Description

Thesis: Long Kingsoft Corp (3888 HK)

Kingsoft Corp (KS) is a Hong Kong holdco with four main assets, of which, its golden goose is the #1 software provider (also the #1 SaaS provider) in China, Kingsoft Office (aka. WPS). By investing through the holdco, we are getting all of the underlying assets with a 55%+ discount to fair market value when compared to KS’s market cap, which provides us with an ample margin of safety while owning one of the fastest growing, innovative, and durable SaaS companies in China. Our investment thesis is as follows: 1) WPS is arguably the strongest SaaS business in China with ~65% of revenue under a subscription model and increasing. WPS benefits from China localization efforts following the rise of China-US tensions and WPS has no meaningful domestic competitor. When compared to Microsoft today, WPS still has the minority market share thus providing ample runway for growth through share gains. 2) SOTP valuation indicates that KS is trading at a 55%+ discount, providing ample margin of safety. 3) KS has RMB9bn of unencumbered cash that can be used to narrow the holdco discount, which the management recently tapped into for a meaningful share repurchase program. As with any investment, there are investment risks to consider: 1) Management has at their disposal RMB9bn of cash they can burn away to the detriment of shareholders. Value may never be realized at the holdco level. 2) There are meaningful intercompany relationships among KS assets and interested parties. There is the risk that KS may sacrifice some of its earnings or unencumbered cash for the benefit of other interrelated companies. 3) One of KS assets, Kingsoft Cloud, is not a top-tier asset as it is subscale and burning cash. It represents ~15% of SOTP value. We have diligenced each of the core risks and believe that management’s track record will help to minimize the overall risk profile. Ultimately, our investment in Kingsoft Corp satisfies our three main criteria of, 1) a company with tremendously long runway for growth were we can see this business grow normalized earnings power at 30%+ for at least five to ten years, which will over time be reflected in the stock price, 2) investing with a meaningful margin of safety due to its SOTP discount of 55%+, and 3) backing a strong management team who is aligned with both shareholders and stakeholders.

Kingsoft Corp Overview

KS was founded by Qiu Bojun (求伯君) in 1988 focusing on office automation software (document processing, spreadsheets, and presentations). In the early 1990’s, Microsoft was looking to enter China by acquiring WPS, which ultimately did not happen, but resulted in Microsoft and WPS signing a perpetual agreement for their source code to be interoperable. Leijun (雷军) joined the company right out of college in the early 1990’s. Fast forward to 2010, Leijun founded Xiaomi. At this point of time KS had the following businesses: 1) A PC game developer (aka. Seasun/ Westhouse/西山居) built around a single IP called Swordsman (剑侠), which is a MMORPG. Playstyle is very similar to Diablo II and World of Warcraft. Revenue is generated through monthly subscriptions (time-based) rather than microtransactions (transaction-based). 2) A PC based internet security business which was shifting focus to mobile security and operations. This business was renamed Cheetah Mobile (CMCM) and was ultimately made obsolete by the Android operating system. 3) WPS, which at the time was still PC based.

Subsequently in 2011, the founder of KS retired and invited Leijun back as the new chairman. In the early 2010’s KS focused on 2 areas: 1) port WPS to mobile, and 2) build a standalone cloud business, which is now known as Kingsoft Cloud (KC). Fast forward to today, we have the following ownership structure; Leijun is the de facto owner of, 1) Xiaomi with ~33% share ownership, which is a completely standalone and separate entity to Kingsoft Corp, and 2) Kingsoft Corp with ~25% share ownership. Kingsoft Corp is a holdco that owns, 1) ~68% of Seasun, which is not separately listed, 2) ~52% of WPS, which is listed on China A-share markets under the ticker 688111, 3) ~40% of Kingsoft Cloud, which is listed on the US markets under the ticker KC, and 4) ~40% of Cheetah Mobile, which is listed on the US markets under the ticker CMCM. Only Seasun and WPS are consolidated while KC and CMCM are both recognized using the equity method. From a business operations perspective, each one of the subsidiaries are standalone entities without shared back-office functions with one another, although the holdco does benefit from some of these back-office functions.

