From Physical Air Cargo Network to Digital Twin and AI: The Trillion-Dollar Supply Chain Transformation Behind FedEx's Unified Network

This material is very worth digging into because FedEx is undergoing the most important "business model redefinition" in its 53-year history. Fred Smith's first phase addressed: How to reliably deliver a package from point A to point B in a very short time. What Raj Subramaniam needs to solve after taking over is: How to combine airplanes, trucks, warehouses, packages, customs, weather, routes, and decades of logistics data into an intelligent network that can predict, schedule, and optimize the global supply chain. This is no longer a traditional "upgrade of a courier company." It is closer to: Upgrading an industrial infrastructure company to an operating system for the physical world. ──────────────── 1. First, calibrate three very important facts 1. The "$2 trillion logistics empire" in the title can be misleading. Raj said in an interview with David Rubenstein in April 2026: The value of goods transported by FedEx in a year is close to $2 trillion. Not: • FedEx's revenue is $2 trillion; • FedEx's market value is $2 trillion; • FedEx manages $2 trillion in assets. This is the Value of Goods Moved. He mentioned that these goods mainly include pharmaceuticals, aerospace, defense, textiles, etc., characterized by: High Value per Pound. That is, the economic value corresponding to each pound of weight is very high. So I suggest upgrading your title to: Precise and Informative Transporting nearly $2 trillion worth of goods annually: How FedEx CEO Raj Subramaniam is reconstructing the global logistics network Or more in-depth: I highly recommend From a courier company to a global supply chain operating system: How FedEx is reconstructing nearly $2 trillion in goods flow with AI This is much more accurate than "Steering a $2 trillion empire" and also more sophisticated. ──────────────── 2. The second important correction: Today's One FedEx can no longer include FedEx Freight. This is the most needed update in this content as of August 2026. Historically, FedEx indeed had: FedEx Express, FedEx Ground, FedEx Freight, FedEx Services and other independent operating systems. On June 1, 2024: FedEx Ground + FedEx Services officially merged into Federal Express Corporation. In other words: The systems of Express and Ground in aviation, ground, sales, technology, etc., began to truly integrate into One FedEx. However: FedEx Freight was officially spun off and listed on June 1, 2026. Stock code: FDXF. FedEx allocated 80.1% of FedEx Freight's shares to original shareholders while temporarily retaining 19.9%, planning to dispose of it within two years. So the real strategy now is not: "To merge all businesses together." But rather a more sophisticated statement: Integrate what should be integrated, and separate what should be separated. ──────────────── 3. This actually reflects Raj's very strong capital allocation logic. Why: Express + Ground need to merge, But Freight needs to be spun off? Because capital allocation is not: Bigger is Better. But rather: Economic Similarity Matters. Although Express and Ground differ in speed and transportation methods, they essentially belong to: Parcel network. A package can: Be transported by plane for a segment, Be transported by truck for another segment, And finally delivered to the door. There are many shared resources between the two networks: Terminals, Routes, Customers, Drivers, Information systems, Sorting facilities. So merging has: Network Synergy. ──────────────── 4. However, LTL Freight is a completely different economic model. FedEx Freight deals with: Less-Than-Truckload, or LTL shipping. It is not a small package. It could be: A pallet, Hundreds of pounds, 1,000 pounds of goods, Industrial parts. Its core operational metrics include: Revenue per Shipment, Weight per Shipment, Revenue per Hundredweight, Terminal Density. In FY 2026, FedEx Freight's revenue was about $8.8 billion, with adjusted operating profit of about $1.1 billion. The mid-term goals set by the company post-spin-off include a 4%-6% revenue CAGR, a 10%-12% adjusted operating profit CAGR, and over $1 billion in FCF. So even though both networks are called FedEx, Their capital attributes are actually different. ──────────────── 5. Here we can learn a very advanced principle of corporate strategy. Many CEOs make the mistake of: Wanting to integrate everything. The truly excellent CEOs ask: Which asset combinations can produce 1+1>2 in the future? And: Which assets, when combined, actually drag down valuation and resource allocation? So Raj is simultaneously doing two seemingly contradictory things: One FedEx And Spin-off FedEx Freight. In fact, there is no contradiction at all. The logic is: Merge assets with strong Operational Synergy; separate assets with weak Capital Market Synergy. This is something that entrepreneurs should learn from. ──────────────── 6. What FedEx is really transporting today is not "packages," but the value of time. This is the most important layer to understand FedEx's business model. Assume: You are transporting a T-shirt worth $15. A day earlier or later: The economic impact is limited. But assume you are transporting: A part of an aircraft engine