The Edge and Asia Analytica have embraced the use of GenAI in our workflow. For instance, the images we have included in our last few articles were created with the help of ChatGPT. The aim is to give readers a quick visual representation and summary of the subject of the article. AskEdge (a newly-launched feature by The Edge) is designed to help investors get a better grasp of the financials of companies they invest in — all the specially customised charts, tables and explanations in simple English as well as peer comparisons are available on demand, with just a few clicks of a button. More GenAI-powered features are in the pipeline. Our overarching objectives are: (i) the use of GenAI has made us more efficient and productive; and (ii) our readers gain more value, information and a better understanding, which is important as time is the most precious resource.
Over the past two weeks, we have explained some of our core investing philosophy, and how they are applied to our stock-picking process for the Malaysian Portfolio and global Absolute Returns Portfolio. To briefly recap, the basic tenet of value investing is buying stocks that are undervalued by the market relative to their intrinsic worth, on the expectations that prices will eventually reflect their fundamental worth. One typically starts by sieving the universe of stocks using a range of filters (for example, valuation metrics such as price-earnings, price-to-book, dividend yields and gearing) and/or algorithms. This is then followed by a thorough fundamental analysis of the financial statements for the short-listed companies; that is, sales, margins, profits, cash flow, assets and liabilities plus a qualitative assessment of management capability, integrity, corporate governance as well as growth prospects and so on. In a nutshell, value investing involves the rational analysis of the trade-off between risks and potential returns (the difference between the prevailing stock price and its expected intrinsic worth).
Most value investors are bottom-up investors. However, there are inherent limitations to the bottom-up approach. For starters, it is difficult to find “green shoots” and truly undervalued stocks because there are many other investors doing the same. The stocks that appear “undervalued” are likely because they have reasons to be so, such as questionable management and lack of trust. And given that much of the filtering process — the fundamental analysis of cash flow, earnings, dividends and balance sheets — is necessarily based on past performances, it misses new discoveries, innovations, technological applications and other significant changes that will have sustained effects on future earnings … and, therefore, future stock values. In other words, the bottom-up approach to value investing is backward-looking rather than forward-looking.
We are value investors. However, our approach is to pair value investing with a topdown approach. Basically, we start with a macro analysis of the economics, trends, new ideas and innovations, lifestyle changes and so forth. And which countries stand to benefit, which industries, the effects of new technologies and so on. In other words, we first decide on the theme and then choose the stocks using value investing methodology.
The biggest theme right now is, without question, artificial intelligence (AI) and, more specifically (at least for now), generative artificial intelligence (GenAI). The biggest beneficiaries thus far have been the “pick-and-shovel” tech companies; that is, those providing the infrastructure — the chips (for example, graphics processing units and tensor processing units), equipment (lithography machines for chipmaking), data centres (cloud and edge platforms) and foundational technologies (such as large language models or LLMs). This group includes the likes of Nvidia, ASML, Amazon, Microsoft and Alphabet — unsurprisingly, the leaders of the US market rally in the past two years — and of course, OpenAI (the company that gave us ChatGPT). We think that much of the near- to medium-term prospects are already reflected in their share prices and elevated valuations. In fact, we believe reality will temper current exuberant expectations. We are negative on Nvidia, for example.
Our stock picks for the AI theme are, therefore, more realistic and focused on the next level of beneficiaries — the AI-powered software platform and solution providers as well as the broad spectrum of end users across the economy. That is, the enterprises employing AI tools to transform existing processes, enhance effectiveness, boost operational efficiency and productivity, drive innovation, create new products-demand-markets and so on.
We think the widely anticipated deregulations under the second Donald Trump administration will accelerate AI advancements and adoption. And the early adopters will gain competitive advantages — better products and higher margins on the back of increased revenue and/or lower costs — and market shares in their industries. The applications are wide-ranging, from improved ad targeting for social media and digital marketing platforms to AI-embedded smartphones and consumer tech, in surveillance and defence systems, education platforms, autonomous vehicles and drones, robotics in manufacturing and warehouses, diagnostic and drugs discovery, wealth management and many, many more, including those in the future that we have yet to even imagine.