Kingsoft Office (aka. WPS)

WPS is the crown jewel of KS and where we will spend the majority of our time as the other pieces add minimal value to the overall SOTP valuation. WPS has a full product suite that mirrors Microsoft Office (Word, Excel, PowerPoint, PDF, document cloud, and collaboration tools).

Business Overview

WPS’s largest segment is the 2C business (represents ~50% of revenue). This segment sells directly to end-consumers on PCs and smartphones. WPS utilizes the freemium model where the basic product is free, but ads will occasionally interrupt the workflow. More advanced features are placed behind the paywall as well. There are three types of memberships: 1) VIP membership which brings added functionality priced at RMB88/yr, 2) Daoke (稻壳) membership which gives unlimited access to over 100m templates priced at RMB88/yr, 3) Super VIP which combines the former two memberships plus additional benefits priced at RMB178/yr. The main benefit for VIP membership is the ability to convert documents and presentations back and forth from PDF. This functionality is quite valuable in China as there is not an abundance of free software alternatives unlike the US. Trying to find free software online may result in getting malware or viruses. In addition, VIP membership provides cloud storage and collaboration functionality. Daoke membership provides templates for any occasion, including presentations, resumes, documents, etc. Through our channel checks, we have learned that templates are a very China specific demand and makes WPS’s template database very valuable. Right now, WPS’s MAU is over 500m and paying ratio is ~4.5%. Based on our understanding of the product, and the historical improvement in penetration rate of roughly 100bps per year, we believe there is strong visibility into future revenue growth. Management has highlighted they expect a 40%+ CAGR for this segment over the next five to ten years. We believe this is credible, but have discounted it meaningfully through our bottoms-up build to ~30% CAGR for our five-year view. 2C segment gross margins is roughly 80-85%, typical for SaaS businesses where COGS represents bandwidth costs and internet infrastructure costs.

WPS’s next largest segment is the 2B business (represents ~40% of revenue). Roughly 1/3 of this segment is on an annual billing cycle (年场地授权). Under this format, enterprises (including government) pay for site licenses, which can be used by unlimited users. Because the enterprise is typically never truthful around the actual number of users, WPS estimates that it only receives revenue for roughly 10% of actual users. For example, a company has 100 users, but only tells WPS it has ten users and pays the annual site fees for ten users. This has led to extreme cases where WPS believes some enterprises are paying as low as RMB10/user vs. the enterprise rate of RMB599/user (typically discounted to RMB200-300/user for large enterprises and government). The way that WPS is looking to rectify this issue is by moving its customers to the cloud. By moving to the cloud, WPS can actively monitor how many concurrent users there are and can ensure that the enterprise is not freeloading with more than the agreed upon user count. This should over time increase the annual billing cycle revenue multiple folds. Roughly 1/2 of the 2B segment is project based for the central government, where the focus is on China localization and technological independence. This business model is quite unique and very important to understand. Together, the Chinese government and government SOEs operate roughly 60m PC units. The government has a mandate to replace Windows with Linux for national security reasons. This conversion has just started and over the next ten years will shift to 80% Linux, which equates to roughly 50m PCs. The pace of the conversion is fluid, but we expect it to ramp over the coming few years. WPS sells to the government on a perpetual license basis, but these licenses are tied to individual Linux licenses, which means that whenever Linux goes through version upgrades a new WPS license is required. This is where it gets interesting: Linux systems in China are far behind Windows in terms of functionality and need to be upgraded every two years with newer versions – Linux is at least two versions behind Windows. Incrementally, PC replacement rate is 3-5 years and a new WPS license is required for each replacement cycle. Each Linux machine will require a WPS license priced at RMB200-300 (discounted from an enterprise rate of RMB599). Over time, management’s goal is to get this business into the cloud to upsell additional products and to make it truly recurring revenue (although technically it is already biannually recurring revenue due to the Linux upgrade cycle). WPS is already testing Linux cloud products and will start migration in 2025 after Linux becomes fully compatible with cloud. Lastly, 10% of the 2B segment is legacy perpetual license sales with minimal ongoing maintenance or support revenue. This segment is already immaterial (10% of 40% = 4% of total WPS revenue) and we expect this to continue to shrink. Overall, the 2B segment has incredibly high gross margins as there are very little costs to incremental perpetual license sales.

