• An epic market overreaction. A ~35% AI-driven sell-off ignores strong double-digit RPO growth, proving enterprise software demand remains highly resilient.

• “Systems of record” are the ultimate moat. Foundational data platforms and cybersecurity remain indispensable, as AI agents rely entirely on their proprietary data and strict guardrails.

• A rare margin of safety. The current valuation compression offers a prime opportunity to accumulate high-quality enterprise SaaS while avoiding vulnerable creative software.

⚠️Disclaimer: Word of Caution!

Please DO NOT take this educational post as a buy or sell signal. When it comes to investing, it is important to have your own judgement. Despite my detailed analysis, mistakes may occur, and blindly following could lead you to make similar errors and financial losses. Furthermore, I AM NOT a licensed financial advisor. I’m merely sharing my experiences and opinions only.

Additionally, please note that I hold positions in the discussed stock, and my view may be biased as a result.

Recently, there has been a crash in US software companies. My holdings in software stocks such as MSFT, PANW, SNOW, NOW, FDS, and SNPS are all showing negative returns this year.

When I look at the software sector ETF (IGV), which I use as a proxy for the broader software market, it is down ~35% since October 2025. This firmly places the sector in bear market territory:

The market dubbed this "SaaSpocalypse". As a long-term investor, the last thing I want to do is “panic”. Instead, I should find answers to 2 critical questions:

• What triggered this sell-off?

• Is this a permanent fundamental deterioration or a market over-reaction?

#1: What Triggers The Sell-Off?

Here’s the timeline of the sell off:

Mid-Oct 2025: Sector Rotation

After the massive bull run from the 2022 bear market lows, tech stock as a whole (including software) were trading at historically stretched forward PE multiples. The market seems to think it has become bubble. As such, many started to rotate out of tech stock.

Below is S&P 500 Index 3-months performance to time time of writing this post:

Sector such as basic materials, energy, industrials and consumer defensive were all being favored by the market. In my opinion, it was a healthy pullback but it was not big enough.

Late-Jan 2026: AI Capex & ROI Fears

As Q4 2025 earnings rolled in, the market suffered severe sticker shock. Alphabet, Amazon, Meta, and Microsoft revealed intentions to spend a combined $680 bn on AI Infrastructure in 2026, a staggering 70% increase from estimates in 2025.

• Alphabet: Guided about $175 - $185 bn capex in 2026. This is nearly doubled the amount in 2025.

• Amazon: Guided about $200 bn capex in 2026, 53% increase from last year.

• Meta: Expect $115-$135 bn capex in 2026. This is also nearly double from last year level.

• MSFT: Expect to spend $75 bn capex in first half of 2026. The annualized amount points to a $140 bn.

• Tesla: Expect capex for 2026 to exceed $20 bn, up significantly from $8.5 bn last year.

• Apple: Expect to spend around $13 bn, which is similar to last year. They are the only outlier but not surprising since they leverage of others to power its Apple Intelligence features.

This spending spree has led to negative impact on companies’ free cash flows (FCF). In the past, the market likes these tech stocks because they are asset light that can generate steady FCF. In turn, these FCF is use to buyback shares.

This narrative seems to have changed. Many of the Mag 7 is now taking on debt to further fund their AI Infrastructure spending. The realization that monetizing AI takes longer than anticipated have caused the market to sell-off aggressively.

3rd Feb 2026: Agentic AI Disruption (The Knockout Blow)

The third trigger point is when Anthropic released new plug-ins for its Claude “Cowork” agent. These plug-ins are tailored for Legal, Sales, Marketing and Financial analysis. When integrated with legacy systems, the plugin can autonomously perform actions as told.

This rattled the market because if AI agents can do more work with fewer people, traditional software that charges per user (or “per seat” basis) may see demand growth slow or worse - a reduction in revenue. It erodes the core business models of legacy SaaS companies, data providers, and workflow tools.

IGV dropped another 11% over the next 3 days to 5th Feb. At the time of writing, the ETF has been consolidating within range $76.3 and $86.1.

Late-Feb 2026: The Cybersecurity Shock…

As if this was not enough, Anthropic launched “Claude Code Security”. It is an enterprise-focused AI capability built into the Claude Code platform that scans entire codebases for security vulnerabilities and automatically suggests targeted software patches to fix them.

Traditional cybersecurity tools rely heavily on rigid rules and matching code against databases of known vulnerability signatures. In contrast, Claude Code Security reads and reasons through code contextually, allowing it to identify complex "Zero-Day" vulnerabilities, such as flaws in business logic or broken access control, that traditional methods often miss.

Consequently, this triggered a violent sell-off on big cybersecurity stocks like Crowdstrike (-19%), Zscaler (-17.2%), Okta (-16.8%), Palo Alto (-10.4%), etc.

#2: Over-reaction or Structural Problems?

To understand whether the software sell-down are justified or not, one must understand how Claude AI works.

How Claude AI Works?

When analysts say “AI agents will disrupt software,” they are often vague and lack details as to how. What has actually changed is that Claude can now connect to a company’s systems and execute repeatable workflows autonomously using two key features:

⛓️‍💥MCP: the “USB port” for enterprise data

Claude’s Model Context Protocol (MCP) provides a standardized way for its AI model to connect to company’s systems (e.g. databases, file drives or internal tools). Just like a “USB port” where any operating system can access the file as long as a compatible USB drive is connected to this USB port.

This MCP reduces the need for custom built application programming interface (APIs) to integrate between two or more systems.

📜Skills: packaged “SOPs” (Standard Operating Procedures)

Claude’s “Skills” are lightweight, reusable workflow packages. They teach the AI how to do a specific job without requiring a long, complex prompt every time.

If MCP provides access to the kitchen & tools, "Skills" are the step-by-step recipes that teach AI how to use those tools effectively.

Anthropic has pre-built some skills (e.g. creating docs, presentation slides, extract data from PDFs or spreadsheets, etc.) and allows developers to create custom ones, potentially creating a marketplace similar to Apple's App Store.

🤖Synergy: MCP + Skills = An Autonomous Worker

Combined, they create a digital employee because:

• Skill defines the workflow (“what good looks like”, risk rules, output format, etc.).

• MCP pulls data and pushes actions into systems (databases, CRM, document storage, etc.) as required by the Skills.

AI isn’t just answering questions. It now can retrieve, decide and execute an action requested by the user.

How Does This Disrupt Software Companies?

For decades, software like Salesforce, ServiceNow or Snowflake monetize 2 big things:

1. Integration between systems (custom APIs)

2. Per-seats (per-user pricing)

Claude AI pressures both:

🔀Reduction of Middleware

In the past, if an ERP system like Oracle JDE needs to talk to HR management system like Workday, one must hire consultants or buy middleware. It is expensive because these require custom built for every system.

This type of integration are hardcoded meaning there is a defined workflow:

“When event A happens in Salesforce > call API B in NetSuite > map fields > write to table C”

Any changes in rules will need to rebuild the whole integration logic.

With the MCP, the AI can now connect to multiple systems via a standard interface and complete the workflow end-to-end autonomously.

User just need to specify an objective, example:

“Generate a renewal risk report and create tasks for the account team”

The AI decides:

• which tools (Skills) to call,

• what data sources to query via the MCP,

• how to transform/ validate data,

• and what actions to execute

This is known as “dynamic orchestration” because the path can change depending on what the AI agent finds. It compresses the value of middleware.

📉Reduced Demand for Seat-based SaaS

Because AI agent is now able to access company’s system of records and perform actions autonomously, companies now require less employees to perform certain tasks.

Here’s an example:

With the use of MCP & dedicated “sales skills”, company can deploy AI agent that autonomously carries below tasks:

• Draft personalized marketing emails based on customer’s past sales transactions.

• During deal negotiation stage, it able to connect into company’s sales database & update the deal stage, logs the summary and flags follow-up actions in real-time.

Traditionally, these tasks are done manually by employees.

As companies downsize, the “per-seat” pricing model used by SaaS giants like Salesforce may face significant revenue compression.

But…How True Are These? 🤔

Reading about these disruption rationale by various analysts seems to make logical sense. However… looking at the recent earnings of these SaaS companies tells a different story:

• ServiceNow reported its Q4 2025 subscription revenue, growing 21% yoy. Its remaining performance obligation (RPO) is at $12.85 bn, up 25% yoy.

• Salesforce reported its Q4 2026 subscription revenue, growing 13% yoy. Its RPO is at $72.4 bn, up 14% yoy.

• Adobe reported its Q1 2026 subscription revenue, growing 13% yoy. Its RPO is at $22.2 bn, up 12% yoy.

• Factset reported its Q1 2026 revenue, growing 6.9% yoy. Its annual subscription value grew to $2.4 bn, up 5.9% yoy.

• Snowflake reported its Q4 2026 revenue, growing 30% yoy. Its RPO is at $9.77bn, up 42% yoy. Additionally, there’s 740 net new customer additions representing 40% yoy growth.

All of these companies are experiencing revenue growth rather than declining as alleged above. More importantly, their RPO — the total value of all signed service contracts yet to be delivered — is still growing.

My Verdict: Over-reactions 😱

The current sell-down is clearly a result of the market being panic. Market tends to sell first and ask question later. A nice way of describing this is “The market is a forward-looking mechanism”. Investors are not selling because of companies’ recent earning results.

They are selling because they believe AI agents will take over white-collar tasks and companies will refuse to renew thousands of software licenses over the next 3 to 5 years.

🔴Misconception of AI

However, I think people tend to forget below key concepts of AI and AI agent:

1. 🎲AI is a probabilistic model. It does not guarantee 100% accuracy. Instead, it generates responses by determining which answer is statistically most likely to satisfy the user's intent.

2. 🗃️AI agents require context to function effectively. This context is pulled from “systems of record” or “internal databases”; without this grounding, the agent cannot execute meaningful workflows.

Because of the above, AI can make mistake and also hallucinate. Example:

Claude AI “Skills” feature attempt to make the AI more deterministic by providing a step-by-step instructions to AI agents. However, this is still at the early stage.

According to a survey done by Prosper Insights & Analytics, most consumers are not yet comfortable with fully autonomous AI agent, especially in high stakes or emotionally sensitive scenarios.

It further stated that “AI cannot deliver consistent performance without human oversight, transparent governance, and clearly defined guardrails”. In other words, AI agent’s work requires human oversight.

I believe AI agent will change workers’ job role from “doer” to “reviewer”.

🔴SaaS Companies Are Evolving Too…

Additionally, SaaS companies are not sitting still. They have been actively defending their moats since 2023:

• Salesforce (Late 2024): Launched Agentforce and Data Cloud, embedding native AI into the CRM for superior, real-time data context that external agents lack.

• ServiceNow (Late 2023): Deployed “Now Assist” natively within its CMDB, enforcing strict access controls and data privacy that third-party agents cannot guarantee.

• Adobe (Late 2023): Released Firefly, trained solely on licensed content, providing enterprise clients with critical legal indemnification against copyright risks.

• FactSet (2023–2024): Embedded proprietary financial data directly into enterprise data warehouses, ensuring it remains the mandatory “ground truth” for any AI modeling.

• Snowflake (Mid-2024): Launched Cortex AI, allowing clients to run LLMs securely inside their own data perimeter, eliminating the risk of data egress to external agents.

That said, I think AI will deeply penetrate creative sector precisely because art is inherently subjective. Since there is no objective right or wrong in artistic expression, AI-generated work can be just as valid as human creation.

SaaS companies that are in this sector such as Adobe, Canva or Figma could face a more severe threat than others. As an investor, I’m steering clear from these companies until they can prove their defensive actions are working.

#3: Conclusion

The "SaaSpocalypse" highlights a profound disconnect between Wall Street’s panic and fundamental business reality. While the ~35% sector drawdown assumes AI agents will instantly vaporize "per-seat" software, the double-digit RPO growth of major SaaS players proves enterprise customers aren't leaving.