worth $80,000. If the plane is grounded for a day because it lacks this part, The loss could be: Hundreds of thousands of dollars or more. Then: Paying $1,000, $3,000, or even $10,000 for urgent logistics could be very reasonable. ──────────────── 7. This is what is called Value Density. Very important. Value Density = Value of Goods / Weight or Volume. For example: Steel: Heavy, Low value per unit weight. Semiconductors: Light, Extremely high value. Pharmaceuticals: Very light, Certain products have extremely high value. Aerospace parts: High value + extremely time-sensitive. So FedEx's proactive move towards: Healthcare, aerospace, automotive, data centers is not accidental. At the 2026 Investor Day, FedEx clearly listed these industries as core high-profit growth verticals, with a target B2B market TAM exceeding $130 billion. ──────────────── 8. Therefore, what FedEx is really selling is not Transportation But rather: Time Certainty. What customers are really buying is: "It must arrive before tomorrow afternoon." "The temperature cannot exceed a certain range." "The factory cannot stop production." "It must arrive before the surgery." "The data center construction cannot be delayed because of a part." So logistics economics should really be written as: Transportation Cost Inventory Cost Delay Cost Stockout Cost Failure Cost. Many companies are willing to pay more for FedEx, not because they don't know about cheaper logistics. But because: The Failure Cost is too high. ──────────────── 9. A very classic case: Aircraft on Ground There is a term in the aviation industry: AOG——Aircraft on Ground. An aircraft cannot fly due to a broken part. Assume a large commercial airliner is grounded because of a small part. That part might weigh: Only 5 kilograms. Its value could be: $30,000. But the loss caused by the aircraft being unable to operate for a day: Could far exceed the price of the part. Thus: Logistics speed itself becomes a production resource. This is why high-end B2B logistics can have higher profit margins. ──────────────── 10. Medical logistics is even more obvious. For example: Biologics, Vaccines, Cell therapy products, Medical devices. Transportation not only requires: Speed. But also requires: Chain of Custody. Who handled it? When did it arrive? What was the temperature? Were there any delays? Did it leave the specified environment? So medical logistics sells: Reliability + Visibility + Compliance. FedEx disclosed in 2026 that its healthcare business exceeded $9 billion in revenue in FY25 and continues to strengthen its pharma and temperature-controlled transportation capabilities. ──────────────── 11. This explains why Raj actively reduces the obsession with low-value e-commerce packages. A high volume of packages does not mean a good business. A very dangerous business metric is: Volume Growth. Assume: You ship 1 million more low-priced packages every day, But: The delivery distance is long, Residential density is low, Customer bargaining power is strong, And the profit per shipment is extremely low. You might: Be busier and earn less. So Raj's core shift is: From: Volume Maximization To: Revenue Quality. ──────────────── 12. This is actually highly similar to UPS's direction over the past decade. What large logistics companies are really competing for is not: "Who has the most packages." But rather: "Who has the most profitable packages." Healthcare, B2B, Urgent transportation, International express, High-value industrial products are usually more profitable than: Ordinary low-priced residential e-commerce is more attractive. Therefore, when researching logistics companies, one cannot only look at: Package Volume. One must also consider: Revenue per Package Cost per Package Route Density Yield Operating Margin. ──────────────── Thirteen, FedEx's most terrifying moat is not its planes. Many people look at FedEx: 700 aircraft. 200,000 vehicles. Thousands of facilities. And thus believe: The planes are the moat. Not true. Planes can be purchased. Trucks can also be bought. What is truly difficult to replicate is: Network Density. ──────────────── Fourteen, why is network density so important? Assume: Company A has only 100 packages in a certain area of Los Angeles in a day. Company B: 10,000 packages. With the same driver going out: For Company B, every mile driven, more deliveries can be completed. Thus: Cost per Stop decreases. After costs decrease: Lower prices can be offered. Lower prices: Attract more customers. With more customers: Package density continues to rise. Forming: Density Flywheel. ──────────────── Fifteen, this is the most important economy of scale in the logistics industry. It is not simply: "Big companies buy fuel cheaply." The real economy of scale comes from: Route Density. Hub Utilization. Aircraft Utilization. Linehaul Utilization. Sortation Utilization. Pickup Density. So even if newcomers finance: $10 billion, they may not easily replicate FedEx. Because you can buy trucks, but you cannot buy: The volume of goods that already exists on every street every day. ──────────────── Sixteen, Fred Smith's greatest innovation: Hub-and-Spoke. What truly changed the world for FedEx was not just: "Overnight delivery." But rather: Hub-and-Spoke. Previously: From city A to city B, A