Description

EPAM Systems is a high quality IT service provider that had been a compounder stock. We believe that EPAM is an AI beneficiary. 

EPAM's winning streak came to an end in 2022. 

With Russia’s invasion of Ukraine the stock fell from $550 per share to $180 per share as the market was worried about EPAM’s workforce/delivery centers in Ukraine, Belarus and Russia. The management at EPAM did a wonderful job derisking the company from this exposure. They exited Russia by not only leaving their clients there but also by closing their delivery centers in the country. By the end of 2022, EPAM’s exposure to these volatile countries had reduced from 60% of overall workforce to 30% of overall workforce. EPAM moved some of the affected employees to neighboring countries and ramped up their hiring efforts in India and Latam. Due to this, the stock moved up to $450 per share a few months later in August 2022.

However, very recently, EPAM reduced their guidance for 2023. They first reduced their guidance during their earnings but then followed this up with a further reduction in guidance a few weeks later. The management is now expecting negative growth in 2023.  

We first got involved with EPAM after it sold off in March 2022 but have recently added to the position. EPAM is not a ‘cheap’ stock in the VIC sense of cheap but it is very cheap given the quality of the business and the tailwinds we see due to AI. EPAM is currently available for 21x depressed 2023 earnings per share. It is a $12 bn market capitalization company with $1.7 bn in net cash and a $500 mn share repurchase authorization. An investor purchasing shares at these valuations can expect to compound with earnings growth with the added call option of multiple expansion if EPAM is once again seen as a durable growth franchise. We believe that is the most likely outcome. 

EPAM is a global IT services firm. It competes with the likes of Accenture and provides IT services to a diversified end-market consisting of clients in the financial services, travel, retail, software, healthcare among other industries. EPAM held its IPO in 2012, and since then, has grown more than 10x until the year ended 2022 with revenues of $4.8 bn and EBIT of $573 mn. Like all IT services companies, EPAM is well placed as a ‘toll booth’ between technology and corporations who increasingly need technology to transform their business. 

Here are the historical financials:

This stellar track record came to an end in 2022. For 2023, EPAM guidance that its revenues will de-grow from 2022 levels. Management mentioned that clients are deferring projects due to which there will be little or no growth for the next few quarters. They guided for growth to come back in 2-4 quarters as clients can pause these projects but not cancel them. 

What happened?

1. IT spending is indeed slow. Like many other industries, IT services also ‘over earned’ during covid and had a stellar performance in 2022. Some of this is now normalizing.

2. Some clients have de-risked from EPAM due to their still substantial exposure to Ukraine.

3. Overall, revenue growth in this industry is very dependent on employee growth. EPAM has not grown its net employees - the losses in Russia and Ukraine have been made up by gains in India and Latam. It takes time in this to train employees and bring them up to high levels of productivity. 

In addition to all these issues, there is debate amongst investors on AI  - Are EPAM and other IT services providers AI winners or AI losers? Most IT services are billed on a ‘time and material’ basis and the ‘bear’ argument is that as software engineers become more productive; fewer software engineers will be needed thus affecting growth for all IT services companies.

In this report, we will highlight EPAM’s strengths and debunk some of the narratives surrounding the thesis. We believe the risk reward on EPAM is very attractive. 

IT services and EPAM’s competitive advantages

All IT services companies look the same. They all provide bodies that specialize in technology to businesses that need to either cut costs or improve processes. Under the hood, however, they are all a little different. Understanding these nuances can be the difference between a value trap and a compounder.

According to Gartner, IT services is one of the fastest growing segments of overall IT spend. The IT service providers play in this bucket and grow due to three factors,

• Increasing IT spends: These spends are typically a function of IT budgets. These budgets are typically a certain percentage of revenues and can be economically sensitive. The trend is that most corporations are allocating a larger and larger part of their spends to technology - software is in fact eating the world. These spends have further tailwinds from Cloud computing, Cybersecurity, Work from home and Digital marketing.

• Increasing outsourcing: Larger businesses can afford in house technology departments but for most businesses technology is not their core expertise. It is also difficult to hire and retain good technology talent. So, they rely on consultants to keep them abreast of the latest technology trends. In the past, most companies let IT companies manage their costs but increasing IT services companies are helping businesses transform the way they do business. 

Confirming this, Gartner mentioned that, through 2025, organizations will increase their reliance on external consultants, as the greater urgency and accelerated pace of change widen the gap between organizations’ digital business ambitions and their internal resources and capabilities. 

The two biggest competitive advantages in this industry are scale and culture.

If you look at the history of IT services companies we observe that the big get bigger. Accenture currently has 700,000 employees. Why? Because scale matters in more than one way. 

On the revenue side it matters because an IT services company will not get a $10 mn project from a client until they have done a $10 mn project in that practice in their industry and can provide a good reference. Many times, an IT services business has to organically get to that level or acquire another company that has a good reputation in that practice and industry. The phrase, ‘no one gets fired for hiring IBM’ came from the fact that until you have a proven history or executing and become a low risk choice for customers, you will not get hired. EPAM has a reputation for being a good partner for corporates. 

