Description

Loral Writeup 5/1/2020

LORL is finally a very compelling investment since, after seven long years of waiting, the company announced on 4/30 that they are in advanced talks of combining the publicly traded LORL with the satellite operator, Telesat, along with a $5.50 special dividend paid out to LORL shareholders in late May. Our sense is that liquidity constraints in other areas of their portfolios finally brought the deal to the finish line with MHR wanting the special dividend and PSP wanting liquidity in Telesat.

When mentioning this idea to a few generalists this past weekend, the pushback I got is that you need a technical understanding to figure out what GEO satellite free cash flow looks like in 5 years and to evaluate whether there’s merit to the LEO investment. My response is that the stock is so cheap and the current owners are so focused on returning cash over the next 5 years, that when you subtract the $5.50 special dividend we get this month and another $2-3 we likely get in the next few months, you pay about $15 for the stock with the company likely to dividend $3-4/year for the foreseeable future. Even if the stock doesn’t move (though I think it would since it’d be a 22% dividend yield), you get your money back in 4-5 years, plus all future free cash flows and a potential incremental 5x more upside should LEO meet company expectations. Essentially, you’re buying this so cheap with real near-term distributions that you don’t need to be a tech expert – all that is a free option. 

I’d also add that 80% of GEO revenues for the next 5-7 years or so are contracted, in backlog so risk to the dividend stream is also minimal while most of longer-term revenue has an extremely long tail. 

We believe the stock is worth $38 at 7x core business free cash flow (this includes capex spend on their LEO constellation) or 144% upside and $95 including the LEO build or 511% upside.

Why is the Stock Misvalued?

1. Massive investor fatigue – as previously noted, investors have waited 7 years for this transaction amidst the hedge fund/special situation carnage Klarman outlines in his 2019 letter. Initially, the stock was in the $60s but it continued to drop as nothing happened, into the $30s. I’d heard of multiple SPVs raised to invest in this situation, one of which I know was wound down; it’s a lot easier to buy than to sell an illiquid hedge fund special situation, hence they probably caused some of the decline.

2. C-Band drama – I’ll outline the C-Band situation later but the stock dropped from the low $40s/high $30s into the low $30s based on C-band volatility as Intelsat practically went to $3. That said, since LORL never really went up when C-band conversion caused Intelsat to rally from $3 to $30, even the much lower proceeds outlined by the FCC is meaningful to LORL at this stock price. 

3. COVID – obviously, the stock’s dropped in 2020 due to COVID; revenue is under contract and people are still watching television though 10% services aero/maritime (I account for this decline in my 2020 numbers).

I think all this fatigue abates since this is no longer a special situation with no end in sight. Once the reverse merger happens, sellside initiates and the market sees how cheap this stock is for a large and liquid asset, the stock will get closer to fair value. This is much better timing than when OF21 wrote the name up in mid-2018, since at that time, we were told in the AGMs in 2018/2019 that the deal was imminent, but now, we actually know a deal is finally here!

LORL is a holding company with a 62.7% stake in Telesat, which operates 16 geostationary satellites with customers in (1) broadcasting/DTH (48% of revenue) and (2) enterprise services (50% of revenue) – government, corporate, backhaul, maritime, aero and video distribution. Check out pages 27-30 in the Telesat 20-F; I like how they map out supply – demand.

DTH Broadcast Business

While the DTH broadcast business isn’t the best business in the world, there are a few reasons why it’s more stable than the cord-cutting trends we see within their customer base. (1) Point to multi-point: One satellite transponder or beam (point) services an entire geographic region (multi-point) and while a satellite TV provider (like DISH) may see cord-cutting across its footprint, it must continue to pay for transponder coverage in a particular region as long as it serves customers, there. (2) HD channels: HD channels demand more bandwidth than SD channels which requires the leasing of additional satellite transponder capacity by satellite TV providers. While benefits in compression technology offset this, the net increase from 720p to 4K more than offsets compression benefits.  