If anything, this sell-down provides an opportunity for me. Here’s how I plan to position my portfolio:

• Buy the “System of Record”: AI is the engine, but proprietary data is the fuel. Companies like MSFT, NOW, SNOW, and FDS provide the mandatory "ground-truth" context AI needs to function. They remain indispensable.

• Hold Cybersecurity: The violent sell-off in names like PANW is short-sighted. Given the massive cost of security errors, AI will augment these platforms, but human oversight and deterministic guardrails remain non-negotiable.

• Avoid the "Subjective" Danger Zone: Because art lacks an objective "right or wrong," creative SaaS faces genuine disruption from AI generation. I am avoiding this sub-sector until clear defensive moats emerge.

Until my next long-winded post! 🙂

Disclaimer: Word of Caution!

Please DO NOT take this educational post as a buy or sell signal. When it comes to investing, it is important to have your own judgement. Despite my detailed analysis, mistakes may occur, and blindly following could lead you to make similar errors and financial losses. Furthermore, I AM NOT a licensed financial advisor. I’m merely sharing my experiences and opinions only.

Additionally, please note that I hold positions in the discussed stock, and my view may be biased as a result.

• A regulated monopoly with predictable cash flow. TNB owns the National Grid, and its return are locked in under the IBR framework

• Key beneficiary of Malaysia’s data centre and energy transition boom. Record capex in RP4 expand its asset base that will translate into higher returns.

• Dividend stock with growth potential. With improved fuel cost adjustments under RP4, dividend visibility and upside look stronger going forward.

⚠️Disclaimer: Word of Caution!

Please DO NOT take this educational post as a buy or sell signal. When it comes to investing, it is important to have your own judgement. Despite my detailed analysis, mistakes may occur, and blindly following could lead you to make similar errors and financial losses. Furthermore, I AM NOT a licensed financial advisor. I’m merely sharing my experiences and opinions only.

Additionally, please note that I hold positions in the discussed stock, and my view may be biased as a result.

In my previous post, I’ve shared in-depth about Malaysia’s energy landscape. If you haven’t read about it, you can read it here. In this post, I want to talk about Tenaga National Berhad (TNB) and why I think this company is a good dividend-paying stock.

🏢Type of Company

TNB is a utility company. It sits indirectly within the AI value chain because AI runs on data that is stored and processed in Data Centres (DC) which are extremely power-hungry. This means rising data centre activity in Malaysia should translate into higher electricity consumption. As such, I see TNB as a key beneficiary of the local DC boom.

Additionally, the company is also a dividend-paying stock. It has a dividend policy to distribute 30% to 60% of its adjusted PATAMI. Since 2018, it has been paying out dividend consistently at above 50%. This is despite the fact that there is a pandemic in 2020. Below is TNB’s historical dividend per share in cents (excluding special dividend) from 2014 to 2024:

Based on the above, this is a “Category 2” type of dividend paying company as described in my book. It pays a constant dividend that doesn’t increase yearly over past 10 years – it is not a straight line up.

However, I think this will change soon. I will explain more in the growth catalyst section.

💡Business Model

TNB makes money through several different “buckets”. One part of their revenue is regulated by EC (remember the ARR?) and the other part is unregulated. The company also has businesses overseas which are mainly related to renewable energy. However, it only accounts for 4% of total revenue in 2024 - not worth looking into it in detail.

Here’s an overview with revenue breakdown as of 2024:

⚡️Bucket #1: Sale of Electricity (The “Monthly Bill”)

This is the monthly electricity bill to consumers at a specified tariff rates based on the IBR framework. In 2024, about 96% of TNB’s total revenue came from electricity sales.

The formula for this tariff rate is as follows:

The base tariff is the rate that TNB use to bill the consumers. Note that not all of revenue collected ends up in TNB’s pocket.

A large portion of this revenue aim to recover the operating costs as depicted in the above formula. These are costs incurred by other regulated functions such as Single Buyer (SB) and Grid System Operator (GSO).

TNB’s true earnings as per the formula above is dependent on below 3 components:

• Regulated Asset Base (RAB) which is the National Grid asset value

• New CAPEX for power grid upgrades & maintenance

• Allowed rate of return which is the WACC

In a nutshell, TNB’s profit is not based on how much electricity consumers use. Instead, it is on how much it spent in upgrading & maintenance of the National Grid.

The bigger the spending (capital), the higher the RAB. This then translates into higher returns for TNB. I will explain more in the growth catalyst section later.

This bucket also includes an unregulated business which is the power generation sale to SB. As explained in previous post, the revenue from this business is dictated in the service agreement between TNB and the SB.

👨‍🔧Bucket #2: Construction of Electrical Systems

Sometimes, TNB is hired to build power-related infrastructure for other companies or the government, like setting up a new substation or moving giant power lines for a new highway.

This is a competitive business where TNB bids for work just like any other construction company. In 2024, this bucket of business only contributes 0.2% of total revenue.

🔌Bucket #3: Customers’ Contribution (“The connection fee”)

Whenever a developer builds a new building such as shopping mall, housing estate or data centers, they have to pay TNB to connect that new area to the main power grid.

In 2024, this bucket of business only contributes 0.5% of total revenue.

🔋Bucket #4: Other Side Hustles (Goods & Services)

This is a mix of smaller unregulated businesses which includes EV charging, rooftop solar installation, fiber-optic internet (Allo), consulting, and maintenance work that are related to electrical infrastructure.

🛡️Moat: Government Regulation

In my book, I listed out 4 common types of economic moats that Malaysian companies can have. Based on TNB’s business model described above, its competitive advantage is government regulation.

As mentioned in previous post, TNB owns 100% of the national power grid in Peninsular Malaysia. It is illegal for anyone else to build a competing power grid. This effectively gives TNB a natural monopoly position in transmission and distribution segments.

However, this does not mean TNB can raise prices freely. Tariffs are still governed by the IBR framework:

• When revenue falls short due to weaker demand, regulators can adjust tariffs to recover the allowed revenue;

• When revenue exceeds the allowed level, the excess is returned to consumers via rebates or lower tariffs.

The upside of this IBR framework is TNB will have very stable and predictable earnings. This is evidenced by their resilient earnings during pandemic period in 2020 despite lower power demand.

🌱Growth Catalyst

The main growth driver for TNB is the expansion of its RAB. As explained earlier, TNB’s true profit drivers are essentially:

(RAB + New CAPEX) x WACC

Under the latest IBR framework (Regulatory Period 4 or RP 4), TNB is permitted to spent RM42.82 bn over 3 years from 2025 to 2027 and earn a regulated return of around 7.3% on its invested capital.

However, not all RM42.82 bn can be spent fully. This approved CAPEX is split into:

• Base CAPEX (RM26.55 bn): For routine power grid upgrades and maintenance purposes.

• Contingent CAPEX (RM16.27 bn): This is driven by specific conditions such as unexpected surge in electricity demand from data centers that require expansion of power grid or battery energy storage systems (BESS).

CIMB analyst is expecting TNB to spend 60% to 70% of the contingent CAPEX in RP4. Based on this, it’s safe to say that TNB’s RAB will expand by at least RM36 bn.

Below is the company’s historical RAB from 2015 to 9M25 which is growing steadily at 4.3% CAGR:

In the past from RP 1 to RP 3, the approved CAPEX has always been hovering between RM18 bn to RM21 bn. The latest RP 4 has doubled this CAPEX amount and I believe it was mainly driven by 2 reasons:

💾Surge in Data Center (DC) investments

Since Singapore’s 2019 moratorium on new DC projects due to land constraints and massive strain on its power grid, Malaysia has experienced a surge in DC projects. The National Grid needs to be upgraded to support DC’s demand of high-voltage and ultra-stable connections.

In TNB’s 3Q25 Analyst Briefing, it highlighted that the electricity demand grew 5.2% yoy from DC. To-date, the company has secured 49 DC projects in the pipeline. This means more power consumption is coming.

🎯National Energy Transition Roadmap (NETR)

This framework, launched in 2023, outlines a roadmap to accelerate Malaysia’s shift from fossil fuel-dependent economy to a high value, low-carbon one. The ultimate goal of this NETR is to achieve Net-Zero greenhouse gas emissions by as early as 2050.

For TNB, it serves as a justification for government to increase TNB’s CAPEX to modernize the National Power Grid infrastructure. You cannot just plug a solar farm into an old grid. The system needs to be smarter to handle intermittent power.

According to CIMB analyst, 64% of the total contingent capex (RM16.3 bn) will be allocated to facilitate energy transition roadmap by TNB.

📈Likelihood of Higher Dividend Payment

As explained earlier, TNB has a target dividend payout of 30% to 60% of its adjusted Profit After Tax and Minority Interests (PATAMI).

In 2024, the company already pays DPS of RM0.51. Question is, can it still increase further for year 2025 & beyond?

I think it is very likely because of below reasons:

• Higher approved CAPEX which will drive TNB’s RAB higher as explained in growth catalyst section above

• Stable cash flows due to the implementation of automatic fuel adjustment (AFA) mechanism which automatically adjust fuel price fluctuations on monthly instead of every 6 months. Additionally, TNB’s reported PATAMI will be more accurate in reflecting its core regulated business.

⚖️Valuation & Risks

Read more

• Electricity moves through five main stages — from power generation to billing, with clear roles for procurement, transmission, distribution, and retail.

• TNB sits at the heart of Malaysia’s electricity system, owning the National Grid and acting as the key player across most of the power supply chain.

• While Malaysia still depends heavily on coal and gas, the government is pushing toward renewables, targeting 70% clean energy by 2050.

Disclaimer: Word of Caution!

Please DO NOT take this educational post as a buy or sell signal. When it comes to investing, it is important to have your own judgement. Despite my detailed analysis, mistakes may occur, and blindly following could lead you to make similar errors and financial losses. Furthermore, I AM NOT a licensed financial advisor. I’m merely sharing my experiences and opinions only.

Additionally, please note that I hold positions in the discussed stock, and my view may be biased as a result.

How is Electricity Generated?

Electricity is generated by converting energy sources into electrical power. Gas and coal plants use steam to spin turbines, hydropower relies on flowing water, and solar farms generate electricity through photovoltaic panels.

Once generated, electricity is stepped up to high voltage for efficient long-distance transmission through the power grid. It is then stepped down at substations and distributed safely to homes, offices, factories, and other end users.

Malaysia uses combination of energy sources:

• Non-renewable source: coal, natural gas, oil & diesel

• Renewable energy source: hydro, solar, wind & biomass

Below is the weightage of each energy sources used by TNB to generate electricity as of 3Q25:

We are still relying heavily on non-renewable energy sources (92.4%) for Peninsular Malaysia. In 2023, Malaysia government launched the National Energy Transition Roadmap (NETR) as a strategic plan to move towards the use of cleaner and more sustainable energy source. The plan highlights some key targets such as the gradual increase usage of renewable energy: 31% by 2025, 40% by 2035 and 70% by 2050.

Energy Landscape in Malaysia

Overview

The power sector in Malaysia is controlled by 3 companies:

• Tenaga National Berhad (TNB) – covers Peninsular Malaysia

• Sabah Electricity Sdn Bhd – covers Sabah and Labuan

• Syarikat SESCO Berhad (or better known as “Sarawak Energy”) – covers Sarawak

The first two entities are regulated by the Energy Commission (Suruhanjaya Tenaga) and Energy Commission of Sabah (ECoS), respectively. While Sarawak Energy operates independently under the Sarawak state government’s jurisdiction.

Notably, TNB holds an 83% stake in Sabah Electricity, making it the dominant player in Malaysia’s power sector with Sarawak being the only exception.

Both EC and ECoS are government regulators that oversees:

• How much utilities company can invest into building power infrastructure

• How electricity prices should be calculated

• Ensures utilities company earn a fair but controlled return

• Protect consumer from sudden price shocks

Since this research is about TNB, I will focus on where TNB has exposure which is mainly Peninsular Malaysia, Sabah and Labuan.

