribution margin including the non-cash GAAP lease costs. This sits at $142mm at 11% margin. Historically it sits around 10%-11% (p.g. 75). That’s considerably lower than 25%. WeWork emphasises that they prefer the margin measure that excludes the non-cash expense (p.g. 72). They argue that during a rent-free period, revenues are low and so are cash lease payments. As a property becomes more mature, revenues increase but so do cash lease payments (p.g. 74). The implication is that the contribution margin excluding non-cash lease costs should be roughly the same at any point during the lease life time.
When a property’s revenue is on the up and up
Let’s consider a property which is at the beginning of its lease and where membership is scaling up. If your revenues increase at the same rate as costs, then WeWork’s preferred margin measure stays constant. Naturally, as long as revenues exceed costs, the margin is always positive. The first measure is good to assess how well you are profiting at a given snapshot in time given your current membership base. You could have the same margin at both low occupancy (and so low costs); as well as at high occupancy levels (with higher costs) if you assume that non-lease operating costs scale with the revenue.
If you use the margin measure including non-cash lease expenses, then you would expect the margin to increase over time (and it may begin as a negative margin as revenues will be low, and leasing expenses are constant over the lease period). The second measure is much more useful to assess over time and assess how well you are utilising your asset to drive revenues. Here, you get penalised at the early stages of a lease as your occupancy rate is low, and a good utilisation of the asset (i.e. rapidly increasing revenues) shows a rapid improvement in this margin measure.
So let’s imagine what a ‘good’ property would look like (i.e. one with revenues growing at the same rate as operating costs) as its membership scales up. We would expect the contribution margin excluding non-cash lease costs to be roughly constant over time. If non-lease operating costs are more of a constant feature of the property as membership scales, then this margin measure would tick upwards. Meanwhile, contribution margin including non-cash lease costs should be increasing over time. The faster the growth of the latter measure, the better idea we have of the utilisation of this asset.
Imagine I add another ‘good’ early-in-the-lease property to my current (growing) property to form a portfolio. If individual contribution margins excluding non-cash lease costs are similar, then we would expect the portfolio’s equivalent to be reasonably unchanged; regardless of where the properties are in their individual lease cycles. However, the growth in contribution margin including non-cash lease costs would be suppressed as every new property initially has a low margin measure (potentially negative) compared to maturing properties. In fact, if I add more and more properties at an ever increasing pace, this margin measure may stay constant for a while, instead of increasing at the portfolio level.
And loan behold, this is exactly what we see with WeWork’s margins. Both margin measures are roughly constant over time.
So why does this matter?
WeWork says that a property ‘matures’ after 24 months (p.g. 11), where the occupancy rate starts to hit upper limits. They argue that they can consistently increase revenues by hiking prices at membership renewals and for new members (p.g. 74) at the later stages of a lease. If you believe that story, then the previous section applies. But let’s imagine that that doesn’t happen.
Intuitively, at some point, you would expect revenue growth to start to flat line. However, rent escalations are more common at the later stages of a lease. So if the cash payments for leases increases, and revenues flat-line for a property, the contribution margin excluding non-cash lease costs will have to fall (assuming other operating costs stay the same). Interestingly enough, since non-cash GAAP lease costs net out to zero over the course of the lease, at the later stages of a lease term, these costs will be negative, and so the contribution margin including non-cash lease costs will actually remain unchanged post-revenue growth stagnation. This sounds counterintuitive. But remember that the way GAAP accounts for lease costs is such that they are expensed at a constant value over the life of a lease. And so if revenues flat-line going forward, this margin measure won’t change either. The non-cash GAAP lease costs adjusts to changes to the actual cash payments such that the combination of the two always equates to this constant expense value.
Since WeWork has chosen the cash margin measure as their unit economics metric of choice, they will eventually see it falling as their property portfolio matures due to rent escalations if revenue growth falters. And so how could they engineer an uplift in this contribution margin? By rapidly expanding and adding more and more properties to their portfolio, with rent-free periods included in these new properties. These new properties will have a higher contribution margin excluding non-cash lease costs, and so the blended margin measure for the portfolio as a whole is kept high. This is especially pertinent since WeWork is adept at filling over 50% of their properties at the point of officially opening the property up to members; which is the point at which the contribution margin starts to be calculated.
And that is exactly what they are doing. Because their metric for unit economics depends on consistent revenue growth for it to remain high, WeWork either can hike prices or add more properties. And for now, they are choosing the latter. If they stop adding properties; and if the price hikes don’t increase revenues as intended, then there will be an issue with unit economics that portfolio growth has so far masked.
We’re not done yet…
You’ll probably need a coffee by now. I needed several, but there is a little more to this under the hood.
When WeWork opens a new property, they need to spruce it up before renting it out. This obviously incurs cost. If there is no rent-free period for the property, then WeWork will have to pay leasing costs before they open the property to members. And second, they will have to pay the direct cost of sprucing up the property in the first place. In this case, a landlord might give a lease incentive to WeWork; basically a kickback of cash (which is amortized) in appreciation of the fact that WeWork is improving the state of the landlord’s property.