On the cost side, scale matters as a company needs resources to train their workforce on the newest technology  - something that changes all the time. The successful companies were able to transition their business from application development to ERP implementation to cloud migrations and now they will have to transition to AI related technologies. All this requires significant resources and a margin buffer that only the larger IT service providers can absorb. 

Warren Buffett had been reading IBM’s annual report for 50 years before he bought the stock. He talked to Berkshire companies and reasoned that most companies do not switch their IT providers. He was absolutely right in his assessment of the industry - it is a very sticky business - but picked the wrong horse (he should have bought Accenture instead!).

EPAM is a founder-led Company. Culture is key. IT services are execution heavy businesses. Projects have to be delivered on time and on budget to satisfy the client. While projects start and finish, good IT services companies remain entrenched with the clients and endeavor to gain a bigger share of their IT budget.  The financial services industry, for example, has a constant need for IT services due to complicated back offices, changing compliance requirements, mergers and acquisitions, or digitizing services. An IT service provider needs to work hard to take its fair share of the client’s IT budget but the good ones do just that. Whatever the client’s need, the IT service provider has a solution. EPAM is good at this! Due to this, the business generates more recurring revenue and is much higher quality than the market realizes. 

With the IT services industry, there are some net winners and then some net losers. IT work is continuously getting commoditized. Companies such as Accenture or Infosys have a ‘legacy’ business that is facing pricing pressures and has negative growth as well as new ‘digital’ business lines that are growing like a weed. Newer IT service companies with very strong engineering capabilities such as EPAM, Endava, Globant, Luxoft (now acquired by DXC) have much more of these ‘digital’ revenues due to which they grow much faster but this also makes them more dependent on discretionary spend. 

EPAM’s historical core competency is full life-cycle software development services including design and prototyping, product development and testing, component design and integration, product deployment, performance tuning, porting and cross-platform migration. 

EPAM is well diversified within verticals (financial services, travel, business information & media, software, and life sciences).  It takes a lot of time and effort to methodically diversify your business in this sector and each vertical can be thought of as a new business line. EPAM also not does not have large concentration risks when it comes to clients with top 10 clients making up about 23% of revenues.

"I would say in my experience in working directly with EPAM is that they have a core engineering expertise that is top shelf. They can bring that to bear on whatever projects or programs that they're involved with. They seem to have a team of core scientific engineers that aren't necessarily dedicated to any particular project but are there to help in any fashion to accelerate or solve very difficult problems that come into play that may stall a project. Their ability to do that is pretty profound."

-EPAM Competitor Jan 2022

“In my opinion, they're probably the best at doing this in the industry. They far away are better than TCS, Cognizant, your typical offshore vendors. They're really a specialist at modernization of product stacks, which has been great. They have a really good practice around DevOps and modernization. They've been fabulous. I use them also at Shutterstock when we did something similar for AWS, and they were just as good.”

-CTO of Priceline.com (June 2023)

The other way to get scale and capabilities is acquisitions. Good providers do many tuck-in acquisitions to gain either expertise or clients. EPAM had done many successful acquisitions over its history and integrated them into their culture. We have not seen large acquisitions succeed in the industry due to cultural differences. 

"When they identified in the past that there's a capability that's going to be in higher demand in the market, and they lack the capability internally, they did the smart thing in my estimation. They went ahead and purchased one or two or several really strong players, and integrated them."

-EPAM competitor (Jan 2022)

We believe that EPAM is a superior business and is shielded from the common problems in the industry. Its services are essential to its clients' success. EPAM has proven to be a good IT services company and invested ahead of new technologies in order to stay relevant with its clients and increase its share of wallet. There is very little reason to believe that this will not be the case going forward. 

AI Narrative

We think IT services companies such as EPAM are AI winners. Indulge me here but I asked chat GPT on how AI can help businesses and here is the response:

“Here are some key areas where corporations can leverage AI for benefits:

• Automation and Efficiency: AI can automate repetitive and mundane tasks, enabling corporations to increase operational efficiency and productivity. 

• Decision Making and Analytics: AI-powered analytics can process vast amounts of data and provide valuable insights for decision-making. 

• Personalized Customer Experiences: By leveraging machine learning algorithms, corporations can analyze customer data, preferences, and behavior to tailor product recommendations, marketing messages, and customer support interactions. 

• Enhanced Customer Service: AI-powered chatbots and virtual assistants can provide instant and round-the-clock customer support. 

• Risk Management and Fraud Detection: Machine learning algorithms can analyze patterns in financial transactions, customer behavior, and cybersecurity data to identify potential risks, anomalies, or fraudulent behavior. 

• Product Development and Innovation: AI techniques like natural language processing and computer vision can assist in product development and innovation. By analyzing customer feedback, social media sentiment, market trends, and competitor data, AI can identify customer needs, suggest product enhancements, and support the development of new offerings."

And on and on and on…..