to C, A to D, required a lot of direct routes. FedEx centralized packages to: Memphis Hub. Then re-sorted them. Thus, many cities can be connected through a super center. This is actually very similar to internet exchange centers. ──────────────── Seventeen, what Raj is doing now is equivalent to upgrading Hub-and-Spoke once again. Fred Smith optimized: Physical Routing. Raj is optimizing: Physical Routing + Digital Routing. Previously, logistics companies answered: Where is the package? Now, digital twins need to answer: Where might the package encounter problems next? This is a completely different level of technology. ──────────────── Eighteen, why is FedEx's 2 PB of data so important? Raj stated in an interview: FedEx generates about: 2 Petabytes of data daily. The 2026 Investor Day also confirmed this scale again. But I want to emphasize: Having a lot of data in itself has no value. Many companies generate PB of data every day, but in the end, it is all garbage. What is truly valuable is: Proprietary + Real-Time + Actionable Data. ──────────────── Nineteen, what is special about FedEx's data? It sees: A sudden increase in cargo volume on a certain route. A delay at a certain airport. Abnormal weather in a certain area. Cross-border growth of a certain product. Inventory in a certain industry starting to recover. Congestion at a certain warehouse. A sudden decline in a certain trade corridor. Abnormal order volume from a certain customer. This is not data scraped from the internet. This is: Physical Economy Data. That is, data generated by real-world economic activities. ──────────────── Twenty, thus FedEx's data is completely different from that of Google, Meta, and OpenAI. Google understands better: Search Intent. Meta understands: Social Intent. Amazon understands: Purchase Intent. Visa / Mastercard understands: Payment Activity. FedEx understands: Physical Goods Movement. These data ultimately form different dimensions of an economic map. This is an asset that has long been underestimated by the capital market. ──────────────── Twenty-one, Digital Twin is by no means just "advanced package tracking." The true meaning of digital twin is: In reality, there exists a network: Planes, Vehicles, Airports, Roads, Facilities, Packages, Personnel. In the digital world, a: Virtual Representation is simultaneously established. It maps the entire physical system in real-time. Then you can in the digital world: Simulate, Predict, Optimize, And feed decisions back to the real world. This forms: Observe → Predict → Optimize → Act. ──────────────── Twenty-two, once AI is added, the value will undergo a qualitative change. Previously: "Chicago snowstorm, packages are already late." This is: Reactive. In the future: AI predicts in advance: Chicago may close some flights in the next 12 hours. Thus: Key medical packages are rerouted to Indianapolis in advance. This is called: Predictive. Going further: The system can automatically reconfigure: Planes, Trailers, Trucks, Warehouse space, Sorting resources. This is called: Prescriptive / Autonomous Orchestration. ──────────────── Twenty-three, this is the true meaning of Supply Chain Orchestration. If FedEx ultimately only: "Helps you transport packages," its business model is still that of a logistics company. If FedEx can tell Nike: A supplier in Vietnam is delayed. Los Angeles port may be congested. A batch of inventory will be out of stock in five days. Some orders should be shifted to air freight. Some goods should be transferred from another warehouse. Then FedEx begins to upgrade from: Carrier to: Control Tower. ──────────────── Twenty-four, if this step is successful, the valuation logic may also change. Traditional transportation companies are usually: Capital-intensive, With low profit margins, And clear cycles. Software platforms usually: Have high gross margins, Are capital-light, And have repeatable revenue. FedEx cannot become purely SaaS. But if more and more value comes from: Analytics, Orchestration, Visibility, Automation, Workflow, It can add a: Digital Margin Layer over its original transportation network. This is where the AI strategy truly deserves the attention of the capital market. ──────────────── Twenty-five, this is also why Dataworks is important. FedEx established Dataworks in 2020 with the aim of turning the data generated by its logistics network into usable software and insights. In 2021, FedEx Surround was launched for real-time supply chain visualization and risk management. So to be precise: It was not that in 2020 they suddenly "completed the digital twin." But rather: Starting in 2020, they systematically began to assetize logistics data as independent assets. It then continuously developed into today's digital twin and AI intelligence. ──────────────── Twenty-six, FedEx's true long-term moat is called Data Exhaust. Every shipment generates data. And this data makes the next shipment: More accurate. Once accurate: More customers are willing to use it. With more customers: More shipping data is generated. Thus producing: Physical Network → Data → Better Network → More Volume → More Data. This is a very interesting: Physical-Digital Flywheel. ──────────────── Twenty-seven, it is even a very rare type of data in the AI era. Internet text is