That said, it will still decline over time. Shaw Communications which operates satellite TV service in rural Canada cut service on Telesat’s Anik1 satellite because of a shift to more cable television in the regions previously serviced through satellite. However, they still lease capacity on Telesat’s Anik2 satellite under long-term contracts and will never likely cut that out (since it services very rural areas where they have no chance of cost effectively rolling out cable). Telesat’s other large DTH customers include Bell Canada and Dish. Most of this revenue is under long-term contract so this will be a slow bleed over 7-10 years with a very long tail. However, the stock more than prices in a MSD decline, here with little maintenance capex.

Enterprise Business

The enterprise business is not as easy to conceptualize as DTH since it’s a hodgepodge of different services but that’s also what makes it a bit stickier. (1) Consumer broadband – through payloads, Telesat offers Ka-band consumer broadband capacity to wholesale customers. There are plenty of investment theses on Echostar and Viasat denoting the merits of consumer broadband in rural areas through Ka-band satellites; it’s a real business with growth and a moat. (2) Backhaul – Telesat offers backhaul to wireless operators using satellite from their more rural towers. This business has been replaced by fiber in most areas so, while it’s not growing, the current footprint is really unreachable by fiber and pretty stable, at least in North America, where Telesat operates. (3) Maritime and Aeronautical – like consumer broadband, this segment grows as ships and planes utilize more broadband (ex-COVID). (4) Government/Military/Natural Resources – lumping all of these together which includes governments adopting new Ka-band satellites for highspeed broadband, remote access for rigs and wells, military satellite usage for drones…etc. While the company says they are innovating to growth satellite services to the military/government, I’d imagine this is pretty stagnant but sticky. 

When evaluating risks to enterprise fixed satellite services, the best example is Intelsat and the other European operators who destroyed value (and their stock prices) by building too much capacity in Africa. Essentially, they built capacity to service rural wireless backhaul from towers where fiber/cable builds ended up disintermediating their satellites. This resulted in a price war and value destruction. We are pretty certain this supply/demand imbalance won’t happen in North America, where Telesat operates, since those irrational actors (mainly Intelsat but also Eutelsat and SES) don’t have the ability to build any new satellites and are still reeling from the issues in Africa (and also Europe). Also, the competitive technology (Ka-band from SATS/VSAT) is focused on rural consumer broadband where revenue per bit is so high that the launch of Jupiter/Viasat 3 expands that addressable market with limited impact on Telesat’s declining data business post 2023.

What’s also important to note in Telesat is there’s no institutional imperative to continue to try to grow (or maintain a dividend). The owners of the company are much happier cutting capex and harvesting cash flows as the industry evolves. 

Financials/Modelling of GEO:

What’s so interesting about this business right now is that they spent a lot in capex in the last several years to launch Telstar 18/19 Vantage which boosted revenue and when we look at the satellites that reach end of life over the next 5 years, through 2025, Telesat can let them runoff without too much impact to revenue since they account for a small percentage of total revenue. What that means is that this isn’t a capex holiday (where capital needs to be replaced in the future), it’s an overall decline in capex that creates a free cash flow windfall that can be dividended out AND utilized for the LEO build.

We went through every satellite, number of transponders, geographic locations, customers, applications and percentages of revenue (I’ll spare including that since it’s all available through public information and estimates from experts) and we view Anik F1/F1R, Nimiq 1/2 and a few others in runoff with no need for replacement capex. This leaves about 7-8 satellites with a useful life from 2030-2040 remaining which generate the bulk of revenue while improving industry supply/demand, so no new builds need to be started. 

As previously mentioned, this means capex capacity available to fund the LEO build without impairing free cash flow to equity holders. I’ll get back to the GEO modelling after a short discussion of LEO since it’s all related.

