Description

Background

Limelight Networks (LLNW) is a legacy content delivery network (CDN).

Akamai (another competitor in the legacy CDN space) provides a good explanation on traditional CDNs on their website: https://www.akamai.com/us/en/cdn/what-is-a-cdn.jsp

“A CDN is a highly distributed platform of servers that helps minimize delays in loading web page content by reducing the physical distance between the server and the user. This helps users around the world view the same high-quality content without slow loading times. The goal of the CDN is to reduce latency (the delay between submitting a request for a web page and the web page fully loading on your device) by reducing the physical distance that the request has to travel. Without a CDN, content origin servers must respond to every single end user request. This results in significant traffic to the origin and subsequent load, thereby increasing the chances for origin failure if the traffic spikes are exceedingly high or if the load is persistent. To combat this, CDNs store a cached version of website content in multiple geographical locations around the world, which are known as “points of presence” (POPs). These POPs will contain their own caching servers and will be responsible for delivering that content in the user’s location.”

Legacy CDNs such as LLNW and AKAM have been around for 20+ years and have spent considerable capital adding POPs globally to remain fast and relevant. LLNW has over 120 POPs globally, with each site containing a high density of servers. The company places these cache servers as close as possible to ISP networks to reduce latency. Content owners such as media and gaming companies pay LLNW to deliver their content to end users. On the cost side, LLNW pays ISPs and network carriers for hosting its servers in their data centers.

Short Thesis

LLNW plays in the low-end content delivery space that is largely commoditized at this point. This segment of the industry is highly competitive with low switching costs and consistently declining prices.

The chart below shows pricing for third-party CDNs (lowest price offered – i.e. largest customers) for the delivery of OTT video and software downloads. While not all companies can get the lowest price offered, the ones that do are the ones that generate the largest overall portion of growth in video/software downloads, and this pricing drives the overall market trend.

Third-party CDNs could theoretically make up for low pricing with volume but this is difficult to achieve since all large customers use a multi-CDN strategy. OTT media companies (Amazon Prime, Disney+, etc.) that use third-party CDNs typically use at least 4-5 providers with traffic share fluctuating constantly based on performance. Switching out CDN providers is easy and inexpensive and media customers play this dynamic to their advantage. Media delivery isn’t a sticky business. OTT video traffic has very low margins, if any at all – it’s almost impossible to make money with commoditized high-volume traffic 

Sell-side analysts are bullish on the name and point to growing traffic trends as streaming TV viewing has increased the year and LLNW has onboarded new OTT customers such as HBO Max (May 2020) and NBC Peacock (Jul 2020). While traffic may have be spiking what hasn’t changed in the market is price compression and revenue growth consistently lags traffic growth. Despite revenues increasing nearly 30% y/y in 1H’20, the company still generated negative operating profit.

LLNW has talked up opportunities in live streaming as areas of growth. This area remains highly competitive as well and at the end of the day latency issues are a commodity now and LLNW’s network doesn’t do low latency better than competitors.

While the revenue side will come under pressure there is little LLNW can do to mitigate its cost of sales which will put pressure on gross margins. There’s a finite amount of last mile ISP providers for an increasing amount of bandwidth demand. This isn’t a cost area where LLNW can extract savings. 

LLNW’s high customer concentration is a liability

Amazon accounted for 30% of revenue in FY’18/19 and LLNW’s top 20 customers account for 72% of total revenues. Some analysts tout LLNW’s relationship with AMZN as a positive and validation that the company is providing a value-add service. Given the market dynamics just detailed, however, this is clearly not the case. If anything, high customer concentration with the largest customers in the industry is a liability given the tremendous negotiating power Amazon, Disney, et al have and for the potential for large customers to bring this service in-house. 

Eventually large OTT content providers develop their own CDNs and cease using third-party providers

OTT providers have indicated that when a service gets to ~20m subs that is the threshold from a size and scale standpoint where it makes sense to consider building your own CDN.

Per industry experts, Disney could build out a CDN for the delivery of Disney+ content for under $100m in initial capex costs. It is not costly or difficult to do so, which is why companies such as Activision Blizzard, Apple, and Netflix among others have done it. The argument that third-party CDNs can provide a lower cost option doesn’t mean that companies won’t bring the services in-house – ultimately content providers care about quality and providing the best user experience more than they care about cost. To that end, what content providers have found is that an in-house delivery system significantly outperforms third-party providers (LLNW, or otherwise) because it can be customized to exactly how the subscriber base for a particular streaming platform uses it.

Alternatively, legacy CDNs can also make up for low pricing for video delivery via selling other higher margin services which is the approach competitors have taken. LLNW, however, doesn’t have enough ancillary products to upsell and hasn’t developed and differentiated itself in any more profitable verticals.

Competitors are bundling video CDN as a loss leader with other higher margin businesses. Legacy CDN Akamai, for example, announced in 2Q that they had achieved a $1bn annual run-rate for their cloud security business. This is a growth vertical where, importantly, pricing is more value focused and there are many ways for vendors to show differentiation amongst competitors in product functionality and customization. This business does 80%+ gross margins which helps offset thin margins on the video delivery side. 

Competitors are involved in video delivery, despite low margins, because it helps in customer acquisition and retention (bundling services) and it also helps on the cost side. On the cost front, fees are paid to ISPs based on bandwidth use, and being able to hit higher minimum commitments drives lower costs.

Ultimately, generating profitability at scale in a business that competitors are using as a loss leader isn’t a sustainable model for long-term profit growth. 

LLNW’s hardware driven network architecture prevents it from pivoting to areas of innovation in the content delivery space that are driving growth for top competitors.

Edge networks that were built from software-defined networking (SDN) architectures better allow for flexibility & customization because they are programmable. All CDNs can’t do this because some (such as LLNW and AKAM) are stuck with their existing network architectures with hardware infrastructure and the massive number of POPs they created that must inter-network and require significant ongoing maintenance. Custom hardware, like routers, load balancers, and security appliances, lack flexibility to support internet traffic demands that are increasingly complex and dynamic.

Growth in the space is tied to providing customers with more functionality and customization and players such as Fastly and Cloudflare have created huge developer-friendly networks where customers can run their own applications and custom code across an entire global network. LLNW has discussed their edge compute capabilities but it lacks the high programmability and large-scale coordination that SDN-based architecture affords and any developments they make in this area are miles behind the curve as they cannot leverage any of their existing infrastructure to compete. 

High capex spend required just to maintain existing infrastructure with growth capex necessary to expand network adding even more to cash outflow

Network depreciation runs at 10-11% per year so LLNW has to spend a significant amount in capex just for its existing network, with even more spend required for network expansion. The company spent $35m in capex in FY’19 for maintenance and expansion with another $25-30m expected this year which fully offsets their EBITDA guidance. Competitors with SDN-based architecture can add capacity without anything close to the capex costs of legacy CDNs. 

Valuation

The replication value for LLNW’s network is around just ~$200m vs. a TEV of ~$875m. The company only has $49m of net PP&E on its balance sheet and 5 years of R&D spend amounts to ~$125m.  The company trades at 22x FY’21 consensus adj EBITDA estimates of $42m – significantly higher vs. its historical range (10-12x) on forward estimates that are unlikely to be met. Akamai, which despite its own challenges is certainly better positioned vs. LLNW, trades at only 12x FY’21 estimates. LLNW valued at 12x on flat FY’21 EBITDA equates to over 50% downside from current levels. 

Catalyst

Earnings misses 

Slower growth leading to multiple compression