becoming increasingly commoditized. Many models can obtain: Web pages, Code, Books, Public data. However: The real data generated by a 50-year global logistics network: Traffic, Delays, Weather, Routes, Packages, Customs, Demand is not easily replicable. So the truly scarce asset in the AI era is increasingly likely not: "Data volume." But rather: Closed-Loop Proprietary Data. ──────────────── Twenty-eight, this is also why AI will increasingly enter the physical industry. The next round of huge value from AI may not all come from: Chatbots. The bigger opportunities may come from: Manufacturing, Logistics, Energy, Healthcare, Defense, Robotics. Because in these industries, a: 1% efficiency improvement, Could correspond to: Billions of dollars. ──────────────── Twenty-nine, why is the scenario for autonomous trucks particularly suitable for FedEx? Raj mentioned: FedEx has already tested autonomous trucking in Texas. The key is not: "The technology for driverless vehicles has finally arrived." What is truly important is: Operational Design Domain. FedEx has a large number of facilities near highways. Raj mentioned in an interview that there are approximately: 638 highway-adjacent facilities. So the scenario can be designed as: Facility A ↓ Highway ↓ Facility B. Highway autonomous driving, with the last mile still driven by humans. ──────────────── Thirty, this is much easier than starting with fully autonomous Robotaxis. Urban autonomous driving faces: pedestrians, bicycles, traffic lights, temporary construction, complex intersections. Highways: the scenarios are relatively standardized. So one of the earliest areas where autonomous driving will truly be commercially scaled is likely: Middle Mile Trucking. Rather than the last mile. ──────────────── Thirty-one, there is also an economics aspect: Driver Utilization. Truck drivers are affected by: working hours, rest, recruitment difficulties. If autonomous driving takes over: continuous hundreds of miles of highway segments, drivers can focus on: pickup, delivery, local operation, then it’s not simply: "cut all the drivers." It is more likely to first optimize: Human Labor Allocation. ──────────────── Thirty-two, Raj mentioned "about 29,000 new jobs added in the U.S. last year," which needs careful interpretation. This is his original statement in the interview. But it cannot be concluded that: "AI led to FedEx adding 29,000 jobs." These two are not the same causal proposition. A more rigorous expression should be: Raj's view is that AI may promote cargo volume and new business growth after improving supply chain efficiency, while the physical network still requires a large number of personnel; he also stated that FedEx added about 29,000 jobs in the U.S. the previous year. This is where facts and opinions need to be separated. ──────────────── Thirty-three, the real revolution of One FedEx is not "changing the organizational structure" It is in eliminating: Duplicate Capacity. Imagine before: A FedEx Express truck goes to the same street. A FedEx Ground truck also goes. Different facilities. Different routes. Different IT. Different management systems. Different network planning. Customers see: FedEx. But inside the company, it’s like two companies. This is a historical legacy issue. ──────────────── Thirty-four, why did it become this way before? Because FedEx Ground was not originally built from scratch by Fred Smith. In 1998, FedEx acquired: Caliber System. Which included: RPS. Later RPS was renamed: FedEx Ground. So Express and Ground essentially come from: two sets of different DNA. After decades of accumulation, integration is extremely difficult. ──────────────── Thirty-five, this illustrates that corporate mergers and acquisitions have a very important long-term cost. Acquiring a company: Deal Closing only takes a few months. But: Organizational Integration may take decades. This is something all entrepreneurs, CEOs, and PE investors should remember: The completion of a merger is not the end of the transaction, but the real beginning of integration costs. ──────────────── Thirty-six, DRIVE and Network 2.0 are addressing this historical burden. DRIVE is closer to: restructuring enterprise-level costs and execution systems. FedEx has achieved a cumulative: $4 billion structural cost reduction in FY24 and FY25. Network 2.0 is more focused on: facilities, routes, sorting, ground network redesign. ──────────────── Thirty-seven, the real indicator to look at in Network 2.0 is not "how many people are cut" but rather: Cost to Serve. This is one of the most important metrics in the logistics industry. For each package shipped, how many: miles, labor minutes, sort touches, facility capacity, linehaul capacity. If the cost per order decreases: $0.20, with 18 million packages a day, over a year, it results in a huge profit change. FedEx's current official data shows that its network averages nearly 18 million packages a day. ──────────────── Thirty-eight, looking at Tricolor, it is actually another very clever capital efficiency reform. FedEx is dividing its international air network into three tiers: Purple highest priority, high-value goods. Mainly using FedEx-owned aircraft. Orange Priority Freight. Slightly slower, truck-fly-truck. White lower time-sensitive