Supply Chain Structure in Peninsular Malaysia

The supply chain of electricity in Malaysia can be broken down into 5 primary segments:

• Generation

• Procurement & Grid Management

• Transmission

• Distribution

• Retail

Below are a detailed insights into each segment of the supply chain:

#1: Generation

This segment is where power is produced through various power plants. In Malaysia, most of the electricity are produced by TNB Power Generation (TNB Genco) and various Independent Power Producers (IPPs) such as Malakoff, YTL Power, etc. It is a competitive segment because TNB has to compete with other IPPs.

They are compensated via the Power Purchase Agreements (PPAs) for the IPPs or Service Level Agreements (SLAs) for TNB Genco. These are set by the Single Buyer and it has two components:

• Capacity – covers the fixed costs such as depreciation & maintenance to ensure it is always available to produce power regardless of demand fluctuations.

• Energy – covers the costs of energy sources used such as coal, natural gas, etc.

#2: Procurement & Grid Management

“Single Buyer (SB)” is an operating model where the procurement of power generation from TNB Genco and IPPs are done centrally. Basically, SB manages the electricity demand forecasting, schedules power dispatch from generators based on a merit order (lowest cost first) and administers the PPAs or SLAs with various power producers.

While SB is an entity placed within TNB, it is “ring-fenced”. This means it operates with separate accounts, governance, and decision-making processes to ensure procurement decisions are fair, transparent, and non-discriminatory. So, it is not treated as “just another business unit” that prioritizes TNB’s own generation.

Once power producers are contracted under approved PPAs or SLAs, their generating units (power producing units within a power plant) are connected to the National Grid.

The National Grid (comprised of towers, cables, substations, etc.) is operated by the Grid System Operator (GSO), another ring-fenced entity within TNB. Its primary duties include operational planning, generating unit dispatch, and coordinating all parties connected to the grid in compliance with the Malaysian Grid Code.

In a nutshell, GSO is the “traffic controller” of the electricity system. It has authority to instruct power plants to maintain, increase, or reduce output based on system conditions to prevent outages.

#3: Transmission

Electricity that is generated will be stepped up and transported over long distances via the National Grid. While this grid is operated by GSO, it is physically owned by TNB. They are responsible for managing the full lifecycle of the grid, including new grid construction, upgrades, and maintenance. In return, TNB is compensated for the costs incurred based on the Incentive-Based Regulation (IBR) Framework.

#4: Distribution

This is the last mile delivery of electricity from transmission substations to end-consumers by lowering down the voltage. This distribution grid is also physically owned by TNB. Similar to “Transmission” segment, TNB is compensated for the costs incurred in maintaining & building these grid.

#5: Retail

Retail covers billing and collection as well as “beyond kWh” solutions such as rooftop solar and EV charging.

Revenue is collected as a bundled tariff and then distributed to the other segments (SB, GSO, Transmission and Distribution) based on their approved compensation as per the IBR Framework.

The Rulebook: Incentive-Based Regulations

Notably, TNB is the only company that owns the National Grid and also involves in the power generation segment in Peninsular Malaysia. Other power companies such as Malakoff, Edra Power or YTL Power participate only in the “Generation” segment. This makes TNB a natural monopoly in Malaysia’s power sector.

As such, a rulebook is required to govern this sector to ensure efficiency and affordability of electricity. This rule book is known as Incentive-Based Regulations (IBR) Framework.

Based on this IBR framework, TNB’s involvement in below segments of the electricity supply chain are classified as Regulated Business Entities (RBEs):

• Procurement & Grid Management (Technically, not part of TNB)

• Transmission

• Distribution

• Retail

This means TNB’s revenues, capital investments, and operating costs incurred within the above 4 segments are tightly regulated by this framework.

Think of this IBR framework as a government contract with TNB that dictates how much electricity tariff should be charged and what profit rates are allowed for TNB. Also, this contract is adjusted every 3 years – known as “Regulatory Period (RP)”. Malaysia is currently in RP4, which runs from 2025 to 2027.

At its core, TNB is not paid based on how much electricity consumer uses. It is paid based on how much it costs to run and invest in the power grid efficiently. The EC calculates the “Annual Revenue Requirement (ARR)”, which essentially reflects how much money TNB needs to run and maintain the power grid, plus a fair profit. Electricity tariff is then structured to recover this amount.

EC uses this “Building Block” formula to figure out the ARR, which comprised of:

The problem comes when there’s fluctuations in below two areas:

• Prices of energy sources such as coal, natural gas and fuel

• Electricity consumption

Because these 2 fluctuations will affect TNB’s profitability. Higher energy costs will lead to TNB paying more for the power generation than amount collected from consumers based on the base tariff rate and vice-versa.

Similarly, lower electricity demand would mean lower revenue collection from consumers and vice-versa. This will cause shortfall or surplus in the ARR.

The IBR framework resolves this by enforcing 2 mechanisms to resolve the shortfall or surpluses:

#1: Automatic fuel adjustment (AFA)

AFA was introduced in July 2025 as part of the RP4. Under this new mechanism, fuel costs are tracked and adjusted automatically on a monthly basis. If coal gets expensive, a small surcharge is added on top of the base tariff; if it drops, consumer gets a rebate.

This is different from the past (RP3) where the fuel costs are only adjusted every six months which does not reflect the actual fuel costs efficiently as compared to the AFA.

This new mechanism will improve TNB’s cash flow stability as the company now avoids funding large fuel cost increases out of its own balance sheet.

#2: Revenue-Cap

This mechanism aims to ensure TNB only earns a fixed amount of revenue each year as calculated and approved by EC in advance. If actual electricity sales are higher or lower than expected, tariffs are adjusted in future periods to ensure TNB ultimately receives exactly the approved revenue – no more, no less.

This revenue-cap will make TNB’s core earnings stable, predictable and not volume driven.

Supply Chain Structure in Sabah & Labuan

The supply chain for Sabah and Labuan is managed by Sabah Electricity Sdn. Bhd. (SESB), which is an 83%-owned subsidiary of TNB. This is similar to Peninsular Malaysia where the procurement of electricity is done centrally – Single-Buyer Model as explained above. Both the SB & GSO is a ring-fenced unit within SESB.

Below are the key differences of supply chain in Peninsular Malaysia, Sabah & Labuan:

That said, Sabah state government is currently working on a financial plan to buy back SESB completely by 2030 from TNB. This takeover is intended to give Sabah total control over its supply chain, from generation to billing, allowing it to align the utility’s operations more closely with local industrial needs. As such, not worth looking into it further.

Now that I’ve covered the energy landscape in Malaysia, my next post will take a deeper dive into Tenaga’s business model and the key risks to take note. Stay tuned!

• KLSE to enter 2026 on stable footing, supported by OPR cut, stronger MYR, and improving domestic demand.

• Consumer spending and corporate earnings should stay resilient, helped by civil service pay hikes and strong investment pipelines.

• Data centre–related sectors remain the key growth engine, benefiting utilities, telcos, renewables, and construction.

This is my second part of my annual market outlook which focuses on Malaysia Market in 2026. If you have not read the first part on my outlook for global markets, click here to read it.

Malaysia in 2025 has been on recovery mode since Apr-25. Here’s my outlook:

MY Market: Remain Stable

Malaysia stock market had a -2.1% loss in 2025. Not surprising because most of the listed companies in Bursa Malaysia are old economy (e.g. mining, agriculture, property development, manufacturing, etc.).

Sectors that are booming in Malaysia is really data center and its related value chain (e.g. renewable energy, utilities, telecommunications, constructions, etc.).

There are several key events in 2025 that sets the stage for 2026:

• OPR cut to 2.75%. Since Jan 2025, inflation has consistently slowed to 1.4% in Nov-25. Unemployment rate has been stable at 3%.

• MYR strengthened against USD by approx. 10%.

• Pay hike for civil service (effective January 2026 onwards)

• Budget 2026 tabled in Oct-25

Based on the above key events, I anticipate the overall Malaysia stock market to be slightly on bullish side. This is because of below views that are more bullish than bearish impact:

💸 Consumer Spending & Corporate Earnings to Increase

Both the OPR cut and pay hike for civil servants will definitely boost consumer spending in 2026.

Corporate earnings will also be resilient in 2026 as it is supported by:

• Pipelines of investments (e.g. RM285.2 billion as of 3Q25) from both foreign countries and domestic investors which created over 150,000 jobs. This is a 13.2% increase from previous year.

• Approximately 62.8% of export goods to US are exempted from the reciprocal tariff rate of 19% set in Oct-25. This will increase the US demand for Malaysia’s goods as prices becomes competitive against countries with reciprocal tariff.

• The stronger MYR against USD will lower input costs for companies that source their supplies in USD

💱 MYR strengths to continue

That said, earnings growth may be more visible among domestically focused companies, while net exporters could come under pressure from higher forex losses and tighter margins in 2026.

According to MARC Ratings, they are foreseeing that USD will continue to strengthen till mid-2026. The MYR’s performance in 2025 was not limited to the USD alone, it also strengthened against several other major currencies.

Below snapshot shows how MYR fares against other currencies as of Nov-25:

Malaysia’s top 10 export destinations for 2024-2025 are Singapore, US, China, Hong Kong, Japan, Taiwan, Thailand, Indonesia, South Korea, and Vietnam. Interestingly, the MYR strengthened against the currencies of most of these trading partners.

Looking deeper, semiconductor products remain among Malaysia’s largest export categories. As a result, I expect the E&E sector’s financial performance to be affected the most in 2026.

💾 Data Center Related Sectors to Continue Booming

I anticipate that sectors within the Data Center (DC) value chain will continue to experience earnings growth due to:

• Budget 2026 tabled in Oct-25 places heavy emphasis on digital infrastructure, AI, cloud, and data ecosystem development which indirectly supports DC growth in Malaysia.

• As an indicator of DC growth, Tenaga has secured a total of 49 DC projects as of 3Q 2025. This is already higher than 2024 by 28.9%.

Below are sectors that I believe will continue to do well in 2026:

📈 Inflation could increase

I believe there’s always a possibility that inflation could rise in Malaysia. In Jul-25, Tenaga adjusted its tariff structure which will result in higher electricity costs for commercial & industrial businesses. As they are the typical high energy usage, especially data centers.

I’m pretty sure business owners will pass on these extra costs to their customers.

My Plan for 2026

Just like I said in my previous outlook post, my goal is to position my portfolio to capitalize on long-term secular trends. This yearly market outlook doesn’t prompt me to alter my portfolio positions; rather, it only serves as assurance whether my long-term view has change or not.

Unless there is a big structural trend change such as “AI is just a hype”, I will continue to focus my portfolio in this sector heavily while maintain some position in other sectors for diversification purposes.

For my paid subscribers, here is what I’m planning to do in 2026 for each market:

Read more

• US Market: Expect to end the year positively but there will be high volatility during the year due to new Fed chair and mid-term elections.

• HK Market: To continue its uptrend despite the market is on the expensive side as the government vows more accommodative policy.

• SG REIT Market: To continue its uptrend recovery as the lagging effect from lower interest rate starts to show in their 2026 financial performance

Now that 2025 is done and dusted, it is time to do a yearly market outlook on markets that I’m currently invested in – US, SG, MY and HK. This annual outlook isn’t about precision forecasting but building a framework to organise my thoughts around the key events from 2025 that could impact my portfolio in 2026 and beyond.

Like the famous quote from Benjamin Franklin, “If you fail to plan, you’re planning to fail”. This annual outlook will help me plan my actions forward. I’m splitting this annual outlook into 2 parts. First part focuses on global markets. The second part will be solely on Malaysia market as combining it will make this blogpost too long.

Here’s my outlook for 2026:

US Market: Volatility Ahead

2025 has been a bullish year despite the market getting hit with Trump’s sweeping tariffs in Apr-25. The S&P 500 has gained 18.1% supported by their thriving economy. GDP was initially shrank by -0.6% in 1Q 25 before turning around and gaining steam, growing at +4.3% as of 3Q 25.

