APPEN - AI-Driven High-Quality Growth Stock Trading At A Fair Price
- Exposure to strong secular growth in a nascent AI/Machine Learning industry.
- Track record of growth and highly cash generative business model.
- No material impact from COVID-19.
Appen Limited (APX:ASX) is a leading global provider of language, data annotation and technology solutions to enterprise and government clients. APX has demonstrated a track record of growth, driven by its strong customer relationships. APX’s business model leverages a crowd workforce to provide training data that is used for the improvement of AI and machine learning algorithms. And the company is well positioned to take advantage of the next stage of growth in the AI industry driven by the broader adoption of AI across non-tech industries including automotive, healthcare, financial services, retail and government.
APX counts eight of the top ten global technology businesses as its customers, and has benefited from three key trends: 1) ubiquitous penetration of search across online platforms; 2) increasing investment in data analytics and machine learning; and 3) broad based acceleration in digital transformation trends. According to APX, the global AI market is expected to grow to USD 169- 191 billion by 2025, with ~10% of AI investment spent on data labeling, this implies a USD 17-19 billion addressable market size.
What is structured and unstructured data?
Structured data is comprised of clearly defined/annotated data types which are easily recognized and searchable. This data regularly exists in relational databases such as CRM and ERP systems, and includes fields such as dates, phone numbers, credit card numbers, names and addresses. AI and Machine Learning can easily be applied to structured data to identify trends, and relationships to provide useful insights or search results.
However, the vast majority of data created is unstructured, this includes audio, images, video, e-mails and social media posts. AI/Machine Learning can be used to turn unstructured data into structured data, but this requires multiple iterations of training, testing and fine tuning of the Machine Learning Model. The key input for this process is training data that is tailored to the needs of the model which APX provides to its customers. It is recognized by the industry that good quality data, at volume, is an important input for the training and optimization of AI and machine learning.
APX currently provides three key data types:
- Content Relevance – data annotation used to improve relevance and accuracy of search engines (web, e-commerce and social media).
- Speech & Natural Language – data used for speech recognition, synthesis and natural language processing. This sector has seen strong growth with the proliferation of voice-activated assistants.
- Image & Video – data used for computer vision models, typically used to automate tasks such as navigation and pattern recognition.
The Relevance business (87% of FY19 revenue) provides data for the improvement of search engines, social media and advertising. I believe this category is likely to see continued growth as data points are constantly being created. This can also be seen as somewhat defensive, given new data points are created even when the macro environment is poor. In addition, APX is leveraged to the growth in investments in machine learning and AI, particularly in Speech/Natural Language and Video. Although some of these data types can be treated like a commodity, many of the projects undertaken by APX are bespoke and the data categorization needs to fit a specific algorithm. As such, I expect the strong investment in AI to be a tailwind for APX’s future revenue growth.
APX is well placed to continue market share gains in a fast growing industry
APX is the leading provider of data to companies in search, AI and machine learning. I expect APX to benefit from the broad based increase in annotated data. There are numerous industry forecasts that suggest rapid growth in AI investment over the next 5 years.
As I mentioned earlier high quality data is important to support this growth in AI and APX estimates that ~10% of AI investment is data labeling, which suggests a potential market size worth USD $17bn to $19bn by 2025. Given APX’s scale, track record and existing relationships, I believe APX is very well placed to capture a significant share of this market. This market growth is driven by new and existing use cases, including:
- Speech related – chat-bots, assistants, natural language applications and translation services
- Video – surveillance, computer assisted vision, social media and video search
- Search, social media and online advertising – constant refresh is required to maintain relevance of search results as new data points are created
- Autonomous vehicles, facial recognition – computer assisted image recognition is gaining a lot of traction, this includes existing use cases such as FaceID
Organic expansion into China
China is the largest AI market outside of the US, and China has made AI a key tenant of its “Digital Silk Road” initiative. I expect APX will be able to leverage its international data labeling expertise to take advantage of the expected growth in the Chinese AI market.
APX has opened its China offices in Beijing and Wuxi and is continuing to build its relationship with large tech companies in China. Given China’s ambition to invest in AI as part of its “Digital Silk Road” initiative, this presents an opportunity for APX to expand into a new market. Specifically, APX is well positioned to provide foreign market knowledge and crowd for technology companies in China that are increasingly looking to expand outside of their core domestic markets.
Competition will increase however this should be outpaced by industry growth
Given the market that APX operates in is relatively new, and growing rapidly, I believe more competition to enter this space as the industry develops. However, many of the new competitors tend to focus on the higher end software and technical side of AI development, rather than the provision of inputs (annotated test data) that is required to improve and refine AI.
In addition, I believe APX has a number of advantages versus its competition being:
- Highly cash generative business model that can be used to fund M&A of new and emerging technologies.
- Large, scalable and diverse crowd.
- Long standing relationships with its key customers, and is a vital part of the search value chain.
- Better capitalized and funded than its smaller technology platform peers who are typically in “start-up” model.
Although APX may not have the internally developed IP and associated barriers to entry, APX’s high quality processes and scale make it difficult to efficiently replicate. APX has over 1 million crow workers in 180 countries, covering 130 languages and a wide range of sectors. Furthermore, APX’s business-critical data engineering function also mean a low-cost alternative is typically not worth the associated risk.
APX has demonstrated a strong track record of growth, driven by repeat work from its key technology customers. APX has generated a 60% top-line CAGR over FY14-FY19 (includes M&A). APX’s crowd-based business is also highly cash generative, and has demonstrated ~90% cash conversion over the past few years.
APX typically generates revenue from 12-month contracts on a per project basis. Some of these contracts are “recurring”, as the training data needs to be regularly updated and refreshed to stay relevant.
In Relevance, APX has multi-year contracts with its customers. However, the volume of work is still variable. APX manages the pipeline of work through regular communication with customers and quarterly business reviews. This allows APX to have good near-term visibility of the project pipeline.
Speech and Image revenues are more project driven, where billing is tracked against activity milestones on a predetermined timeline. There are fewer “ongoing” projects in Speech and Image, however demand for data in this space is accelerating as AI and Machine Learning applications become more sophisticated.
Although APX's revenues are project driven, many of these projects lead to recurring revenue. By APX's estimates, one third of data requires regular refreshes (weekly or monthly), with the balance requiring less frequent refreshes. This helps to provide a base level of “recurring” projects and revenue and also allows APX to be constantly engaged with its clients to win new projects.
APX has demonstrated a strong track record of earnings growth, reporting a 5-year CAGR of +60% to FY19, although this does include a number of acquisitions. EBITDA margins have plateaued around the high teens, and I expect a similar trend for FY20E as APX continues to reinvest in its business to take advantage of this fast growing industry.
APX has a relatively clean capital structure, reporting net cash of $75m at the end of FY19.
Impact of COVID-19
The company provided a business update in mid-April. To date, it has not seen any slowdown in digital advertising or IT spending and as such has reinstated its prior full year earnings guidance, which is very rare in this period.
APX is trading at a forward EV/Revenue of 4.8x which is favorable relative to other ASX tech peers.
I believe 4.8x multiple is a fair price for this high quality company with a very long growth runway.
This article was written by
Analyst’s Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.