Description

Summary overview & thesis

Vivint (NYSE: VVNT) is the 2nd largest player in the residential security / smart home market in the US with close to 2m subscribers. The company is majority owned by Blackstone (~48% ownership), who first invested in the company back in 2012; in late 2019/early 2020, Vivint went public through a SPAC.

While Vivint has had a mixed reception in its short life in the public markets, it appears to us that it has become an orphaned stock. It has a small float (~25% or ~$300m); is in a niche industry that historically has had more activity in the private markets and continues to be in the shadow of Amazon and Google entering the space; no natural shareholder base (not growth-y enough for growth investors, but also not cheap enough for value investors) or sell-side coverage (pretty much varies between business services and hardware); and pretty complex accounting that makes it difficult to understand and accurately track the unit economics. To top it all off, in the second half of 2021, Vivint changed the structure of its financing program with Citizens, which has had the impact of depressing FCF this year and masking the true underlying earnings power of the business.

We have covered Vivint for a long time in the private markets, and believe that at current levels, the equity presents very attractive risk. Vivint has been one of the fastest growing companies in the space and pioneered the transition from traditional home security to smart home. The company has a fully integrated model: producing the hardware; owning the sales channel and installation/service; built the operating system to connect all the various devices; and provides its own security monitoring. While the traditional home security industry has been stuck at ~20% penetration for years, I believe we are in the beginning stages of the smart home penetration curve, and Vivint has a long runway to grow (even alongside competitors like Amazon/Ring and Simplisafe). With very attractive unit economics producing mid-20% returns, the company has a clear opportunity to deploy capital at attractive rates for many years.

With ~$2.5b net debt and ~$1.1b market cap, the current enterprise value is $3.6-3.7b. We believe the company will triple levered FCF to $225m by 2026 compared to ~$70m guidance for 2022. This ~$1 per share in FCF, at a ~12x multiple, together with ~$3 excess cash per share, translates to a ~$15 stock or ~3x.

[We note that Vivint was posted by compass868 last September, but we believe our write-up still adds value / takes a slightly different approach to the analysis.]

Industry overview and unit economics

It is helpful to have a quick background on the industry and unit economics to understand Vivint’s positioning. For the “traditional” home security industry, Vivint is the 2nd largest player with close to 2m subs. ADT is the largest player in the space with >6m subs. Monitronics/Brinks has close to 1m subs and after that, there is a long tail of regional players. These home security players grew in a “push” model where the product was sold, not bought. Companies went to market either directly, with their own door to door sales force or phone sales, or through the dealer channel, where regional dealers originated accounts and then sold them to ADT/Monitronics for a small profit (and then ADT/Monitronics service the account and bill the monthly revenues). While most companies had a mix of dealer and direct creation, Vivint was unique in that they were all direct (originally just a door to door sales force, but more recently built out a direct phone channel).

In terms of the unit economics, the cost to create an account was typically >$1,500. The breakdown is roughly $750 of commissions, $500 of equipment and $250 labor/installation. You then charged the consumer $50 monthly and had ~$12.50 of servicing costs. Churn was typically low teens, so close to 8 year life, so as long as your creation multiple was reasonable, you could generate a decent IRR in the mid to high teens.

Vivint changed the model as they transitioned from simply home security (panel, couple door/window sensors, etc) to a more complete “smart home” which was really just a lot more devices (indoor and outdoor cameras, doorbell cams, smart lights, etc). Vivint sold much more equipment to each subscriber (>$1k) but still fully subsidized it with its own balance sheet, and then increased the monthly charge to the consumer ($70 vs $50). Much of Vivint’s early life in the private markets was characterized by cash burn and raising debt to scale the business in this model.

In 2017, Vivint introduced a financing plan with Citizens for the equipment purchase. Citizens would pay Vivint $1,500-2,000 for the equipment, Vivint would pay a monthly fee on the loan balance (Monthly Discount Rate or MDR) and share in a portion of loan losses, and the consumer would pay the same $70, but ~$25 went to Citizens over the 60-month financing period and ~$45 went to Vivint for the recurring monthly fee. Vivint’s net creation went to <$1k (as they sold the equipment for a profit), and despite the MDR/loss share (which they booked as a 25% holdback on the financed amount), it was positive for the unit economics for the company, as they took advantage of Citizen’s lower cost of capital, and dramatically improved Vivint’s free cash flow profile.

In mid-2021, Vivint changed the model with Citizens to a revolving line of credit loan instead of an installment loan, which made it easier for the consumer to add on additional equipment purchases to their financing plan (whereas before they would have needed to get a new loan). But Vivint now needs to pay the MDR and loan loss share upfront. 25% of $1,500-2,000 in equipment financing is $375-500 additional per subscriber added for Vivint. For 2022, when the new changes of the program will be in full effect (in 2021 they only impacted 2H), Vivint will see a double impact on FCF – they are paying the MDR and loss share on previously financed accounts, while also paying the full upfront MDR/loss share on new subscribers added. This distorts what the true steady state cash flow Vivint can generate, and is one of the reasons for the mis-valuation, in our view.