Inflation was under-controlled. The core CPI (excl. food & energy prices) rises to only 2.6% as of Nov-25, which is less than anticipated 3% increase. The US Fed carried out below actions in 2025:

• Cut interest rates 3 times, bringing down the interest rate to 3.5% – 3.75%.

• Ending the quantitative tightening (QT) from 1^st Dec 2025 onwards

The Dot plot now shows 2 more cuts – 1 in 2026 and 1 in 2027. However, I anticipate that the market in 2026 will be volatile. This is because:

📉 Technical Pullback is Due

Based solely on technical chart, SPX is hovering just slightly above its previous high of 6,917 points (potential resistance). Since the market don’t move in straight line, I think there will be a pullback – how much? I was hoping for -5.4%, given how expensive the market is right now.

At the time of writing, the SPX Index is trading at PE of 31.28x, which is above the 10-yr median PE of 23.81x. The market is definitely not cheap right now.

🏦 New US Fed Chair in May 2026

The current US Fed Chair, Jerome Powell is expected to leave in May 2026. Trump is expected to put someone that agrees with his views of having rate cut to 1%. The general market does not like uncertainties and so, I anticipate there will be volatility during this time until we know who the new US Fed Chair is – “Sell in May and go away?”

🗳️ US Mid-term Elections in Nov 2026

According to Carson Investment Research, the years with mid-term elections are generally weakest in terms of the market returns but it also depends on whether it is the first term or second term president. For Trump’s case, it is the latter. Hence, not too bad.

That said, anything can happen. I somehow have the feeling that 2026 will be like 2022 where there’s a big correction for the whole year before recovering in 2023. The chart below from Carson Investment Research shows that the correction for mid-term years tends to last until later part of the year.

Nevertheless, I think the US market will still end the year 2026 positively because:

#1: Lower US Fed Fund Rate & Higher Liquidity in Market

The US federal funds rate is expected to see one more cut in 2026, and the QT program has already ended. This is positive for markets, as liquidity conditions are becoming more supportive of growth because the money supply increases.

This typically benefits risky assets by lowering financing costs, improving access to capital, and encouraging investment flows into equities and longer-duration assets.

#2: Extra Booster from Trump’s One Big, Beautiful Bill

This bill introduced by Trump could boost the US economy through tax cuts on:

• Individuals which could result in $100 billion in tax refunds. This will increase the US consumer spending.

• Corporation which could result in $137.2 billion reduction in tax liability.

HK/ China Market: Uptrend Continues 🚀

The Hong Kong market had a strong run in 2025, climbing roughly 27%. On a 5-year PE basis, valuations are starting to look a little expensive, sitting above the average of 11.02x. Still, market don’t reprice overnight. With HK and China still recovering from years of crackdowns and property-related crisis, higher valuations can persist longer than expected.

At the same time, earnings recovery and continued government policy support could help justify higher valuations, as investor confidence gradually rebuilds.

There are few bullish points that I believe HK market will continue its uptrend in 2026:

• China’s PBOC cut its key lending reference rates for its one-year and five-year’s Loan Prime Rates (LPR) once by 0.10% in May 2025. This will ease the burden of borrowers.

• In Dec-25, adjustments were made to lower the risk factors for insurers’ stock investments. This will encourage greater stock market participation from insurers.

Additionally, China’s major tech companies such as Tencent, Alibaba, and ByteDance could benefit from renewed momentum in AI development if Trump administration allows Nvidia to sell H200 chips to these Chinese firms. We may see more innovations in the AI space.

It’s still uncertain whether this would invite new competitors to copy Nvidia’s GPU designs, mainly because China does not yet have access to ASML’s EUV machines needed to manufacture advanced chips. Until that hurdle is cleared, China remains behind in the semiconductor race.

If China eventually cracks this challenge, greater GPU supply could normalize today’s tight demand environment. This is similar to how China’s EV manufacturers have increased competitive pressure on US and European automakers. I digress.

SG REIT Market: Benefit from Lower Interest Rate

This section is intentionally narrow to just REIT market in Singapore (SG). I’m not making a call on the entire Singapore market because my exposure is limited to SG REITs, and that’s where my analysis sits.

SG REITs tend to move with interest rates, since borrowing costs play a big role in their growth and distributions. 2025 turned out to be a recovery year, helped by rate cuts across the US, Europe, Hong Kong, and China. Over time, this should ease financing costs for REITs with exposure to these markets.

Still, this low interest rate relief doesn’t come instantly. For REITs with fixed-rate debt, lower interest costs only show up when loans mature and are refinanced.

Below are some SG REITs’ debt profile I took from CIMB research:

I highlighted my current position in SG REIT – Mapletree Industrial Tr (MINT) and Frasers Centrepoint Tr (FCT). MINT is going to refinance 34% of their total debt between 2025 & 2026. While FCT is refinancing 16% of their total debt in 2026.

I’m not worried about FCT because the retail sector in SG still resilient. In October 2025, retail sales, excluding motor vehicles, grew 3.7% picking up from September which is 1.8%.

The one I’m watching closely is MINT’s performance because of reasons I have listed in my previous post on “Portfolio Review” – you can read it here. The only positive thing I see is the continued rate cuts in 2026. It could offset some of their loss of income from asset divestments.

In summary, I think SG REIT will continue its recovery supported by these rate cuts globally. All eyes on their Budget 2026 – this is one uncertainty ahead.

Closing Thoughts

For 2026, I anticipate all 3 markets – US, SG and HK/ China – to be bullish for reasons as above. That said, there will be some volatility coming from US due to the unpredictability of Trump’s administration.

Stay tuned for my outlook on Malaysia market and my plan for 2026.

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Summary

• Global macro surprised to the upside, with softer inflation and multiple rate cuts instead of the tightening I expected.

• Geopolitics played out as anticipated, but market impact was milder despite renewed US-China trade tensions.

• Malaysia defied my expectations, with an OPR cut and a stronger MYR supporting market sentiment.

• Sector calls were mixed, as data centres and renewables outperformed while exporters recovered for unexpected reasons.

At the end of 2024, I wrote Navigating 2025 Markets as a way to organize my thoughts across the US, Singapore, Hong Kong, and Malaysia before stepping into another unpredictable year. I didn’t expect to get everything right. I just wanted a framework that made sense at that moment.

Now that 2025 has finally closed, I went back to compare what I believed should happen against what actually unfolded. And honestly, reviewing these predictions felt like reading old journal entries. Some parts I nodded at, others I laughed at.

Here’s a look back at how my 2025 calls aged:

#1:🌎Global Markets

Heading into 2025, I had a clear narrative in my mind. Trump had just won the US election. Tariffs were back at the center of policy conversations. Inflation looked like it might flare again. And with that, I thought below would happen:

• The Federal Reserve would slow down rate cuts or even hike if needed

• Stronger USD as markets priced in higher-for-longer interest rates.

But the market had other plans.

📉Inflation didn’t return… and the Fed cut rates three times

Instead of tightening, the Fed delivered three rate cuts of 25 bps each, bringing the federal funds range down to 3.5% – 3.75% in:

• September 2025

• October 2025, and

• December 2025

That was the first major curveball that threw my prediction off. The softer inflation trend persisted, the US economy held steady driven by AI narratives, and markets embraced the idea of a soft landing. As a result, the USD weakened, not strengthened – another miss on my part. The USD Index has dropped 10% year-to-date:

Central banks around the world also cut their interest rates:

• European central bank cut its interest rates 4 times. Its main refinancing operation rates was cut from 3.15% to 2.15% as of June 2025.

• Monetary Authority of Singapore had eased the pace of SGD Nominal Effective Exchange Rate (S$ NEER) appreciation twice earlier in 2025.

• China’s PBOC cut its key lending reference rates for its one-year and five-year’s Loan Prime Rates (LPR) once by 0.10% in May 2025.

• Hong Kong Monetary Authority cut its base interest rate 3 times in 2025 from 4.75% to 4% as of December 2025

• Malaysia’s Bank Negara cut its overnight policy rate (OPR) once in July 2025 from 3% to 2.75%.

In a nutshell, inflation was contained despite Trump’s sweeping tariff threats.

⚔️But on geopolitics… I was spot on

If there was one global theme I expected to intensify, it was the US-China trade war. And this prediction turned out to be accurate.

China retaliated against Trump’s renewed tariff threats with measures that were sharper than I envisioned:

• Export controls on rare earth minerals

• Adding US firms to its “Unreliable Entities” list and impose export ban

• Blocking certain soybean imports through suspended permits

But global markets got a bit of relief later in the year when both sides agreed to a temporary truce.

In hindsight, the global story of 2025 became a mix of humility and confirmation:

• Macro? I was too early and too defensive.

• Geopolitics? I called it right, but the market impact wasn’t as severe as expected.

#2: Malaysia Market

I expected Bank Negara Malaysia (BNM) to keep OPR at 3.0% throughout 2025, given how conservative they were in past cycles. But in July 2025, BNM surprised the market with a 25 bps cut, lowering OPR to 2.75%.

This aligned Malaysia more closely with global easing trends and provided a sentiment lift to interest-sensitive sectors. Another miss on my end.

💵MYR: I expected it to weaken… it strengthened

This was my third major surprise of the year. Instead of weakening due to tariff risks and inflationary spillovers, MYR strengthened roughly 9.4% against USD year-to-date.

This stability helped import-heavy businesses and eased the pressure on companies that previously suffered large forex losses. It was the opposite of what I anticipated but a welcome development for the broader market.

#3: Sector Specific Reflections

Some parts of my outlook aged well. Others didn’t. Here’s a narrative version of the year’s sector story:

💾The Data Centre Boom: This one hit the mark

I expected the data centre value chain (e.g. construction, utilities, telco & renewable energy) to stand out in 2025. And this turned out to be correct.

A few notable highlights:

• Tenaga saw electricity demand grow 7.7%, with 5.2 percentage points coming from data centres as of 3Q25.

• TIME dotCom delivered strong 3Q25 results, backed by higher data revenue and robust demand from AIMS (its 30% associate).

• Renewable energy players recorded some of their best earnings in recent years. Examples are Solarvest Holdings Berhad’s Q2 2026 net profit has already surpassed its FY2024 net profit.

The momentum from 2024 clearly carried into 2025 driven by increasing adoption of AI and EV.

📦Export-oriented companies: The reversal I didn’t anticipate

In my original outlook, I expected exporters to recover due to a stronger USD because many of whom suffered heavy forex losses in 2024.

But since USD weakened, the recovery came instead from:

• Stabilizing MYR

• Improving external demand

• Better cost management

The direction was correct, they did recover but the reason was not what I predicted.

Smaller Themes: Compressed but worth noting

• Inflation drivers (RON95 subsidy removal, minimum wage, civil service pay hikes, sugary beverage tax) played out, but their impact was more muted than expected.

• Dividend-oriented sectors such as retail and industrial REITs remained resilient.

These didn’t materially change the market narrative but shaped the overall texture of 2025.

📌Closing Thoughts

Most of my big predictions were wrong. A few played out as expected. The year 2025 didn’t go the way I thought it would, but it gave me clearer insight into how unpredictable and fascinating markets can be.

I’ll be doing another prediction for 2026. Stay tuned and happy new year!

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

• Revenue was flat YoY at $1.08B, landing at the top end of guidance, with stable performance across consumables and instruments.

• Non-GAAP profitability exceeded expectations, with operating margin at 24.5% and EPS at $1.34 — supported by cost controls, favorable NovaSeq X mix, and share buybacks.

• NovaSeq X adoption remains the key growth driver, with the migration running ahead of schedule and representing over half of revenue this quarter.

• Management has raised its full-year guidance after this solid quarter result.

Illumina has released its 3Q25 earnings. Doing an update on their latest performance. If you want to know in-depth about their business, click here.