Software platform and solutions providers
Our choice of software platform and solutions provider is Palantir Technologies, which operates a software-as-a-service (SaaS) subscription-based business model. Palantir specialises in big data analytics. Its software platforms, such as Gotham and Foundry, enable customers in the commercial sector as well as government agencies — notably, in defence, intelligence and counterterrorism — to transform massive amounts of information into an integrated data asset. We see the company benefiting from the rapid advancements in GenAI and machine learning in two main aspects.
See also: Part 2 — Investing is always about valuations based on intelligent and rational assumptions
One, integrating GenAI into its existing platforms enhances their performances through machine learning (learn and improve over time without manual intervention) and, therefore, value to customers. For instance, the company started deploying its latest Artificial Intelligence Platform (AIP) in 2023, integrating GenAI models (including LLMs) into its existing platforms. By focusing on cognition and the ability to gain and comprehend knowledge, Palantir helps users identify patterns and connections within large datasets and thus, make better data-driven and faster real-time decisions.
In other words, Palantir has improved its competitive advantage — its platforms are both predictive (integrating, managing and structuring big data) and prescriptive (analyse relationships, logic, simulations in the context of real-life internal and external conditions and execution).
Leveraging GenAI internally also reduces operating costs by automating its own workflow processes. For instance, generating and customising customer proposals and presentations, AI chatbots for routine customer assistance, coding, speeding up the research and development cycles (exploring datasets and generating hypotheses) and so on, all of which will allow it to scale more effectively — and boosting margins.
And two, the AI revolution is underway and the divide between the AI haves and have-nots will accelerate the winner-take-all AI global economy. As we mentioned above, early adopters will gain significant competitive advantages over their peers. Palantir gave an example of how this works for one of its customers: “We have automated the insurance underwriting process for one of America’s largest and most well-known insurers with 78 AI agents, taking a process that took two weeks to three hours. More than the labour savings, this presents the customer with an asymmetrical advantage in the marketplace to buying contracts before the competition has even gotten through 15% of their process.”
In short, we foresee a race among enterprises (end users) to integrate AI into their businesses and decision-making processes. Similarly, the US military is increasingly adopting AI-driven solutions across its complex to improve advanced capabilities and maintain its strategic edge against adversaries.
To be sure, there is also a certain degree of speculation that has driven Palantir’s share price sharply higher since the Nov 5 US presidential election. Its co-founder, Peter Thiel, is a known supporter of Donald Trump. The perceived close ties between the company and the incoming administration fuelled market speculation that Palantir will win more government jobs from expected Trump policies such as increased federal spending on national security, defence, and immigration enforcement. Government agencies accounted for 54% of the company’s revenue in the last 12 months (see Chart 1).
For more stories about where money flows, click here for Capital Section
Incidentally, Palantir obtained a higher rating for secure cloud computing services from the US government this month. The higher authorisation means the government can now process the most sensitive unclassified workloads on the company’s platforms, thereby further strengthening Palantir’s positioning in the public sector.
All of the above bodes well for demand growth. Palantir turned profitable for the first time in 2023, on the back of increased sales since integrating GenAI into its platforms. Rising adoption by both enterprises and the government will underpin growth for the foreseeable future while the scalability of its software platforms will translate into further margin expansions and earnings growth.
CrowdStrike is another software platform and solutions provider in our Absolute Returns Portfolio. The company is a leading global player in cybersecurity. Like Palantir, it operates an SaaS subscription-based business model. The company’s flagship product, the CrowdStrike Falcon Platform, is a cloud native (meaning it operates entirely in the cloud, making it highly scaleable and can be deployed quickly and efficiently), AI-powered cybersecurity solution designed to prevent breaches, detect threats and respond to security incidents in real time. It also provides endpoint security tools to detect malware on laptops, mobile phones and other devices that access corporate networks.