goods. More using commercial airline partners. ──────────────── Thirty-nine, the capital logic here is very beautiful. In the past, one way of thinking was: I have planes, so I want to load as much cargo as possible onto my own planes. Raj's thinking is: Only the most return-generating cargo is worth occupying the most expensive owned assets. This is a completely different capital discipline. ──────────────── Forty, why do airlines fear low utilization? An aircraft: After buying it, whether it flies or not, there are: depreciation, financing, maintenance, insurance, crew costs, airport fees. So an aircraft is: High Fixed Cost Asset. The most dangerous situation is: spending hundreds of millions to buy an aircraft, and then loading: low-yield cargo. Thus, Tricolor is essentially optimizing: Yield per Aircraft Hour. ──────────────── Forty-one, this is why Raj now emphasizes ROIC. Traditional large enterprises easily pursue: Revenue. Truly excellent capital allocators look at: ROIC—Return on Invested Capital. Because if: investing $10B, in the last year only earns an additional $200M, the business is large, but value is destroyed. FedEx's target ROIC for 2029 is approximately: 11%. An increase of about 200bp from the FY26 baseline. This is the metric that capital markets should really focus on. ──────────────── Forty-two, the "14% net profit CAGR" in your materials also needs slight correction. The official target for the 2026 Investor Day is: by 2029: Revenue ≈ $98B approximately: 4% CAGR. Operating Income: approximately: $8B. Under adjusted standards, CAGR: approximately 14%. GAAP operating income CAGR: approximately 17%. So it would be more accurate to call this: Operating Income CAGR rather than simply writing: "net profit CAGR 14%". ──────────────── Forty-three, but this 4% vs 14% is one of the most important data points in the entire investment story. Why? Because it means: FedEx is not telling investors: "I want to crazily expand revenue." But rather saying: "I want to let profit growth speed far exceed revenue." This is called: Operating Leverage. ──────────────── Forty-four, it can be understood with the simplest model. Assuming now: Revenue: $100. Cost: $94. Profit: $6. If revenue increases: 4%. Becomes: $104. Costs only increase: 2%. Becomes: $95.88. Profit: $8.12. Revenue only increases: 4%. But profit increases: 35%. Of course, the actual FedEx model is not this extreme, but the principle is: Small Revenue Growth + Slower Expense Growth = Large Profit Growth. ──────────────── Forty-five, so what Raj is really doing is the Margin Story, not the Growth Story. Many tech CEOs talk about: growth. Raj talks about: Productivity. Margin. ROIC. FCF. This is what a CEO of a mature heavy-asset enterprise should be doing. ──────────────── Forty-six, why is $6B FCF in 2029 particularly critical? Because for decades, FedEx has had to continuously build: aircraft, hubs, vehicles, IT, facilities. These consume a lot of cash. Raj said an important thing in the interview, roughly meaning: "We have basically covered this planet." Future capital investment will be more: incremental, rather than: rebuilding the global network. ──────────────── Forty-seven, this is the most beautiful lifecycle stage for large infrastructure companies. First stage: Build the Network. Cash flow is poor. Second stage: Fill the Network. Increase cargo volume. Third stage: Optimize the Network. Improve utilization. Fourth stage: Harvest Cash Flow. If FedEx can truly enter the fourth stage, capital returns may undergo structural changes. ──────────────── Forty-eight, signs of this have already appeared in FY26. FedEx FY26 Capital Expenditure: approximately: $3.8B. A decrease of about: 6%. CapEx / Revenue: 4.0%. The company claims this is the lowest level in FedEx's history. So the real logic behind the $6B FCF target is not: "Suddenly making a lot of money." But rather: Increasing profits while reducing capital consumption. ──────────────── 49. In the investment world, this is a very strong combination. The value of a business ultimately comes from: Cash Flow. And free cash flow can be roughly understood as: Operating Cash Flow − Capital Expenditure. If a company: Has revenue growth, Improves profit margins, While the CapEx ratio decreases, Then FCF will exhibit very strong elasticity. This is called: Free Cash Flow Inflection. ──────────────── 50. The spin-off of FedEx Freight further strengthens this logic. One very smart aspect of this spin-off is: FedEx Freight paid FedEx approximately: $4.1B in special cash dividends before the spin-off. The funds mainly came from new debt issued by Freight, etc. FedEx subsequently used this money to advance: Debt Reduction. So the spin-off is not just: "Giving shareholders a new stock." It is also: Balance Sheet Engineering. ──────────────── 51. As of August 2026, the market has even given FDXF a market value close to $20 billion. Around August 25: FedEx Freight's market value was approximately: $19.6B. Enterprise Value was approximately: $25.7B. This explains why the capital markets favor: Pure Play. A pure LTL company, Investors can separately: Estimate growth, Estimate profit margins, Estimate industry cycles, Estimate capital returns. No longer buried within the large FedEx group. ──────────────── 