📊Financial Highlights

Overall, Illumina delivered a clean beat quarter. Revenue hit the top end of guidance. Both operating margin and EPS are above expectations.

Despite revenue being mostly flat, the company is becoming more profitable, thanks to ongoing cost controls and a more favorable mix from the NovaSeq X platform.

The company repurchases 1.24 million shares for $120 mil in 3Q25, contributing to non-GAAP diluted EPS growing 17.5% yoy despite their quarterly revenue being flat.

In my previous post, I highlighted 3 main challenges Illumina faces:

#1: US Funding Cuts💵

Illumina’s target customers can be broken down into 2 segments:

• Research & Applied

• Clinical

The Clinical segment is more resilient, as it is directly integrated into healthcare applications such as oncology testing, genetic disease testing & reproductive health diagnostic. In contrast, the Research & Applied segment is cyclical because its customers are research institutes. Their spending on Illumina’s products depends on government funding.

As of 3Q25, the clinical segment continues to be the one doing the heavy lifting. Revenue in this segment is growing at double-digit (excluding China), +12% growth. This is driven by broader adoption of comprehensive genomic profiling and growing use of sequencing-intensive applications like MRD.

The sequencing consumables revenue for research & applied segment continues to decline in high-single digit (excluding China), -8% declined, due to funding uncertainty. One positive note from the management is the demand seems to be stabilizing. Nonetheless, they still expect this funding uncertainty to continue in 2026.

#2: Tariffs👮

While the US sweeping tariffs do not have significant impact on Illumina’s performance, the Management mentioned for 3Q25 it had a 2.2% impact on gross margin.

#3: China’s Regulatory Restrictions🚫

Revenue contribution from China continues to decline and is now only 5% of Illumina’s total revenue for 3Q25. However, management mentioned that:

“Illumina has received approval to manufacture certain instruments locally for OEM partners in China.”

This is not a full solution, but it’s a small step forward while instrument exports to China remain restricted. Overall, I still view China as a “bonus upside” at this point rather than a core driver. Thus, my valuation will continue to exclude this market until there’s some level of certainty.

Growth #1: NovaSeq X Transitioning Well📋

This continues to be Illumina’s biggest growth engine. As of 3Q25:

• More than 55 NovaSeq X instruments were placed in the quarter. This is in-line with their goal of 50 to 60 placements per quarter.

• Roughly 78% of volumes and 51% of revenue in Q3 was sequenced on NovaSeq X.

• 91% of research volumes & 64% of clinical volumes have already transitioned to NovaSeq X. This means there is still room for growth in instrument sales within the clinical segment.

🎯Future Guidance Raised

Overall, this quarter’s result is very solid with expansion of operating margin. The management has again raised their full year’s guidance:

⚖️Valuation

As mentioned in my previous post, this remains a waiting game. But judging from this 3Q25 result, the only major problem that hinders the company’s growth is its research segment. With the demand normalizing, I believe Illumina will do better from 2026 onwards.

I am not adjusting my valuation of Illumina. This is because my previous assumptions are already conservative (at the worst case scenario). The intrinsic value remains in the range of $156.98 to $180.75.

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Additionally, please note that the author holds a position in the discussed stock, and his view may be biased as a result.

• Revenue declined by 2.7% YoY, with flat performance in consumables and instruments

• However, non-GAAP diluted EPS grew by 9% YoY despite revenue declined

• The company faces on-going challenges from US funding cuts and tariffs but supported by resilience in clinical market segment.

• Management has raised its full-year guidance after a good quarter result.

Illumina has released its 2Q25 earnings. Doing an update on their latest performance. If you want to know in-depth about their business, click here.

📊Financial Highlights

Not surprising, most of its revenue segments are either down or flat. On the bright side, its non-GAAP operating margin and diluted EPS are above expectations. The management’s action in controlling their expenses as mentioned in their previous quarter’s earnings release is working.

Non-GAAP operating expenses are down 6% yoy. The company repurchases 4.5 million shares for $380 mil in 2Q25. This has resulted in non-GAAP diluted EPS growing 9% yoy despite their quarterly revenue down by 2.7% yoy. The remaining amount Illumina is authorized to repurchase is approx. $800 mil.

In my previous post, I highlighted 3 main challenges Illumina faces:

#1: US Funding Cuts💵

Illumina’s target customers can be broken down into 2 segments:

• Research & Applied

• Clinical

The Clinical segment is more resilient, as it is directly integrated into healthcare applications such as oncology testing, genetic disease testing & reproductive health diagnostic. As of 2Q25, this segment accounts for roughly 60% of sequencing consumables.

In contrast, the Research & Applied segment is cyclical because its customers are research institutes. Their spending on Illumina’s products depends on government funding.

As Trump administration cuts the National Institutes of Health (NIH)’s budget, Illumina’s research customers continue to face funding constraints. Thus, lower sequencing instruments & consumables demand.

#2: Tariffs👮

Apparently, the US sweeping tariffs do not have significant impact on Illumina’s performance. The Management mentioned for 2Q25 it only had a 1.1% impact on gross margin. I think this provides some comfort for investors.

#3: China’s Regulatory Restrictions🚫

China now only contributes 6% of Illumina’s total revenue for 2Q25. It has become the smallest segment. I think it is not worth keeping track anymore. The only positive point is Illumina can still sell sequencing consumables to its China customers. The export of instruments into China is still restricted. I will still treat this market as gone completely.

Growth #1: NovaSeq X Transitioning Well📋

Both the funding cuts and tariffs are beyond Illumina’s control. Their main growth drivers are coming from clinical segment. As mentioned, 60% of sequencing consumables are from this clinical segment. Management is anticipating more to come relative to the research segment.

One factor I see is the company’s NovaSeq X transition progress. As of 2Q25:

• Roughly 55% of sequencing volume for clinical segment already transitioned to this NovaSeq X series

• For research segment, over 80% of sequencing volume has already transitioned

This means more instrument sales will be coming from their clinical segment moving forward. In 2Q25, approx. 44% of high throughput consumables revenue was on NovaSeq X series and the management anticipate achieving 50% by end of 2025.

Growth #2: MiSeq i100 Plus Sales📋

This is Illumina’s latest low throughput device. In 2Q25, the company has placed about 500 instruments. Management are seeing customers order additional units after just few months of use. Feedback remains positive.

Growth #3: Acquisition of SomaLogic📋

This company is a key player in high throughput proteomics. According to NIH, Proteomics is the study of the interactions, function, composition, and structures of proteins and their cellular activities. Proteomics provides a better understanding of the structure and function of the organism than genomics.

Somalogic’s Somascan able to analyze over 9,500 unique human proteins from small biological samples & deliver insights for drug discovery, diagnostics and health monitoring. How this contributes to Illumina’s growth remains to be seen.

🎯Future Guidance Raised

Overall, this quarter’s result is consider good. The management has raised their full year’s guidance:

⚖️Valuation

As mentioned in my previous post, this remains a waiting game. I still believe that the challenges Illumina is currently facing are temporary. How long until resolution? My best estimate would be until Trump is replaced with a new president – in 2028? Hahaha…

Despite the above uncertainties, I am not adjusting my valuation of Illumina. This is because my previous assumptions are already conservative (at the worst case scenario). The intrinsic value remains in the range of $156.98 to $180.75.

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Additionally, please note that the author holds a position in the discussed stock, and his view may be biased as a result.

• Three main categories of AI use cases: Generative, Predictive, and Physical AI, each with distinct use cases and revenue models

• Generative and Predictive AI, relying on subscription-based models, face customer retention challenges amid intense competition

• Business risks of physical AI are highly industry-specific

This is the fourth and final instalment of my in-depth overview of the AI value chain, focusing on Applications. It explores how Machine Learning (ML) and Deep Learning (DL) are integrated into real-world use cases, driving value across industries and enhancing our daily lives.

At its core, AI applications empower businesses to streamline operations and improve decision-making, resulting in substantial value additions to their bottom lines. These applications can be standalone AI software or embedded as add-on features within existing software systems.

AI applications can be broadly categorized into three key areas:

⚛️Generative AI

Generative AI is what most people are actively using today – examples include ChatGPT, Gemini, and DeepSeek. These applications enable the creation of new content and typically operate on a subscription-based revenue model.

For instance, Adobe has introduced its own generative AI solution, Firefly, which allows users to produce images, audio, and video content.

Statista projects that the generative AI market will reach nearly $66.9 billion by 2025, growing at CAGR of 37% from 2025 to 2031. This growth is fueled by increasing enterprise adoption, particularly in the realm of digital assistants and AI agent.

📊Predictive AI

This refers to AI technologies that are embedded into existing software applications. It’s like an additional feature. It is able to recognize patterns in large dataset and predict the outcome. Yes, it is the ML and DL algorithm – foundation elements of all AI technologies.

For example:

• 🎬Netflix uses the ML algorithm to recommend shows that their users would like to watch. They don’t charge this AI feature. Instead, it’s already part of their content subscription fees.

• 🛡️CrowdStrike integrates ML into its cybersecurity platform, Falcon, to analyze massive datasets, detect patterns, and adapt to emerging threats. Predictive AI in this context enhances the functionality of existing systems rather than being a standalone application.

🤖Physical AI

Physical AI involves the integration of artificial intelligence with robotic and automation systems that interact directly with the physical world. Applications span multiple industries:

• 🚗Automotive: Autonomous driving technologies, like Tesla’s Autopilot and Waymo’s driverless taxis, leverage DL algorithms embedded in safety and navigation systems

• 🏥Healthcare: Companies such as Intuitive Surgical develop robotic surgical systems that use AI for precision and safety in minimally invasive procedures

• 🏭General Manufacturing: AI-driven machine vision systems, like those from Cognex Corporation, play critical roles in automated quality control, inspection, and optimization of production processes.

⚠️Key Business Risks

Both the generative AI and predictive AI are mostly operating a “software-as-a-service” (SaaS) model. From an investor’s perspective, the primary risk associated with this model is customer retention, given the intense competition among similar applications.

What makes a customer want to continue the subscription & perhaps buy more add-on features? There’s an article from Sequoia Capital – a venture capital company, it highlights two key factors that create stickiness and loyalty which is:

• Workflows – how AI tools integrate into users’ work processes

• User networks – the community and ecosystem built around the AI tools

As an investor, these two will be the key metrics to monitor & determine whether these AI companies have competitive advantage.

As for Physical AI, in my opinion, its business risks are very industry specific. For example, within the manufacturing industry, automation is important to enhance productivity & efficiency. However, competition can be very intense as there are many companies that provide such automation solutions to these manufacturers.

Additionally, the income stream is often non-recurring in nature since these automation solutions require bidding (project based) while the maintenance can be performed by others or at a discounted pricing.

🔎Conclusion

As a long-term investor, I prefer companies with high income predictability because:

• It will not give me surprises such as sudden drop in revenue

• Forecasting the company’s earnings growth would be much easier & reliable

Based on this, I think most of the investment opportunities will be in the generative and predictive AI categories.

This concludes my comprehensive analysis of the AI value chain and marks the beginning of my new investment strategy. Moving forward, I will be posting my in-depth research into AI-related stocks. If you’re interested in following my journey, be sure to subscribe for more insights!

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Additionally, please note that the author holds a position in the discussed stocks, and his view may be biased as a result.

• Robust data infrastructure is vital for AI, enabling massive data collection, storage, and processing.

• Data centers and cloud services provide scalable, on-demand computing resources for AI.

• Data platforms and analytical tools streamline efficient model training and deployment.

• Cybersecurity is crucial for protecting AI data in an increasingly threat-prone digital landscape.

Welcome to the second part of my in-depth exploration of the AI value chain which is about data infrastructure.

Data is the lifeblood of AI applications, and robust data infrastructure is essential for collecting, storing, processing, and managing the vast amounts of data required for ML and DL.