We acquired the stock back in August 2024. Its shares had been beaten down sharply after a widespread IT systems outage around the world due to a glitch in one of its routine updates. Most investors are short-term oriented, emotionally driven by knee-jerk reactions to events. And this is something longerterm investors can exploit and profit from. For sure, the incident has some negative impact on the company, including delays in signing contracts and expectations that customers will seek price discounts when renewing contracts to help cover the cost of the business disruptions. Nevertheless, as we wrote back then, the outage was not caused by a security breach, which is the most critical factor for a cybersecurity firm. Its shares have since recouped lost ground, though still below the all-time high. Neither Palantir nor CrowdStrike is cheap by most traditional valuation yardsticks. But we see similar robust longterm growth stories and competitive advantages for both companies in their respective industries. Sustainable earnings and cash flow growth are the most important variables in fundamental stock valuation.
Demand for cybersecurity solutions by both enterprises and government agencies will only increase over time, with the need to protect the integrity of systems and data from ever-growing threats from hackers and increasingly sophisticated cyberattacks. Case in point: Existing customers are increasingly spending more by adding on new modules. The total addressable market (TAM) for AI-native security platform is estimated to increase from US$100 billion in 2024 to US$225 billion by 2028. CrowdStrike’s revenue in the latest 3Q2024 crossed US$1 billion for the first time in the company’s history and annual recurring revenue (calculated as the annualised value of the prevailing subscription contracts, assuming any contract that expires during the next 12 months is renewed on existing terms) topped US$4 billion. The company is cash flow positive (see Chart 2).
CrowdStrike describes its platform as “the industry’s most complete AI-native defence, trained on the world’s highest-fidelity security data and augmented by the company’s elite threat hunters, incident response experts and the best managed detection and response service”. The company applies AI and machine learning to identify and stop the most advanced emerging attacks, creating new Indicators of Attack (IoAs) at machine speed and scale. This is critical, given that cyberattacks, also aided by AI, too are faster, stealthier and more sophisticated.
There is a veritable arms race between hackers and cybersecurity platforms. We see CrowdStrike gaining from growing network effect — the more its protection measures are tested, the more data (security events) is fed into its ecosystem and the more intelligent its platforms become, strengthening the company’s competitive advantage in the cybersecurity market and enhancing value proposition to customers.
Its GenAI security analyst, Charlotte AI (launched in May 2023), is designed to help customers better understand the threats and risks, shorten investigative times, make better and faster decisions and stop breaches while reducing security operations complexity — all using simple plain language (no technical expertise required). Detection and response can be automated at scale, improving the overall effectiveness of the customer’s cybersecurity efforts.
Applying Ai-driven tech to improve efficiency and productivity
Uber Technologies is one of the early adopters of GenAI advancements across its operations, including in its ride-hailing, food delivery and out-of-app (OOA) personalised marketing businesses. The integration of GenAI — it uses both open- and closedsource models from OpenAI, Google and other third-party providers — into its centralised machine learning platform (Michaelangelo) is the latest stage of the company’s continuous development process. The evolution started from simple tree models in its early years (2016-19) to advanced deep learning models (2019-23) and most recently, GenAI. Basically, the company fine-tunes the LLMs by leveraging its own massive proprietary data, to achieve high levels of accuracy and performance on Uber-centric tasks, at a fraction of the cost and lower latency.
The company has a dedicated AI arm (Uber AI) where it uses AI-powered innovation, technologies, research and its applications to solve challenges across the whole of Uber. The primary objective is to improve internal efficiency and productivity, streamline operations with automation and enhance user experience to drive revenue growth, raise returns on assets and equity.
For example, the company employs machine learning algorithms in its ride-hailing operations to detect fraud, forecast demand patterns, route optimisation, estimated time of arrival (ETA) computation and dynamic pricing for optimal resource allocation. This ensures that supply meets demand efficiently, reducing wait times and improving customer experience — and, therefore, usage and revenue. After the trip is completed, machine learning aids in payment fraud detection, chargeback prevention and powers the customer service chatbot. Uber currently operates in more than 70 countries spanning over 10,000 cities, serving 25 million trips on the platform each day with 137 million monthly active users.
The same philosophy is applied to Uber Eats, where an AI assistant — powered by LLM and fine-tuned with its user database — provides personalised recommendations by analysing past behaviour and preferences, accurately predicts delivery time based on real-time data and efficiently completes the delivery to enhance customer engagement and satisfaction and builds trust.