52. This is the classic Conglomerate Discount problem. If a company contains: Aviation, Packages, LTL, Software, Logistics, Print shops, It is difficult for investors to value. Sometimes after a spin-off: 1 + 1 > 2. It is not that operations have changed. But rather: Capital Market Visibility has improved. GE, eBay/PayPal, United Technologies have historically experienced similar logic. ──────────────── 53. Looking at FedEx's macroeconomic value. FedEx has a very special identity: Real-Time Economic Sensor. GDP data is: Lagging. Quarterly data. FedEx sees every day: Business orders, Industrial parts, Imports and exports, Cross-border trade, E-commerce volumes, Medical freight. So in a sense: FedEx sees: Physical Economy Turning Points earlier than government statistics. ──────────────── 54. This is what Raj refers to as the significance of Stealth Industrial Growth. In an interview in 2026, he said: Previously, the U.S. industrial ISM experienced a long period of stagnation, But FedEx has already seen: B2B activity showing signs of recovery. He calls it: Stealth Industrial Growth. Note: This is an observation from a logistics CEO based on his own network data, It cannot be directly equated to: "U.S. manufacturing has fully recovered." But it is very worth paying attention to. ──────────────── 55. Why might FedEx be a leading indicator? When companies start to replenish inventory: They first order parts. Then transport. Then produce. Then sell. So freight may precede: Financial report revenues, Employment data, GDP. This is why investors have long focused on: FedEx, UPS, Railroads, Trucking companies Freight data. They are: Industrial Economy Pulse. ──────────────── 56. The Middle East case truly reflects not whether "delivery can still be made" It demonstrates: Network Optionality. Raj said: FedEx has a large hub in Dubai and operations in Riyadh. When certain routes cannot use their own planes due to the situation, They will switch to: Commercial carrier partners To deliver the goods. Note here: Riyadh is more accurately called operations, It is not recommended to write as "two large aviation hubs in Dubai and Riyadh." ──────────────── 57. A truly powerful logistics network does not mean it never has problems But rather: It can reroute after problems occur. This is called: Resilience. After the pandemic, global companies began to realize: The cheapest supply chain Is not necessarily the best supply chain. A truly good supply chain requires: Cost + Speed + Optionality + Resilience. ──────────────── 58. This is also why globalization is transitioning from the "efficiency era" to the "resilience era." From 1990 to 2019, the core of global supply chains was: Just in Time. Wherever it is cheapest, Production occurs there. In the future, more and more companies are turning to: Just in Case. Nearshoring. Friendshoring. Dual Sourcing. Multiple Suppliers. Regional Hubs. This means supply chains are becoming: More complex. ──────────────── 59. And the higher the complexity, the more valuable FedEx as an Orchestrator becomes. This is a very important counterintuitive point: If global supply chains become simpler, The value of logistics platforms actually decreases. If: Manufacturing in China, Direct sales to the U.S., Fixed routes, Are easy to manage. But in the future, it may be: Mexico, Vietnam, India, China, Europe Multi-center procurement. The more complex the routes: The more valuable the Intelligence Layer. ──────────────── 60. This is also why Raj often talks about Re-globalization rather than De-globalization. Globalization has not simply disappeared. But is: Re-routing. In the past: China → USA Had a very high proportion. In the future, more will be: Vietnam → USA, India → Europe, Mexico → USA, China → Southeast Asia, China → Mexico. The supply chain map is being redrawn. Logistics companies are at the forefront of this change. ──────────────── 61. "Panda Express" may seem like a fun anecdote, but it actually has significant commercial value. In 2024: Bao Li and Qing Bao flew from Chengdu to Washington. FedEx used a dedicated: Boeing 777F "Panda Express." The flight took about: 19 hours, Approximately 8,250 miles. Raj mentioned in an interview that FedEx has participated in about: 15 similar panda transports. ──────────────── 62. Why is FedEx willing to engage in this seemingly low-revenue business? Because it is: Capability Signaling. If you can safely transport: Giant pandas, Then customers will naturally associate: Medicines, Precision equipment, High-value artworks, Aerospace parts Can also be entrusted to you. Such projects are equivalent to: Brand Proof. ──────────────── 63. Moreover, Raj himself said something very interesting. He did not emphasize: "The panda business is particularly profitable." But rather referred to it as: More of a diplomatic move. In other words: Diplomacy, Brand, Culture Value may be more important than direct freight profits. This is the difference between how a CEO views a project and how an ordinary salesperson views a project. The salesperson asks: How much profit does this order make? The CEO will also ask: What can this do for the overall brand? ──────────────── 64. Raj's own story is actually more worth studying than "an Indian immigrant becoming a CEO." His career path is