Before we continue, while on the industry discussion, we think it makes sense to address the risk of Amazon, Google/Nest, & Simplisafe, as that is a common investor pushback. The high level argument is that these companies approached the market with a DIY model, and that is a fundamentally different end market than what Vivint is targeting. DIY models typically only have a handful of equipment, no installation (as the name implies) and offer optional professional monitoring. Because of less equipment and a ”pull” model without sales commission for reps, they can offer their solution for a lower recurring monthly cost ($10-30 instead of the typical $40-50). First, consistently in our research, we have heard that this targets a younger demographic that tends to be more renter focused, and therefore acts as a gateway product to a professionally installed home security system. Second, if you look at ADT and Vivint’s subscriber trends since these players entered the market, they have not materially changed. And third, it just makes sense to us intuitively that people would pay up for a white glove service with more gadgets to protect the largest investment in their portfolio and make it more enjoyable/convenient to live in. We think it is telling that Google decided to make a strategic investment and partner with ADT to pivot Nest to be a part of a more professionally installed product/service set.

Accounting

Normally we would save a discussion on the accounting for the appendix, but here we believe it is such an important part of the thesis and a key point of investor confusion. In our analysis, we try to highlight 2 things: the true recurring cash EBITDA of the subscriber base and the actual customer acquisition cost that Vivint incurs.

Recurring cash EBITDA of the subscriber base:

Take average recurring monthly revenue (RMR) per sub and deduct the service cost per sub to get to a gross profit $ contribution, then deduct cash G&A and financing costs (MDR/loss share). The service cost per sub is a non-GAAP KPI that the company provides in its investor presentation, and through conversations with IR, we understand that there is some G&A allocated to this metric (quantum unknown), but this is the best approximation we can get. Cash G&A needs to remove SBC allocated to G&A. Then there is also expensed subscriber acquisition costs embedded in G&A that need to be removed, by taking the total expensed SAC (non-capitalized contract costs in the company’s covenant adjusted EBITDA bridge) plus SBC in selling expenses, less total selling expense. Finally, there are the finance costs associated with the consumer financing program that need to be deducted – this is the Settlements in the disclosure on the Consumer Financing Program Contractual Obligation (and we have confirmed with the company that both the loss share reserve as well as the MDR is included in the liability figure here).

The result is that, for 2021, the company has close to $800m GP$ ($47.5 recurring monthly revenue less $10.5 service cost, so $37 GP$ per sub per month, across an average of 1.78m subs). This has increased from ~$600m in 2017, as the company’s sub base has grown from 1.2-1.3m to 1.85m, but GP$ per sub per month has come down from ~$40, as recurring monthly revenue per sub has decreased with the move to financing (i.e. Vivint gets low $40 RMR vs $70+ when financed on its own B/S), but service cost per sub has declined as well (from $15-16 to $10-11 today). Cash G&A excluding SAC was about ~$150m for 2021 and financing cost of ~$90m gets us to a recurring cash EBITDA of ~$550m, up from ~$450m in 2017 (but down from close to $600m in 2020, as the company temporarily cut G&A during COVID).

For comparison, Vivint’s disclosed Adjusted EBITDA for 2021 is $669m, but includes non-cash deferred revenue, does not strip out expensed SAC, does not deduct financing costs, includes standalone equipment GP $ and includes other non-recurring / non-equipment revenues. Similar story for Covenant Adjusted EBITDA, which is >$1,000m for 2021. So these EBITDA figures are completely useless to assess the earnings power and cash generation of the business; they are nearly impossible to model with confidence, and when investors see research analysts valuing the company on these figures, it is easy to see why there is frustration / confusion for how to understand and value the company.

Subscriber acquisition cost:

Next is calculating what Vivint actually spends on SAC. There are 3 main components here: capitalized costs, expensed SAC costs, and the equipment upfront. The first two are easier, as they are disclosed on the cash flow statement and covenant EBITDA bridge, respectively. The equipment upfront is a little trickier. We start with deferred revenue on the cash flow statement. But this is effectively a net number (of deferred revenue recognized in the period), while we want a gross number (equipment financing that has come in). We add quarterly non-service revenue in the period, by taking the delta between total monthly revenue and total monthly service revenue (found in the earnings releases); this serves as an approximation for deferred revenue in the period (but there is likely some non-deferred and non-service revenue in here, such as one-time equipment sales…we will adjust for this later). Next we add in the additions to consumer financing obligations (representing the portion of the consumer financing that Vivint reserves for MDR & loss share). We then subtract the change in Retail Installment Contract receivable (these are Vivint financed customers, that the company no longer does), as that change would be a change in deferred revenue, but does not come with any financing inflow as Vivint was financing that on B/S. Finally, we make an adjustment to quarterly non-service revenue, as some of this is not deferred revenue, but actual cash revenue coming in from one-off equipment sales…we do not have an exact figure here, but estimate it to be in the $50-100m range annually. You can see below that with this bridge, we get pretty close to the mgmt disclosed KPI of avg proceeds collected upfront.