Lastly, WPS’s smallest segment is the advertising business (represents ~10% of revenue). This segment is related to the freemium model for its 2C customers where ads will occasionally interrupt the workflow. WPS has started to decrease its ads inventory and wants to focus more on product quality and user experience, leading to a decline in ads revenue in 2020. Going forward we should see ads inventory stabilize, but become less and less important over time as other segments continue to grow. Roughly speaking the ads business has mid- to high-70% gross margins.

Competitive Environment

Microsoft is the largest and realistically the only competitor to WPS. WPS currently has a smaller share of wallet, but due to the government’s focus on China localization and through our channel checks, we believe there will be significant share gain opportunities over the next five years. For example, through our discussion with a large government SOE, Microsoft is currently 80% of their budget with the remaining going to WPS. They expect that within five years, WPS will represent over 80%+ of their budget. Our diligence indicates that within enterprise, 50-60% of share is still from piracy, followed by Microsoft with 20-30% share and WPS with 15% share. Current indicators show that WPS will gain share from all buckets (Microsoft and piracy). There is a similar share gain story for WPS in the SME segment where piracy is closer to 70-80%. Note that WPS is significantly cheaper than Microsoft, which will help with adoption. When comparing prices between WPS and Microsoft, WPS is roughly 65-80% cheaper across all subscription tiers (individual, business, and professional). From a product perspective, through our own primary product diligence and through industry experts, we have concluded that WPS and Microsoft are interchangeable in terms of functionality. There are some areas where WPS is better and others where Microsoft is better, but those are along the fringes. Lastly, WPS has been added to some national student examinations as of November 2020 and by March 2021 was in all national student exams. This point is crucial as it will train the younger generation to learn how to use WPS over Microsoft.

When comparing WPS to its collaboration peers, including: Bytedance’s Feishu, Alibaba’s Dingding, and WeChat’s Wecom, we conclude that these products serve a different requirement for the enterprise vs. WPS. Most of these products focus more on instant messaging and employee tracking/management (HR and OKR). They are more of a messaging tool instead of a document processing tool. If you need to work with anything official, these products do not satisfy user functionality requirements. For example, the collaboration document processing tools for these products are more akin to Microsoft Notepad or Paint vs. full function document processing tools where you can create professional formats and track changes. Furthermore, these collaboration products are not fully compatible with WPS or Microsoft, which makes them less useful outside of internal workflows (i.e. client facing). The collaboration peers understand their shortcomings and have partnered with WPS to white label WPS’s products as the underlying infrastructure for some of their products. Lastly, WPS has been focused on collaboration and created a product called WPS Docs, which has collaboration and cloud functionality, helping to future-proof its business. Although WPS Docs works better than Microsoft Teams, its user friendliness and collaboration capabilities are still lacking when compared to its collaboration peers. That being said, even though WPS’s collaboration functionalities may be weaker, it does provide a full function document processing product and you can easily share documents through links in WeChat which is much more convenient than using two separate products.