This data infrastructure encompasses:

• Data Centers & Cloud Service Providers

• Data Platforms and Tools

• Foundational Models

• Cybersecurity

💾Data Centers & Cloud Service Providers

As data required for training increases, companies would need a place to store and process these data. The most cost-effective way is through cloud-computing. It provides scalable, on-demand access to computing resources via internet.

It eliminates the need for significant upfront capital expenditure on hardware such as servers. Companies that require more computing resources can just subscribe with cloud service providers (CSPs).

According to Synergy Research Group, below are the top cloud providers worldwide:

CSPs are the ones constantly looking to purchase more computing resources to cater for their increasing customers’ demand.

Goldman Sachs is forecasting that revenue from cloud computing will hit $2 trillion by 2030. The total addressable market of cloud computing is expected to grow at CAGR of 22% from 2024 to 2030. Approximately 10% to 15% of the revenue is from generative AI.

With the expected increase in cloud computing demand, it also drives the demand growth for data centers. These are facilities that house critical infrastructure required to store, process and manage vast amounts of data generated worldwide.

It has the following main components:

Beyond the tech sector, this AI-driven transformation is creating a significant spillover effect across industries such as power, real estate development, and property management, all benefiting from increased demand for data centers.

Investment Outlook

Mckinsey is projecting the global demand for data center to increase at an annual rate of between 19% to 22% from 2023 to 2029. While the demand is increasing, there is one uncertainty remains.

As mentioned in my previous post, the Trump administration is replacing Biden-era’s U.S. AI Diffusion Policy with a new one. It remains to be seen what this new policy entails and its potential impact on overall market growth.

Any restriction on chip exports could slow down the future development of data centers. At the moment, there are no issues with existing orders since this AI Diffusion Policy has been rescinded.

Regardless, I think the impact would be minimal for those supporting components such as electricity and network connectivity as these are largely subscription-based tied on existing data centers.

Overall, companies within this data center and cloud services segment benefit from highly recurring revenue models, making them well-suited for long-term investment. However, one should remain cautious of ongoing geopolitical uncertainties that may impact future growth.

📊Data Platforms and Tools

These are used to manage and facilitate the effective use of data. Companies such as Snowflake, Databricks, Palantir, etc. provide platforms & tools to facilitate the processing and analysis of data which serves as the foundation for training and deploying AI models.

These platforms and tools has the following benefits other than training AI models:

• Help companies consolidate various systems - from point-of-sale (POS) solutions to backend ERP systems - into a single, integrated environment for streamlined data analytics.

• Businesses can leverage data to drive strategic decisions. Take TikTok, for example - it harnesses data to recommend videos and advertisements, keeping users engaged on its platform for hours.

Investment Outlook

In my opinion, this segment will have companies with highly recurring and "sticky" business models. Their products often create high switching costs, making it difficult for users to change providers.

Migrating large volumes of data from one platform to another is both time-consuming and costly. Additionally, adopting a new platform comes with a learning curve, further discouraging users from making a change.

The global data platforms and tools market is projected to grow at a compound annual growth rate (CAGR) of approximately 12.5% from 2024 to 2032.

🧠Foundational Models

These are pre-trained AI models which form the basis for a wide range of AI applications across diverse industries. Rather than developing AI from scratch, one could use this foundational model as starting point to develop AI that power new applications more quickly and cost-effectively.

Companies like OpenAI, Google, Microsoft and Amazon are at the forefront of developing these foundational models, which serve as the building blocks for tasks such as natural language processing, computer vision and predictive analytics.

There are not many companies that develop these foundational models or at least within my knowledge. So, it is not worth looking into it further.

🔐Cybersecurity

As AI technologies advance and become more integrated into various sectors, the need for robust cybersecurity measures to protect sensitive or proprietary data and systems become increasingly critical.

Especially when AI models are trained on vast amount of data. Cybersecurity plays a crucial role in protecting these data from tampering.

There are 7 core segments in cybersecurity:

• 🪪Identity & access management

• 💻Endpoint security

• 🛜Network security

• 📱Application security

• 🗂️Data security

• ☁️Cloud security

• 🛡️Security operations

Each of these segments is interconnected. For example, a user wants to access company’s resources. He uses the company’s laptop to access it by typing in his password (🪪identity & access management).

Then, the laptop connects to the internet via a Virtual Private Network which encrypts the internet connection and hides the IP address. This makes it less susceptible to interception by cyber threats when downloading data (💻Endpoint security & 🛜Network security). I think you got the picture.

I like how Public Comps - website that organizes and visualize SaaS metrics - put in place an overview of cybersecurity value chain, credits to the writer (Eric Flaningam):

Companies like Palo Alto Networks, CrowdStrike, and Fortinet are leading the charge in developing AI-driven cybersecurity solutions that proactively detect and mitigate threats in real-time.

Investment Outlook

Similar to all the above segments, companies within this cybersecurity space often has highly recurring and “sticky” business model. Companies do not often change their cybersecurity providers.

Instead, companies often engage with multiple service providers to diversify their cybersecurity defenses. This approach minimizes the risk of disruption in the event of a cyberattack, ensuring greater resilience against potential threats.

The global cybersecurity market is projected to grow at a compound annual growth rate (CAGR) of approximately 14.3% from 2024 to 2032. This growth is driven by the increasing number of cyber-attacks, the proliferation of e-commerce platforms, the emergence of smart devices, and the deployment of cloud technologies.

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Additionally, please note that the author holds a position in the discussed stock, and his view may be biased as a result.

• Revenue declined by 1.4% YoY, with flat performance in consumables and instruments

• The company faces challenges from US funding cuts, tariffs, and China's regulatory restrictions on its sequencing devices.

• Management has lowered its full-year guidance due to these obstacles but continues to implement mitigating strategies.

Illumina has released its 1Q25 earnings. Doing an update on their latest performance. If you want to know in-depth about their business, click here.

📊Financial Highlights

Most of its revenue segments are either flat or down. Its free cash flows are down by 35% when compared against 4Q24 free cash flows (at $322 mil).

There are 3 main issues that cause Illumina’s poor results:

#1: US Funding Cuts💵

In my previous post, I highlighted that Illumina’s target customers can be broken down into 2 segments:

• Research & Applied

• Clinical

The Clinical segment is more resilient, as it is directly integrated into healthcare applications such as oncology testing. In contrast, the Research & Applied segment is cyclical because its customers are research institutes. Their spending on Illumina’s products depends on government funding.

The Trump administration has been cutting the National Institutes of Health (NIH)’s budget. To date, they terminated $1.81 bil of grants. To make matter worse, Trump’s new budget proposal is calling for an overall reduction in NIH funding to $27 bil which is approx. $18 bil cut.

This funding cuts have forced research institutions to cut down their sequencing activities. Thus, lower sequencing instruments & consumables demand.

#2: Tariffs👮

The baseline import tariffs of 10% are increasing costs for Illumina. The management is estimating an increase of $85 million in operating costs for FY2025.

#3: China’s Regulatory Restrictions🚫

In retaliation against US’s 145% import tariffs against China, the Chinese government has banned Illumina from importing its sequencing devices into the country. It is safe to assume that this market is gone completely.

Mitigating Actions📋

Below are the on-going actions undertaken by the management:

🎯Future Guidance Lowered

Due to the above 3 issues, the management has lowered their full year and next quarter’s guidance:

⚖️Valuation

As mentioned in my previous post, this remains a waiting game. I still believe that the challenges Illumina is currently facing are temporary. How long until resolution? My best estimate would be until Trump is replaced with a new president – in 4 years’ time? Hahaha…

Despite the above uncertainties, I am not adjusting my valuation of Illumina. This is because my previous assumptions are already conservative (at the worst case scenario). The intrinsic value remains in the range of $156.98 to $180.75.

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Additionally, please note that the author holds a position in the discussed stock, and his view may be biased as a result.

• Hardware powers AI by supplying the computational strength required for ML and DL.

• The semiconductor value chain comprises fabless, foundries, and OSAT companies.

• Supportive segments — IP suppliers, design tool vendors, and equipment providers — enable chip production.

• AI hardware revenue is set to grow at a 24.3% CAGR, driven by major tech investments and policy shifts.

This is the first instalment of my in-depth exploration of the AI value chain, beginning with hardware.

Hardware forms the backbone of AI development, providing necessary computational power to train and deploy AI models. These hardware are mainly chips designed to handle the massive data and complex computations required by ML and DL algorithms.

In a typical computer, there are 3 major components (FYI, these are also collectively known as semiconductor chips):

• 💻Computation (CPU and GPU)

• 💽Memory (RAM)

• 💾Storage (Drives)

The computation component (in particular, GPU) handles the training of AI models. While the storage drives are used to store massive data required for ML and DL. The memory ensures efficient processing by reducing the time to access data from slower storage drives.

Additionally, specialized AI accelerator chips are often used as well to handle more intensive computational demands.

Below are some of the specialized chip design for AI development purpose:

Companies that design and manufacture these chips are those from semiconductor industry. Like any industry, it has its own value chain. Below is an overview:

The semiconductor value chain has 3 core segments:

✍Fabless

The term "fabless" refers to companies that design and market semiconductor chips but do not own the fabrication facilities (fabs) to manufacture them. Instead, they outsource the manufacturing process to specialized foundries.

This business model allows fabless companies to focus on innovation, research, and design without the significant capital investment required for building and maintaining fabs.

Some of the notable fabless companies include NVIDIA, Qualcomm, and AMD. In recent times, AAPL also is considered to be a fabless company due to their introduction of Apple Silicon in all their hardware.

🏭Foundry

Foundries are specialized companies that manufacture semiconductor chips for other companies, including fabless firms. They provide the infrastructure and expertise needed to fabricate chips according to the specifications provided by their clients.

Below are the leading foundries:

This is one of the bottleneck that contributes to the shortages of semiconductor chips worldwide during the pandemic. Currently, only TSMC and Samsung are capable of producing advanced chips which are 5nm and smaller nodes.

TSMC produces an estimated 90% of the world's super advanced semiconductor chips which are used to power AI applications.

🔬Outsourced Semiconductor Assembly & Test (OSAT)

OSAT companies provide third-party services for the assembly, packaging, and testing of semiconductor devices. They bridge the gap between semiconductor foundries and end consumers by ensuring that chips are properly packaged and tested before being shipped to customers.

Major OSAT companies include ASE Technology Holding, Amkor Technology, and Powertech Technology. The OSAT market is driven by the increasing demand for advanced packaging solutions and the growing complexity of semiconductor devices.

However, there is little to no product differentiation. Hence, companies in this segment often compete by being the lowest costs producer. They do not have pricing power.

There are 3 additional segments that complement and support the 3 core components mentioned above. They are known as:

📱IP Suppliers and IC Design Tools

The manufacturing process of a chip begins with the creation and verification of a digital integrated circuit (IC) blueprint. A photomask is created with the IC design pattern on it.

An ultraviolet light is then used to print the IC design from the photomask to the silicon wafer. Thereafter, the silicon wafer will go through process like etching and deposition to form an operational chip.

Below is an example of Apple M3 silicon IC design patterns:

Tools like Electronic Design Automation (EDA) are essential for such blueprint design because it allows engineers to simulate the circuitry until the layout is perfected. Companies that provide such tools include Cadence Design Systems, Synopsys, and Mentor Graphics.

If companies do not prefer to create IC blueprint from scratch, they can choose to purchase the license from Intellectual property (IP) suppliers. These suppliers provide pre-designed and pre-verified building blocks, known as IP cores, that semiconductor companies integrate into their own chip designs.

IP cores are like “Lego pieces” where it has pre-made blocks like CPU core, a memory controller or a communication interface. Companies just need to focus on the parts that make their chip unique while relying on proven blocks for common functions.

Some of the leading IP suppliers include ARM, Synopsys, Imagination Technologies, CEVA Inc and Cadence Design Systems.

🛠️Materials and Specialized Equipment Suppliers

This segment supplies the raw materials and specialized equipment needed for semiconductor manufacturing, including silicon wafers, photomasks, and deposition equipment.