You might have thought that WeWork would include any pre-operating locations expenses in the contribution margin calculation. They don’t:
“Our contribution margin measures exclude items that are not directly attributable to the membership and service revenue we realize from a given location in a relevant period, such as general and administrative expenses, pre-opening location expenses, …” (p.g. 72)
You’re also probably wondering, what’s in the pre-opening location expenses? In fact, its lease non-cash expenses:
“During the build phase, locations incur pre-opening location expenses consisting primarily of lease cost expense. Given the impact of free-rent periods, lease cost expense recorded in accordance with GAAP typically exceeds cash payments required to be made during this phase. Of the lease cost expense included within pre-opening location expenses in the year ended December 31, 2018, approximately 85% was non-cash expense.” (p.g. 82)
So this means that the contribution margin calculation excludes the non-cash GAAP expenses during a property’s rent-free period, and is only considered once the property is open to members. That means that the non-cash GAAP expense of $371mm for the 6 months until 30 June 2019 is in fact lower than what it would be if we included the pre-opening non-cash lease expenses. And therefore, the contribution margin including non-cash lease costs is higher than what it would be if those pre-opening location expenses are accounted for. The contribution margin excluding non-cash lease costs won’t be affected since there were few cash payments during these pre-opening periods anyway as these properties are in their rent-free periods.
Let’s carry out a thought experiment and see the impact of this. And let’s give the benefit-of-the-doubt and apply the H1 2018 pre-opening location expenses to the H1 2019 financials, to ‘mimic’ the case where WeWork is opening the same number of new properties in both periods. Using the H1 2018 figure of c. $157mm (p.g. 91) and assuming 85% of this represents the non-cash lease expense portion, we get $133mm of non-cash expenses. Taking that away from the $142mm (p.g. 72) contribution margin including non-cash lease costs and we get $9mm of this new contribution margin measure. Which is 0.7% over H1 2019 revenues of $1,349mm (p.g. 72). Hmmm…
Remember when we talked about adding new properties to a portfolio putting downward pressure on the contribution margin including non-cash lease costs? This is it in action. And because it places such a downward pressure on this measure, it makes complete sense (from a financial engineering perspective) to separate out pre-operating location expenses from the contribution margin calculation. And also why WeWork prefers the margin measure where the non-cash expenses are excluded.
And there’s one more thing to go …
But what about the lease incentives that are amortized over the term of a lease? Are they removed from the contribution margin calculation?
“While we separately present our contribution margin both including and excluding non-cash GAAP straight-line lease cost, we do not adjust for the benefit related to the amortization of lease incentives included in location operating expenses in our calculation of contribution margin.” (p.g. 74)
That’s a no then. And since the amortisation of lease incentives is a benefit to WeWork, then it decreases the cash payment portion of their lease costs (i.e. the $638mm adjusted lease cost, p.g. 72). Which puts uplift on both contribution margin measures. And why do they do this?
“Had we chosen to structure our lease agreements without these incentives and made the decision to instead finance that portion of our capital expenditures through a traditional loan or through draws under our credit facility, we believe that our overall lease cost payments would likely be reduced, and the incremental cash could have instead been used to pay down principal and interest, which payments would not have impacted our calculation of contribution margin.” (p.g. 74)
So they do this because they believe they would have secured lower leases from landlords. I leave it to the reader as to whether they have faith in that explanation or not.
What’s fascinating with this though is that WeWork are able to separate out the costs (non-cash lease expenses) and benefits (amortisation of lease incentives), with the former being excluded from contribution margin calculations, and the latter being included. With both decisions having a positive impact on the contribution margin calculation, and thus unit economics.
TL;DR
WeWork’s prospectus, and the way it treats unit economics, is incredibly interesting. To summarise:
- WeWork bases their preferred unit economics measure on a metric called ‘contribution margin excluding non-cash straight-line GAAP lease costs’. In other words, the gross profit from their properties where the lease costs are the actual cash payments that need to be made. This is considerably higher than the contribution margin including non-cash GAAP lease costs since a lot of WeWork properties are earlier in their leases and so benefit from rent-free periods
- As a newly leased property scales up membership, the contribution margin excluding non-cash lease costs should be pretty resilient and high (assuming costs grow at the same rate as revenues), and the margin including these costs should increase over time from a low/negative base. As a property matures and revenue flat-lines however, the story is different. We would expect the contribution margin excluding non-cash lease costs to fall (as cash lease payments ramp up), and the margin including these costs to stagnate
- WeWork can keep their contribution margin excluding non-cash lease costs high across their portfolio by consistently adding more properties (who have high margins individually due to 50%+ occupancy rates at the point of opening), which is what appears to be happening. This does put downward pressure on the contribution margin including non-cash lease costs, however
- WeWork excludes pre-opening location expenses from its contribution margin calculations. Since these expenses mainly include just non-cash lease expenses, removing them from the equation would give uplift to the contribution margin including non-cash lease costs. Taking them into account (as a thought experiment) could put this margin measure close to zero. Hell, it could even be negative depending on the number of new properties they are opening
- WeWork includes the amortisation of lease incentives (which is a benefit to WeWork) in their contribution margin calculations. This provides an uplift on both contribution margin measures. WeWork argues that if they used financing to do up their properties instead, it wouldn’t matter, and the margins would still be as high, as they would have been able to negotiate lower leases from their landlords…
For dessert, I’ll leave you with my most favourable tidbit from the whole prospectus. All the way from page 1:
“Our mission is to elevate the world’s consciousness”
Yes, they really did write that too…
Ciao, Dom