very rare. Born in: Trivandrum / Kerala. Undergraduate: IIT Chemical Engineering. Master's: Syracuse University Chemical Engineering. MBA: University of Texas at Austin. Joined FedEx in 1991. In 2022: Became the second CEO in FedEx's history. ──────────────── 65. His first job at FedEx was indeed very grassroots. Position: Associate Marketing Analyst. International Marketing. He himself said: It was almost: "The lowest possible level." This detail is very important. ──────────────── 66. The story of "snatching the roommate's interview call" is basically true and very worth telling. In 1991, during the U.S. recession. His roommate received a campus recruitment call from FedEx, But was preparing to leave the U.S. Raj directly took the phone, Asked the other party: My roommate is not going, I'm still here, can you let me send a resume? He then got an interview. More importantly: At that time, he did not have a green card. Other companies had already rejected him for this reason. Sixty-seven, he started the interview without any small talk and directly told the interviewer: “I don't have a green card.” If this is a problem, let's not waste each other's time. The core meaning of the interviewer's response was: First, let's see if you have the ability to do this job, then we can talk about paperwork. In the end, FedEx hired him. What is truly worth learning from this story is not: "Be a little braver." But rather: Don't reject yourself on behalf of others. ──────────────── Sixty-eight, many people's problem is Self-Rejection Job requirements: 10 items. They only have: 7 items. So: They don't apply. Funding might be rejected: They don't approach VCs. Clients might not respond: They don't send emails. Truly outstanding people often just: Ask. Let the other party say No. Don't say No for them. This is a very practical career principle. ──────────────── Sixty-nine, Raj was promoted to CEO in 31 years, and another core reason is cross-functional He was not: Marketing Analyst all the way to: Chief Marketing Officer and then stop. His experience covers: Asia Pacific, Canada, Marketing, International, FedEx Express, Operations, Corporate Management. In the end: CEO. This is called: T-Shaped Executive Development. ──────────────── Seventy, a true CEO cannot just be an expert in one department CFO needs to understand: Finance. CTO: Technology. CMO: Marketing. But a CEO must understand: Interdependencies. If marketing does promotions: Can operations handle it? If we increase packages: Will the margin worsen? If we buy planes: What about ROIC? If we automate: How will employees cope? If we raise prices: Will customers churn? The CEO's job is always: Trade-offs. ──────────────── Seventy-one, Fred Smith chose Raj not because he resembled Fred Smith the most but possibly because: He does not resemble Fred Smith. Fred Smith is: Founder Builder. Building networks. Expanding assets. Inventing overnight delivery. Continuously expanding the empire. The era Raj took over needed: Optimizer. Integrating networks. Increasing profit margins. Reducing capital intensity. Divesting assets. Using AI to enhance efficiency. This is a completely different type of CEO. ──────────────── Seventy-two, this is a very important rule in corporate history Startups in the first phase need: Builder CEO. Once mature, often need: Operator CEO. When further mature, may need: Capital Allocator CEO. The real danger is: The company has entered a mature phase, but the managers still operate as if in the startup phase: Constantly expanding, Constantly acquiring, Constantly increasing assets. As a result: The scale gets larger, ROIC gets worse. ──────────────── Seventy-three, Raj's One FedEx is actually dismantling the old structure left by Fred Smith This is very difficult. Because: Fred Smith is not an ordinary former CEO. He is: FedEx itself. But Raj said that one important promise Fred made to him is: After the handover is complete, Raj is: The Final Decision Maker. This is a very important aspect of an excellent founder's succession. ──────────────── Seventy-four, the hardest thing for a founder is not retirement but rather: Allowing successors to overturn their designs. Many great founders end up destroying their successors: They verbally retire, but in reality continue to: Micromanage, Veto decisions, Bypass the CEO, Control employees. As a result: The new CEO can never truly take power. Fred Smith did this very well. ──────────────── Seventy-five, the story of the "two-minute appointment on the plane" is basically established Fred was going to Los Angeles, invited Raj to sit on the plane. He even joked: "I looked at your calendar, you have nothing going on." After getting on the plane, he directly asked: If he was willing to be the successor in the direction of President/CEO. Raj answered that he was willing. Subsequently, in 2019, he became President & COO, and three years later truly took over as CEO. ──────────────── Seventy-six, after Raj took over, the capital market's evaluation is currently generally positive Barron's listed Raj among the Top CEOs in 2026. Their report pointed out that since the market clearly knew he would become CEO, FedEx's stock price has cumulatively increased by about 102%. So your: "Increase of about 73%" is no longer the latest