Companies like Applied Materials, ASML, and Lam Research are some of the key players in this segment. Especially ASML. This company is a specialized equipment supplier which manufactures and sells ultraviolet lithography machine that are used to print the IC design onto the silicon wafer.

Currently, ASML holds a monopoly in supplying extreme ultraviolet (EUV) lithography systems, which are essential for manufacturing advanced semiconductor devices. These machines are the only one capable of printing extremely fine integrated circuit patterns on silicon wafers, enabling the production of chips at 7nm process nodes and below.

There are no other suppliers that manufacture such machine.

🔎Testing Equipment Suppliers

Companies in this segment provide tools and systems used to test semiconductor chips, either during or after production. This includes automated test equipment (ATE), handlers, and probers.

Testing equipment ensures that semiconductor chips meet quality and performance standards before they are integrated into electronic products. Companies such as Advantest, Teradyne, and Cohu are prominent suppliers in this segment.

Unfortunately, it’s a tough business. Companies in this segment depend heavily on foundries’ orders and there aren’t many big foundries in the world. When there are many companies competing for a share of the foundries’ orders, price wars are inevitable. Of course, this only happens when the demand for chip starts to peak.

📊AI Hardware to Grow at 24.3% CAGR

The AI hardware are projected to grow at a CAGR of 24.3% from 2024 to 2033 in terms of revenue. The US tech giants like Alphabet, Microsoft, Meta, and Amazon are still increasing their AI and data center investments to $320 billion in 2025, up from $230 billion in 2024.

This is despite of the Deepseek-R1 model – China’s version of ChatGPT-o1 model – launched in January 2025 which raises questions about the necessity of exorbitant AI development budgets. The AI model claims to have only spend approximately $6 million in training cost while OpenAI’s ChatGPT costs an estimated $100+ million.

However, there’s another concern that could impact the growth of these chip makers’ revenue which is the U.S. AI Diffusion Policy. It was announced in January 2025 and is expected for US companies to comply starting 15 May 2025.

The policy essentially restricts the export of advanced chips used to develop AI. It divides countries into 3 tiers:

On the bright side, Trump’s administration has announced on 7th May 2025 that they plan to replace this AI Diffusion Policy with a new policy.

It remains to be seen what does the new policy entails and its potential impact. Only time will tell…

Disclaimer

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Additionally, please note that the author holds a position in the discussed stock, and his view may be biased as a result.

🤖What is Artificial Intelligence (AI)?

According to Google, they defined AI as follows:

“It is a field of science concerned with building computers and machines that can reason, learn, and act in such a way that would normally require human intelligence or that involves data whose scale exceeds what humans can analyze.”

AI is not a singular technology but several subsets, each with its own scope and applications. But the foundation to all AI subsets are Machine Learning and Deep Learning.

• 📊Machine Learning (ML): Involves teaching computers to learn from data & then making predictions or decisions without being explicitly programmed. There are many types of ML such as supervised, unsupervised or semi-supervised learning as well as reinforcement learning

• 🧠Deep Learning (DL): This is a subset of ML. It’s inspired by the way human brain works, using neural networks to analyze large amounts of data. It can analyze more complex patterns

Because of ML and DL, we have applications or other AI subsets such as generative AI, robotics, computer vision, natural language processing, etc.

🌊The Three AI Waves

Spear Invest – a fund management firm, has identified that there are 3 AI waves:

The first was about “Prediction” where AI models were trained to recognize patterns in data and predict the outcome. For example, recognizing images, understanding speech, recommending video or item to buy, etc.

The second wave is about “Generative”. This is what we are currently experiencing with all the generative AI models from openAI, Google, Microsoft, Deepseek, etc. These models can learn patterns and generate new content such as images, videos, etc.

The next wave would be “Physical”. This is where AI interacts with physical world such as robotics and autonomous driving. While we have seen these already in action, it is still not perfect.

For example, there are five levels of autonomous driving technology:

• Level 1: The lowest level of automation, providing basic driver assistance like adaptive cruise control.

• Level 2: Advanced driver assistance where the vehicle controls steering and speed, such as Tesla’s Autopilot.

• Level 3: Conditional driving automation where vehicles can make decisions based on the environment, but human intervention is needed if the system fails.

• Level 4: High driving automation, such as driverless taxis operating independently within approved areas and specific conditions.

• Level 5: Full driving automation, where vehicles are entirely self-driving and require no human input, though this remains a goal yet to be achieved.

Most modern cars today are still in level 2 where human attention is still required. In this sense, we are still at the early stage of physical AI.

⚛️The AI Value Chain

Developing an AI application begins with choosing an appropriate ML or DL algorithms. The algorithm is then undergo training using prepared “data” and continuously tuning it to improve accuracy. Thereafter, it is deployed.

A powerful AI application depends on 3 key factors:

• 📊Quality, diversity and quantity of data used for training

• 🛠️Methodologies and algorithms employed during training

• 💾Computational resources and infrastructure supporting the training process

The third factor represents a critical bottleneck with substantial investments from major tech companies worldwide for its development.

This is because data is the lifeblood of all AI applications, and robust data infrastructure is essential for collecting, storing, processing, and managing the vast amounts of data required for ML and DL.

From an investor’s standpoint, this area presents promising opportunities to identify high-growth stocks. It includes businesses that provide services like cloud storage or create special computer chips designed for AI.

These areas are getting a lot of attention because they play a key role in advancing AI technology, making them promising options for long-term investments.

Below is an overview of the AI value chain. It is categorized into 3 broad areas; hardware, data infrastructure and applications:

Read more about each of this area in-depth starting with:

1. AI Hardware: Semiconductors

2. AI Data Infrastructure

3. AI Applications

Disclaimer

The information provided in this blog post is for educational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

• Financial Performance: ILMN's FY2024 revenue was $4,372 million, down 2.9% from FY2023. Free cashflow will soon reach its historical high of $891 mil

• NovaSeq X Transition: Sales have slowed, with 278 orders shipped in 2024 compared to 352 orders in 2023

• China Market Challenges: Revenue from China decreased, contributing only 7% of total revenue in FY2024. Trade tensions have impacted ILMN's share price

• Valuation: Despite disappointing revenue growth, the company’s valuation remains attractive but might take longer to unlock value

In my last post, I discussed the potential turnaround ILMN might experience. Since then, they have released their Q4 2024 results. Here’s an update on their performance so far.

Financial Highlights

The lower reported revenue in FY2024 was mainly due to lower instrument sales. Nevertheless, this is offset by an increase in consumables of approx. 2% y-o-y.

On the bright side, free cashflows have jumped when compared against FY2023. If this continues, it will soon reach its historical high free cashflows of $891 mil (in FY2020).

Higher free cashflows allow the company to pare down its debt or perform stock buybacks. During the year, it has repurchased 134,000 shares at average price of $129.02. This is good, as it will increase the EPS. Hope to see more in 2025 onwards.

NovaSeq X Transition Progress

There seems to be a slowdown in sale of NovaSeq X series instruments which led to the decline in instrument sales. In 2024, there were 278 orders shipped. This is lower than in 2023 when 352 orders shipped. The total installed base now stands at 630 units.

As of 4Q24, approx. 40% of high-throughput consumables revenue was on NovaSeq X series. Unfortunately, ILMN does not disclose the breakdown of consumables revenue by instrument type. Regardless, at just 40% of high-throughput consumables, it means that there are still room for growth in instrument sales.

Market Headwinds - China woes

China market continues to decline in revenue. For FY2024, it only contributed 7% of total revenue. This is a 2% reduction from FY2023.

The recent announcement of 10% tariff on China’s products by Trump has made the Chinese government to retaliate. Unexpectedly, ILMN was placed under the “unreliable entity list” which could risk it being sanction.

Because of this, ILMN’s share price tanked further. I guess the market is pricing ILMN as if there’ll be no more revenue from China going forward (at least during the trade war period).

ILMN’s CEO said that they are currently in a dialogue with relevant parties from Chinese Ministry of Commerce to reach a solution.

Regardless, China only accounts for 7% of total revenue and it has been struggling to recover even before this issue. So, this will be a bonus for ILMN if ever it could recover. Even if it doesn’t, the max exposure is still just 7%. Not a big deal in my opinion.

2025 Full Year Guidance

Some catalyst to boost its revenue are:

• New consumables (e.g. flow cells) for advancing multiomic applications

• MiSeq i100 was launched in 4Q24 and so far, has already placed 70 units

Lowered Valuation

It’s a rather disappointing result, especially on the revenue growth. The only positive is its free cashflow soon reaching historical high level of $891 mil in FY2020. It looks like ILMN is going to take a detour before fully turned around.

Taking the worst case scenario as follows, the intrinsic value is now lowered to between $153.47 to $180.75:

• Illumina completely out from China market (7% reduction in revenue) while revenue from other regions remain flat. The revenue per share would be approx. $25 per share for FY2025

• Cashflow from operation continues to increase albeit at slower pace since China market is assumed to be doomed. I’m guessing at $6 per share for FY2025

• Using -1 STD of 15 Yr median Price-to-Sales & Price-to-CF Ratio of 6.1x and 30.1x, respectively

With this extremely conservative valuation, ILMN is still undervalued with margin of safety between 34% and 44% at current price of $101.78 (at the time of writing). Nevertheless, this is a waiting game now for those who are vested. I will continue to update the company’s 1Q25 result.

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Additionally, please note that the author holds a position in the discussed stock, and his view may be biased as a result.

• Refocused strategy – Spin-off of Grail allows renewed focus on core Illumina business operation

• Revenue growth potential – Transition to NovaSeq X series progress well

• Improving margins – There are signs of recovery in terms of non-GAAP gross profit margin and operating margin

• Valuation upside – Conservative estimates suggest there’s approx. 57% average margin of safety based on share price at the time of writing (December 2024)

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Additionally, please note that the author holds a position in the discussed stock, and his view may be biased as a result.

Business Overview

DNA is a molecule found in every cell of every living organism. It contains the genetic code that determines an organism's traits, such as eye color, hair color, and height. DNA is also responsible for passing on traits from parents to offspring.

Sequencing DNA means reading the order of the "letters" in the genetic code. This information can be used in many ways, including:

• Understanding diseases: Identifying genetic mutations linked to diseases like cancer or Alzheimer's.

• Developing new medicines: Creating targeted therapies based on a person's unique genetic makeup.

• Improving healthcare: Personalizing medical treatments and predicting an individual's risk for certain conditions like cancer.

• Agriculture: Improving crop yields and developing disease-resistant plants.

Illumina develops and manufacture machines (they called it “instrument”) that can quickly and accurately read DNA sequences. These machines have revolutionized the field of genomics, making it possible to study DNA on a large scale at lower costs. There are 3 categories of instruments developed by Illumina:

1. High-throughput

2. Mid-throughput

3. Low-throughput

Along with the instruments are its consumables – flow cells and reagents. The former are glass substrates where sequencing reactions occur. While reagents are chemicals used in the sequencing process.

The company generates revenue by selling these instruments and consumables. A bulk of its revenue is from consumables. This business model is like the “razor and blade” business where the blade gets replaced regularly after usage, giving the company a recurring income.

Illumina also sells a wide range of products to support various aspects of DNA sequencing activities. Below is an illustration of a typical workflow of DNA sequencing & the products offered by Illumina across the workflow:

Apart from the above, the company also provide other services such as:

• Sequencing service: for customers that wish to outsource DNA sequencing workflow

• Training: Tutorial on how to operate the machine

• Support: Maintenance of instruments

Its target customers are in:

• Research & Applied: This includes academic labs, research centers, government institutions, biotech and pharma companies

• Clinical: This includes oncology, reproductive healthcare and Infectious disease & microbiology testing

Below is an overview of its financial as of June 2024:

Illumina seems like a good business to invest in but the market seems to think otherwise. Since Aug 2021, the company’s share price has fallen 82% from its peak of $526 to just $89. At the time of writing (December 2024), the share price has recovered some ground to $134.96.