As with Palantir and CrowdStrike, the adoption of GenAI also enables the company to scale more efficiently. And its efforts have borne fruit, as evidenced by the improved sales and earnings. Uber has been profitable for the last five straight quarters on the back of rising revenue. Its share price has rallied from the lows in mid-2022, driven by positive sentiment from its growing profitability, returns on equity as well as growth prospects (see Chart 3).
Looking ahead, Uber is laying the foundation for future expansion by incorporating autonomous vehicles (AV) and delivery robots into its businesses aimed at further reducing costs, strengthening its competitive advantage and market share — driving sales, margins and earnings. Uber currently has numerous different AV partnerships in domestic and international markets, including with Waymo, as well as a collaboration with Serve Robotics, the developer of AI-powered autonomous sidewalk robots for food deliveries on its platform.
Summary
None of the three stocks — Palantir, CrowdStrike and Uber — would have popped up using the typical value investing filters, which are mostly based on historical financials and traditional valuation metrics. The problem with a bottom-up approach is that it relies too heavily on past performances while value is the sum of future discounted cash flows. So, our approach is top-down value investing. We decide on the theme and then choose the stocks using value investing methodology.
We do not disagree that there is currently a lot of hype surrounding AI. While some of the well-known brands or companies may be excessively overvalued, the AI revolution is real and underway. Its effects are also very real. The three companies we highlighted here clearly showcase how they have enhanced their products and services by leveraging AI technologies. That gives them a competitive advantage over competitors that have not incorporated AI into their operations and strengthens their positioning in the global market.
GenAI and machine learning will keep accelerating innovation and betterment of products and services, creating a powerful network effect. We can envision a future where enterprises with better AI-powered products will win more customers, contributing more data to their ecosystems, which improves the training of GenAI and machine learning models at a faster pace to create even better and differentiated products, resulting in higher barriers to entry. Economies of scale, especially for highly scaleable software platforms, will mean greater competitiveness and market share. Proprietary data can be an enduring competitive moat. We may well end up in a winner-takes-all (or just a handful of mega enterprises) AI-driven world. The US and China are the clear leaders in embracing AI technologies, driving innovation. American and Chinese enterprises will dominate in each of the two emerging ecosystems, with the tech decoupling. Where does that leave the rest of the world and especially smaller economies, in terms of future competitive ness, investments, jobs, incomes and growth? This is a topic for a future article. Stay tuned.
The Malaysian Portfolio gained 0.4% for the week ended Dec 18, outperforming the benchmark FBM KLCI, which fell 0.2%. United Plantations (+7.0%), Kumpulan Kitacon (+3.4%) and Insas Bhd – Warrants C (+3.3%) were the top gainers for the week; while the three losing stocks were IOI Properties Group (-5.1%), UOA Development (-2.7%) and KSL Holdings (-2.3%). Last week’s gains lifted total portfolio returns to 206.4% since inception. This portfolio is outperforming the benchmark FBM KLCI, which is down 12.6% over the same period, by a long, long way.
The Absolute Returns Portfolio, however, fell 1.8%, reflecting the broader US market selloff. The market bellwether Dow Jones Industrial Index has fallen for 10 straight trading days, its worst losing streak in 50 years. As we have highlighted previously, investor expectations and valuations are elevated, making stocks vulnerable to any negative news flow. Last week’s selloff accelerated after the US Federal Reserve pared back its interest rate cut projections for 2025, on the back of higher inflation expectations amid stronger-than-anticipated economic growth and low unemployment rate. Last week’s losses pared the portfolio’s total returns since inception to 16.7%. The top gainers were MAP Aktif Adiperkasa (+2%), Swire Properties (+1.3%) and OCBC (+0.6%). Talen Energy (-7.2%), CRH (-5%) and CrowdStrike (-3.8%) were the biggest losers for the week.
Disclaimer: This is a personal portfolio for information purposes only and does not constitute a recommendation or solicitation or expression of views to influence readers to buy/sell stocks, including the particular stocks mentioned herein. It does not take into account an individual investor’s particular financial situation, investment objectives, investment horizon, risk profile and/ or risk preference. Our shareholders, directors and employees may have positions in or may be materially interested in any of the stocks. We may also have or have had dealings with or may provide or have provided content services to the companies mentioned in the reports.