Total Addressable Market

The overall addressable market is substantial. There is no perfect way to quantify what is possible (a large portion is piracy), but no matter how we cut the cake we see, 1) a multi-decade runway for WPS, and 2) multiple times the revenue vs. what the business is generating today. Through our diligence, we understand that Microsoft generates roughly USD600m of revenue per year in China, of which USD200m is from the Windows operating system. The remaining USD400m is from Microsoft Office enterprise customers only. WPS’s total enterprise revenue is only USD200m. We estimate that enterprise customers are only paying 10-20% of their total true per-seat usage. Assuming WPS expands to 2/3 market share (over Microsoft) and overall enterprise penetration increases to 80%, we achieve greater than USD1.6bn of revenue (vs. USD200m today) on the enterprise side. On the consumer side, WPS currently has over 500m MAU out of a total internet population of over 1bn in China. WPS’s paying ratio is ~4.5% with an average annual ARPU of ~RMB60/yr. We know that the cheapest subscription package is RMB88/yr, which means that as user stickiness increases, they will use WPS for a greater portion of the year (i.e. paying for 12 months instead of only eight months). Penetration ratio has been consistently increasing at ~100bps per year. Benchmarking to other freemium subscription apps, WPS has line of sight to at least 10% to 15% paying ratio. Using somewhat conservative estimates, WPS’s consumer business can grow 5x+ in revenue over the medium- to long-term.

Management Team

The new CEO, Zhang Qingyuan (章庆元), who took over in early 2020, is a game changer for WPS. He graduated in 1998, started as a programmer at WPS in 2000, and has been with the company ever since. He was always focused on the R&D side of the business and prior to being promoted to CEO, he was the CTO. He spearheaded all of WPS’s core product offerings including, WPS for PC, WPS Mobile, and WPS Docs. WPS is following Microsoft’s playbook on management transition, which should benefit WPS tremendously over the long-term. Microsoft was founded by Bill Gates. He was replaced by Steve Ballmer, who was more focused on marketing and pushed the product into everyone’s hands. After everyone had a Microsoft product, Steve was replaced by Satya Nadella, who focused on customer retention through continuous product improvements, cloud, R&D, and interoperability. WPS has followed a similar path. Qiu Bojun founded WPS. He was replaced by Ge Ke (葛珂), who was more focused on marketing and through his relationships successfully pushed WPS into the enterprise and government channels. Now that everyone in China has access to WPS in some capacity, Zhang Qingyuan has taken over and has been driving a meaningful push on product and functionality. Since Zhang Qingyuan joined, he has already doubled the R&D team from 1,500 to 3,000 employees. In addition, he modified employee incentive structures and linked S&M and R&D employees’ KPIs together. This is a meaningful change as it pushes the R&D team closer to the customer therefore making the whole organization more nimble at responding to user complaints, which leads to an increase in the pace of product functionality improvements and ultimately, user stickiness and retention.

Valuation

Using a SOTP approach, KS is trading at a 55%+ discount to fair value. Historically, this discount has ranged from 40% to 60%, which means that we are currently at the higher end. While creating a position in KS with a meaningful discount provides us with a margin of safety, we do not underwrite to a narrowing of the discount to help us realize our required returns. We are predominantly focused on the growth in normalized earnings power for our underlying businesses. In this case, WPS (which represents the majority of KS’s value) has meaningful runway to grow normalized earnings power at 30%+ CAGR for at least five to ten years, which will over time be reflected in KS’s stock price. From a downside perspective, our SOTP discount provides us with ample margin of safety. Case in point, if we assume that, 1) WPS’s valuation multiple gets cut in half, 2) zero value for KC, 3) zero value for Seasun, and 4) 50% discount to the RMB9bn of unencumbered cash at the holdco level, we arrive at KS’s stock price today. Lastly, KS’s management has been quite prudent with their cash balance. So far, they have used the cash balance for the benefit of both stakeholders and shareholders through share repurchases. Management is well aware of the holdco discount as they keep track of this metric closely. When KS’s stock price dropped to its trough, management announced a share repurchase program of HKD1bn on August 30, 2021, which is quite meaningful! Over the last few months they have already repurchased HKD120m worth of shares (more than 10% of the program) at roughly HKD30/shr, which indicates a strong willingness from management to actively shrink the holdco discount.

Catalyst

1. Continued compounding of Kingsoft Office (WPS)

2. Continued share repurchases at Kingsoft Corporation