What Happened? The drama from Grail Acquisition…

In Sep 2020, Illumina proposed the acquisition of Grail for a consideration of $8 billion. Grail is a healthcare company that focuses on developing diagnostic tests for multi-cancer early detection. This proposal prompted the US Federal Trade Commission (FTC) and EU regulators to launch anti-trust investigations due to concerns over potential monopolistic effects.

Despite the ongoing investigations, Illumina proceeded to close the deal in August 2021 without considering the potential consequences. Both the US FTC and EU regulators subsequently ordered Illumina to divest Grail. In addition, Illumina was slapped with a record fine of €432 mil from EU regulator for hastily closing the deal.

Although Grail has significant potential for future growth, it remains a loss-making business with substantial R&D expenditures. Consequently, Illumina incurred a total of $4.7 billion in goodwill impairment related to the Grail acquisition across fiscal years 2022 and 2023. This contributed to Illumina reporting substantial losses.

The acquisition of Grail was also criticized by activist investor Carl Icahn, who launched a proxy fight and successfully secured a seat on Illumina’s Board of Directors. Unsurprisingly, the company’s CEO, Francis deSouza, resigned shortly thereafter.

With an activist investor now influencing the Board, Illumina finally spun off Grail in June 2024. I believe this drama has diverted the management’s focus from its core business, resulting in stagnant revenue since 2021.

Fun Fact: Grail was founded by Illumina in 2016 & was later spun-off in the same year!

Macro Headwinds

This flattish revenue is also the result of below macro-environment that Illumina is facing at the beginning of 2022:

• Unfavorable exchange rates as USD strengthened against other currencies

• Lower covid surveillance demand as the world recovers from pandemic

• China market disruption as the country strive to achieve zero covid cases

• China market faces strong competition in mid-throughput

• Russia-Ukraine war has led to company unable to conduct business due to sanctions

These macro headwinds have affected its customers’ funding priorities. Thus, limiting the demand for instruments and consumables.

APAC and China market are hit the hardest compared to other region. From 2009 to 2020, the average revenue contribution from both markets has been approx. 18% of total revenue. But this has drop to just 9% as of fiscal year 2023.

Transition to NovaSeq X Series

In addition to facing macroeconomic headwinds, Illumina has been navigating a transition to its newly announced NovaSeq X series instruments since 2022. Customers using the previous version of high-throughput instruments (NovaSeq 6000) have been reducing their purchase of consumables as they migrate to the new high-throughput instruments, with plans to ramp up sequencing activity again once the transition is complete.

Can Illumina Turnaround? Maybe…

From 2009 to 2019, Illumina was growing at CAGR of 18%. Since then, its business has been flattish. For the company to revert back to its growth story, it needs to resolve the above 3 challenges.

Now that Grail’s spin-off has completed, there’s 2 more challenges to overcome:

Macro Headwinds

These are uncontrollable by the company. It will continue to affect customers’ purchasing behavior. With 2025 expected to be more challenging due to Trump 2.0, I think Illumina’s growth will continue to be tapered.

One concern is Illumina’s China market. Even if the Chinese government showing initiatives to revive the economy, Illumina still needs to overcome the strong competition in the mid-throughput segment.

In 1Q24, the management has brought in new leader to manage China market. This needs to be monitor closely.

Transition to NovaSeq X

Based on the company’s August 2024 strategy updates, there are approx. 2,180 active high-throughput instruments installed as of 30^th June 2024. So far, they have shipped a total of 527 NovaSeq X series as at end of 3Q24. This means the migration is only at roughly 24% completion.

Already there are signs that this transition to new instrument has been progressing well as of 3Q24, for instance:

• Out of the 527 installed base, 40% are from clinical customers. This is good as clinical market tends to be more stable in terms of sequencing activities, especially in the oncology test which is the fastest growing segment

• Approx. 35% of consumables revenue is from NovaSeq X. The management anticipates almost half of consumables revenue will be transition to NovaSeq X by mid-2025

This transition matters to Illumina because I believe historically high-throughput instrument & its consumables represent the largest portion of the company’s revenue. Unfortunately, the company does not provide such breakdown. I can only make my assumption.

Other Recovery Signs

When it comes to turnaround play, monitoring the financials closely for recovery signs is key. In the August 2024 strategy updates, the management highlighted some recovery signs besides the transition progress to NovaSeq X:

• Non-GAAP gross margin has been improving from 66% in FY2023 to 70.5% as of 3Q24

• Non-GAAP operating margin also seen improving from 20% in FY2023 to 22.6% as of 3Q24

In addition to the above, I noticed the company’s free cashflows have been improving. As of 9M24, it generated a free cashflows of $374 mil which is higher than FY2021. Perhaps it could recover back to pre-covid level of free cashflows by FY2026, which then allow the company to buy back more shares.

Another big positive for Illumina is that they avoid paying the €432 mil fines as the EU court has ruled in favor of Illumina on September 2024, stating the EU regulator has no jurisdiction over the acquisition. While this comes after Illumina has spun-off Grail, I think it’s a blessing in disguised. With Grail under Illumina, it would make the financials look bad.

Nonetheless, looking at the below financial targets for FY2025 to FY2027 set by the management, it seems like Illumina’s historical double digit growth rate (CAGR at 18%) will not happen anytime soon.

Valuation

Before Illumina faces all the above challenges, its 14-year median price-to-sales & price-to-operating cashflows were at 11.8x and 43x, respectively.

Assuming Illumina would turnaround & go back to its pre-Grail valuation as above, I think the intrinsic value will range from $293 to $335. Below are some additional assumptions made:

• Revenue will be flattish despite management anticipating it will grow at high single digit. This is because I’m trying to be conservative in my valuation

• Cashflow from operation will return back to pre-covid level (within the range of ~$1,000 mil). Considering the growing free cashflows in its latest 9M2024, I think it is achievable

What can go wrong?

There are 3 things that I will monitor closely:

• Installed base of NovaSeq X – This will have direct impact on Illumina’s consumables revenue. Unexpected slowdown in the transition would impact the company’s future consumables demand

• Macro headwinds – Though beyond the company’s control, this plays a significant role in Illumina's turnaround narrative. Any additional economic headwinds could further constrain customers' funding

• China market – If Illumina is able to overcome the stiff competition here, it would be a bonus for them. As of FY23, China account for just 9% of total revenue

  ---

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.

Additionally, please note that the author holds a position in the discussed stock, and his view may be biased as a result.

• Trump 2.0: Trump's tariffs on Canada, Mexico, and China could cause higher inflation leading to higher for longer interest rates

• US-China Trade War: Expected to escalate, with potential retaliation from China

• Malaysia Market: Inflation expected to rise due to removal of blanket subsidy on petrol & diesel and higher minimum wage hike

• Strategy: Focus on investing in sectors that align with long-term secular trends

Since I’m starting this journal, I want to make it a point to do a yearly market outlook on markets that I’m currently invested in – US, SG, MY and HK. So far, below are the events happened in 2024 that I believe will have lasting impact in 2025 and beyond:

Global Markets (US, SG and HK)

• US Fed has cut its interest rate by another 0.25% this year. Currently, the interest rate is at 4.25% to 4.50%. They also signal fewer rate cuts in 2025

• Trump won the US presidential election. His republican party also gain control of the Congress. It’s a clean sweep

• China government unveils various stimulus packages to revive the economy. It seems they are committed to support the economy again

• Geopolitical tension still on-going in Middle East and between Ukraine and Russia. It does not seem to be resolving anytime soon

• In early Aug-24, USD & SGD weakens against MYR but later rebounded in Oct-24

Local Market (MY)

• OPR remained unchanged at 3%. It does not seem like BNM will cut or increase

• RON95 blanket subsidy to be replaced with targeted subsidy in 2025

• Higher minimum wages were announced during budget 2024

• Pay hike for civil service (effective December 2024)

• Many export oriented companies’ financials got hit by forex losses due to MYR strengthening

Trump 2.0

Perhaps the most impactful event would be Trump winning the US presidential election. From my reading on countless articles and videos explaining about his policies, the one keyword that keeps appearing is “Tariffs”.

These are taxes imposed by the US government on goods imported into the country. Trump intend to raise the tariffs on all goods from US’s 3 largest trading partners – Canada, Mexico and China. Especially China, considering they are still in a tech-war.

Because of this, I expect below could happen this year (2025):

Inflation making a comeback

Inflation will make a comeback and the US Fed would have to slow their rate cut or, in worst case scenario, increase the interest rate.

• If this indeed happens, USD will strengthen against other countries’ currencies. This is good for export oriented companies in Malaysia.

  • But then again, high tariffs are not good for export companies because it will make the product less competitive in terms of price. An example is how Malaysia charges an exorbitant taxes on continental cars making it less appealing in terms of price when compare against local car makers such as Proton.

  • While stronger USD can increase the export companies’ revenue when it translates back to its local currency, having lower revenue could potentially offset this gain.

  • Companies that have strong moat such as switching costs, network effect & economies of scale would continue to do well.

• For Singapore REITs, I expect their earnings to continue facing pressure by the higher for longer periods of borrowing costs. On the bright side, this would be offset by those REITs that are able to negotiate for positive rental reversion higher than the increase in borrowing costs.

US-China trade war to escalate

There’s no doubt that this would happen. At the time of writing this journal (December 2024), Trump has already made some comments on the tariffs that he will impose once he inaugurates.

• I expect China to retaliate further but what they will do remains unknown to me. But the first casualty will be US companies that are in semiconductor sector.

• According to Reuters, below table lists S&P500 companies with largest China exposure in fiscal 2023:

• Another company not in the list but has the highest exposure to China is ASML. It is a Dutch company with 47% of its total revenue comes from China, as of 3Q24.

• Interestingly, leading AI chipmakers such as NVDA, AMD and Micron have the lowest exposure to China compared to the list above:

• I guess having position in these two companies, NVDA and AMD, are not as bad as I thought it would be.

Malaysia Market Outlook 2025

Back in Malaysia, I think inflation would rise due to:

• Removal of blanket subsidy to RON95 petrol

• Higher minimum wage hike

• Pay hike for Civil Service

• Higher tax on sugary beverages

Despite this, I expect BNM will maintain its OPR at 3% given how prudent they have been. I still remember banks globally are increasing their key interest rate but BNM chose to delay it. This also explains the weakening of MYR.

While MYR has recovers against USD, I suspect this is just temporary given the high inflationary policies to be adopted in Trump 2.0. I expect companies in Malaysia that have made huge forex losses recently to recover back again in 2025.

However, companies where its input costs are USD denominated would underperform. I guess brewery sector will be flattish in second half of next year.

Data Centre Boom

Perhaps the most notable trends that are growing fast in Malaysia is Data Centre. According to DC Byte’s 2024 Global Data Centre Index, Johor is the fastest growing market within Southeast Asia.

I think companies within the Data Centre value chain will continue to do well. Especially construction, utilities, telco and renewable energy sector. Regretted that I did not invest into Tenaga back in pandemic time.

My Plan for 2025

When investing, my goal is to position my portfolio to capitalize on long-term secular trends. This yearly market outlook doesn’t prompt me to alter my portfolio positions; rather, it only serves as assurance whether my long-term view has change or not.

Furthermore, the markets are dynamic. Anything could happen. Market crashes are often caught by surprise. With this in mind, I think the uncertainties in the US could present me with opportunities to invest into stocks that are beneficiary of long-term secular trends.

For Malaysia market, I continue to like dividend growth stocks. If opportunities exist, I would like to position my holdings more into sectors that are within Data Centre value chain.

Disclaimer:

The information provided in this blog post is for informational purposes only and should NOT be construed as financial advice. Investing in stocks and ETFs involves risk, and there is no guarantee of profits. Past performance is not indicative of future results. It is important to conduct thorough research or consult with a qualified financial advisor before making any investment decisions. The author is NOT a financial advisor and is sharing his personal experiences and opinions only.