LEO Constellation

There’s been a lot of talk of LEO builds over the last 5-6 years with little knowledge as to whether they will work and I’ve been a OneWeb skeptics (I have notes from 2017 that walk through all my issues with their architecture that I’m happy to send). That said, I think Telesat’s LEO constellation is really the only viable design (there are some interesting presentations you can find on the FCC website which walk through Telesat’s design). Essentially, industry groups and experts have ranked Telesat’s build higher than SpaceX or now bankrupt, OneWeb. Telesat is planning to build 200 or so (by 2023, probably delayed by COVID), while the other two are/were expected to build 5x that. Telesat’s LEOs are larger with a 10-12 year lifespan (vs. 5 years for the others); Telesat’s will solely be B2B (since antennas for LEOs are so expensive, it doesn’t make sense to build B2C); Telesat LEO latency is lower at 16ms (very important for LEOs), they have intersatellite links (so the satellites can talk to each other rather than requiring each to talk to a ground station, the latter requiring a ton more ground stations) and they have bendable beams which means less stranded capacity (OneWeb LEOs covered the whole Earth equally which means lots of stranded capacity over areas with higher population density). All of this means a much higher ROI for Telesat which means a viable business. Also, since Telesat already has a network of customers, distributors and ground stations, they aren’t creating an entire business from scratch like competitors. Finally, Telesat has priority rights for LEO for 4GHz in the Ka-band, supported by the ITU and the Canadian government. All of this points to why their build will likely be a success and why there are rumors that some part of FAANG will partner with them rather than going at it, alone. 

https://www.satellitetoday.com/mobility/2019/11/21/gilat-achieves-fastest-modem-speeds-over-telesat-leo-satellite/

http://www.mit.edu/~portillo/files/Comparison-LEO-IAC-2018-slides.pdf

Back to Modelling:

Running off some of the GEO satellites as discussed above results in the following revenue, EBITDA and free cash flow results. 

Using 40% of free cash flow to pay off debt, keeps us well under the senior debt covenants but with ample free cash flow for a 20% dividend yield, even including $83m in capex/year to be used for building the LEO constellation (and $27m maintenance capex). 

The cost to build the LEO constellation is $3bln CAD. With over $1bln CAD on the Telesat balance sheet, $525m coming from C-band (based on Ajit Pai’s tweeted numbers) and $600m CAD to be raised through debt, the funding gap covered by capex is $440m CAD or $88m/year for five years. Given the limited GEO capex over the next five years, this still means significant free cash flow for equity holders even after such a large undertaking as a LEO fleet.

LEO Projections

In mid-2019, the government of Canada awarded Telesat a $1.2bln LEO contract over 10 year with $600m coming directly from the government. Using our estimates of the number of satellites (and therefore the number of Gbps) over Canada, along with estimates on the amount of Canadian satellite capacity leased to the government, we get to about $56/Gbps/month in wholesale revenue for Telesat’s LEO constellation. This compares to about $330 for Ka-band GEO capacity which makes sense because that’s consumer focused while Telesat’s LEO capacity is wholesale. We use this revenue figure based on total constellation capacity and a conservative 25% utilization to get to the 2025 revenue and free cash flow estimates listed above. This jives with a mid-teens unlevered IRR the company once mentioned a few years ago as a hurdle/lower bound. We use current share count, not shrunk share count post buyback to get to our $8 in free cash flow per share estimates. 

The stock was in the $60s when people thought a reverse merger would happen soon so a $40 price target excluding LEO isn’t too egregious and apparently Rachesky turned down bids in the $70s years ago so a $90 overall price target doesn’t seem aggressive, either given the value of the LEO build.

Risks: They can’t use the restricted cash to finance the LEO build which means more debt or capex. Compression improves and reduces demand for DTH satellite capacity. Similarly, additional data capacity is built or terrestrial fiber builds expand. Too much LEO capacity from Amazon, SpaceX and Telesat means pricing wars.

Catalyst

Payment of special dividend and reverse merger of Telesat into LORL