We can now arrive at a net SAC per sub number, with just two more final minor adjustments. We take capitalized costs (disclosed on cash flow statement) and expensed SAC (disclosed in EBITDA bridge) to get to a gross SAC figure. We then add in SAC equipment COGS, as this is running through the I/S, but we exclude in our calculation of recurring cash EBITDA (and we estimate the gross margin here to be in the 40-60% range, based on comments from mgmt/IR). We also subtract commissions related to energy/insurance pilots (which are effectively pass thru commissions to sales people for cross selling solar/insurance). Finally, subtracting the equipment upfront gets us to our net SAC per sub number.

Why do our figures differ from mgmt disclosed net SAC per sub? If you dig in to their definition with mgmt/IR (they do not disclose this methodology, but explain it in investor calls if you ask), you will learn that they exclude ~15% from gross SAC, as ~10% relates to service/upgrade costs of existing subs, ~3% is related to pilots, and the rest is residual commission to DTH reps for prior period installs. If you run this adjustment on our gross SAC number (and excluding the adj. to SAC equipment COGS), we can get pretty close to their disclosed KPI.

Valuation & Returns

Now that we have a solid foundation for understanding the economics and accounting for the business, we can discuss our projections and view on valuation.

First, let’s level set on tying out their 2022 levered FCF guidance, using our economic framework. We estimate recurring cash EBITDA to be ~$560m, which is pretty straightforward, save for the estimate on financing cost. We assume that this goes away gradually over the next 5 years (as prior subs that were financed in this model churn off and the typical loan was for 60 months), and cumulatively we estimate close to ~$300m of cost here over 5 years, compared to the YE 2021 derivative balance of ~$217m. This could be conservative, but it could also reflect higher than expected customer defaults (which might become more of a base case if we go into a recession).

We then have equipment gross margin $ (what we excluded from net equipment upfront in our calculation of SAC above).

The last major item of note is SAC. We assume ~75% financed, at a creation cost per sub a bit north of $1k, and then 25% paid in full, which have creation costs closer to $500. Blended you are at a little under $1k, so your 360k of gross adds cost about $340m.

Our base case is that the company will triple its annual levered FCF over the next 5 years (from the 2022 guidance of ~$70m to ~$225m). This ~$150m increase is from unlevered FCF going from ~$210m today to ~$360m by 2026, which is the result of 2 primary drivers.

First, gross margin $ increase by ~$60m (from ~$800m to ~$860m). This in turn is driven by subs growing from ~1.9m to ~2.3m, which assumes gross adds flat at ~360k per year and a slight increase in churn to ~12.5%. Offsetting the sub growth is avg GM$ per sub going from ~$35 to ~$31.5 (avg. RMR continues to come down as Vivint moves towards 3rd party financed subs instead of subs previously financed on B/S…but service cost per sub assumed to stay flat at ~$11).

Second, financing cost comes down as Vivint no longer pays out MDR and loss share over time. We have this going from ~$100m to ~$10m, or a ~$90m benefit to recurring cash EBITDA.

In 2026, the company will have built up close to $900m in cash from $200m today (assuming no cash used for buybacks or debt reduction), representing about ~$3.5 per share (on a share count of ~240m from ~209m today, so factoring in expected SBC). The $225m of levered FCF will be close to ~$1 per share. If you value that ~$1 per share at ~12x, and add in $3.5 per share in excess cash, you have a ~$15 stock, which would be a ~3x over 4-5 years. We like that risk for a business with attractive unit economics, large underpenetrated TAM, and in the top 2 category leaders in the space.

Risks

Aggressive sales tactics - this was noted in the comments of compound’s write-up and we do agree that it is a valid risk. We have been around this industry for a long time though and would just note that this is pretty normal (again, this is a product that is sold, not bought). In our view, the Vivint product is a lot better than traditional home security / alarm monitoring, so there is a more valid value proposition to the consumer and it’s not like the whole model is based on selling an overpriced, sham product (which for some traditional alarm companies, that is pretty much the reality).

Customer defaults - Vivint shares in the losses with Citizens (subject to certain thresholds, but the full details are not public). In our write-up we noted that we were factoring in more losses than Vivint has accounted for, but we could still be too low, depending on how bad of an economic environment we are in. The fact that the Vivint product is “protecting” people’s homes gives us some comfort that it is viewed as critical, but it is still a costly monthly expense for a consumer, so certainly there could be larger than expected defaults in a recession. But we would also note that moves typically go down in a recession, and moves can represent around half of churn, so there could be that offsetting benefit to Vivint as well.

Sales retention/commissions - a major strength of the Vivint model has been the sales machine it has built. However, there has been noise recently around sales people leaving, in particular to the solar industry, as the commissions there are outsized relative to home security commissions. We think this is a major reason for Vivint’s return to solar through a partnership and letting sales people cross sell (and therefore capture more commissions), but this is something we will be watching closely going forward. 

Catalyst

Company posting a more detailed accounting reconciliation to their non-GAAP KPIs 

Share buyback (but Blackstone has wanted to increase the float, not decrease it...)