AI is great but how does a corporation incorporate AI into their business? It seems like machine learning is an integral part of the process. So, what does one need for machine learning? I again asked chat GPT:

“To effectively implement machine learning, several key components and considerations are necessary. Here are the main requirements for machine learning:

• Quality Data: This involves relevant, accurate, and properly labeled or annotated data. The data should be representative of the problem domain and cover a wide range of scenarios and examples.

• Data Preprocessing: Before feeding data into machine learning algorithms, preprocessing is often required. This includes cleaning the data by removing noise, handling missing values, normalizing or standardizing features, and performing feature engineering to extract meaningful representations.

• Machine Learning Algorithms: A variety of machine learning algorithms exist, including supervised learning, unsupervised learning, reinforcement learning, and more. Selecting the appropriate algorithm(s) for the specific task at hand is crucial. 

• Training Set: A labeled training set is needed for supervised learning, where the input data and their corresponding target values are available. The training set is used to train the machine learning model by adjusting its parameters or weights based on the provided examples.

The other key points mentioned are Model Selection and Tuning, Model Training, Model Evaluation, Deployment and Integration and finally, Monitoring and Maintenance”

In essence, what we read above is that there is a LOT of work that needs to be done in order to enable a corporation to reap the benefits of AI. All this work has to be performed by someone! We think EPAM and other engineering focused IT services providers will be the key enablers of this move to AI. Further to this, corporations that still have their data on premises on different platforms and rhey first need to bring it in one place and on the cloud to even begin implementing AI. This work is also performed by EPAM. 

There is an excellent transcript on Stream by the CTO of Priceline.com that sheds light on these factors:

“I've had prior experience working with EPAM in prior companies. They are what I'd say a best-in-class company for modernizing tech stacks and moving them to the cloud. We hired them to accelerate our move out of our data centers and into Google Cloud.”

“They were doing cloud modernization, I'd say, about 10 years ago before it was a big thing and now it's a big thing. They have really good relationships with Google and AWS. They bring EPAM into deals as well with customers. What winds up happening is for AWS or Google contracts, if you lock in to, say, a five-year deal worth $500 billion, Google then gives you what they call PSF, which is professional service funding, in order to make the migration to the cloud. They'll give you, say, $60 million, $70 million to do that. You can only spend that PSF money against certified Google partners. Quite honestly, there's maybe a half dozen that are really good and EPAM is one of those.”

“They're (EPAM) building out our real-time data infrastructure for us now to train and build out all of our machine learning models. They're doing that right now and they're well-positioned to do that. The generative AI stuff, we're going to do in-house just because we have a large data science population already and have been doing machine learning for quite some time. Any of the heavy lifting around modernizing our data infrastructure over the next couple of years, EPAM will be doing that work.”

-CTO of Priceline.com (June 2023)

The current narrative is that corporations are going to hire a team of software engineers and then with the help of chat GPT or Bard they will be successful in enabling corporations to build AI capabilities. We believe this is totally wrong. 

The IT services industry has suffered from these false narratives in the past. A few examples:

Initially, all IT work was application development but when SAP and Oracle brought packaged software the narrative was that IT services is done but IT services pivoted to implementing ERP and made out like bandits. Then cloud software started to become dominant and this software did not require a lot of implementation and the narrative again shifted to the death of the IT services business. But, IT services pivoted to cloud migrations and digitally enabled custom projects that were increasingly possible because of the flexibility offered by the cloud. We believe this current shift of narrative where investors believe that IT service companies are net losers from AI is misplaced. 

As long as IT budgets are increasing, the better IT service providers will find things to do. Cloud providers such as AWS and GCP bring EPAM into deals so that they can more effectively modernize tech stacks for clients! We believe AI is a net enabler and will lead to accelerated growth in revenues and earnings for IT services companies like EPAM.

Conclusion

EPAM remains a high quality founder led business that has hit a rough patch. We believe that this is a small bump and EPAM will get back to growth in a few quarters. 

Overall, the company can be had for 21x depressed 2023 guided earnings. This is on a non-gaap basis. Before 2022, analysts had estimates going to $16 per share in earnings for EPAM by 2025. While this is no longer possible, we model $14 per share in 2025 and $16+ per share in 2026. In addition to the reduction in growth, EPAM is also not optimized when it comes to margins. We believe it can move on to high teens margins in a more normalized environment. 

In all this, AI remains a big variable and we think the surprise is likely to be on the upside. 

This is a $320-400 stock in three years if they can do $16 in earnings per share in 3 years and get valued at 20-25x earnings. It is more likely than not that the multiple will be much higher if they do return to 20% plus growth. 

Importantly, EPAM has a strong balance sheet with $1.7 bn in net cash and can buy back more shares or do tuck in M&A to further enhance shareholder value. EPAM has a $500 mn buyback authorization in place which should help with dilution and earnings accretion. 

Economic cycle is a big risk. IT service providers are dependent on IT budgets which are then dependent on revenues. If there is a downturn in the economy and revenues come down - that will affect IT budgets. 

Catalyst

Return to growth

Narrative change around AI