data. Of course, it must be emphasized: Stock Price ≠ CEO ability's perfect measure. Because it also includes: Economic cycles, Fuel prices, Interest rates, Market valuations, Industry cycles. However, long-term relative performance is still an important feedback from the capital market. ──────────────── Seventy-seven, Raj's CEO time management logic is also worth learning A true large company CEO should not spend every day: Revising PPTs, Monitoring every warehouse, Approving every project. His core responsibilities should be highly focused on: Direction Where to go. People Who is responsible. Capital Where to put the money. Communication Helping hundreds of thousands understand why. ──────────────── Seventy-eight, this is also why "not micromanaging the team" is important FedEx is: A system composed of over 220 countries and regions, Nearly 18 million daily shipments, Thousands of facilities, Hundreds of thousands of employees. The CEO cannot manage operations personally. If a CEO has to personally resolve: Warehouse issues every day, It indicates a failure in organizational structure. What a CEO should truly design is: Decision System. ──────────────── Seventy-nine, therefore, the most important product of a CEO is not the company's product but rather: Organization. What a CEO ultimately creates is: Who has decision-making power? How does data flow? What behaviors are rewarded? Who is responsible when problems arise? Which businesses receive capital? Which businesses are closed? If the company falls apart the day the CEO leaves, Then that is not a great CEO. That is: Human Bottleneck. ──────────────── Eighty, why is the phrase "Change / Extinction" particularly suitable for FedEx? Raj often quotes the change philosophy left by Fred Smith: If you don't like change, then you will definitely hate extinction even more. Now FedEx itself is the best example. Because in the 1970s: FedEx rose based on overnight letters. After the internet appeared: Paper documents plummeted. Then e-commerce emerged. Then Amazon built its own logistics. Then AI. Each generation could potentially kill FedEx's original business model. ──────────────── Eighty-one, the real danger for a company is not competitors but rather: Its own historical success. If a company once succeeded through: Express Network, It easily feels: It should always use the Express Network. If it once succeeded through: Own Aircraft, It feels: All goods should fly on their own. If it once succeeded through: Independent business units, It is unwilling to merge. This is called: Success Trap. ──────────────── Eighty-two, Kodak died here Kodak was not unaware of digital photography. In fact, Kodak itself participated in early digital camera technology. The problem was: The film business was too profitable. So new technology would: Cannibalize Original profits. In the end, others completed the cannibalization for it. So truly excellent CEOs must dare to: Cannibalize Yourself. ──────────────── Eighty-three, what Raj is doing is exactly this Integrating Express and Ground: Moving the old organization. Tricolor: Reducing the absolute priority of some owned aircraft. Freight Spin-off: Dismantling group assets. AI: Redesigning positions. This is essentially a: Institutional Self-Disruption. It is very difficult for mature enterprises to achieve this. ──────────────── Eighty-four, but FedEx's biggest competitive threat still cannot be ignored: Amazon UPS is a traditional competitor. DHL is an international logistics competitor. Amazon is a more dangerous type: Customer → Competitor. Amazon was once a huge logistics customer. Later, it began to build: Planes, Warehouses, Distribution centers, Last-mile networks. Today, it has also started selling logistics capabilities to third-party merchants. So it is evolving from: Internal Logistics to: Logistics Platform. Recent investment research still views Amazon's external logistics expansion as a significant competitive risk for FedEx and UPS. ──────────────── Eighty-five, so how does FedEx compete with Amazon? It's not about competing with: Amazon packages. FedEx's smartest move is precisely to: Go into areas that Amazon finds hard to penetrate. Healthcare. Aerospace. Automotive. B2B. International customs. Time-critical freight. Complex supply chains. These are much harder than: Regular e-commerce cardboard boxes. ──────────────── Eighty-six, this is a classic example of Strategic Escape. If a giant enters your low-end market: Don't always engage in a price war with it. You can move towards: Higher Complexity. Because the higher the complexity: The more customers value: Reliability, Expertise, Integration, Service. Price sensitivity decreases. ──────────────── Eighty-seven, so FedEx's future is not about becoming a "bigger courier company." This is, in my opinion, the biggest insight from the entire interview. What Raj really wants to build is: Physical Infrastructure + Proprietary Data + AI + Orchestration. Four layers stacked together. First layer: Planes, vehicles, facilities. Second layer: Global package and trade data. Third layer: AI prediction and optimization. Fourth layer: Customer supply chain decisions. ──────────────── Eighty-eight, if the fourth layer is achieved, FedEx's customer relationships will undergo a qualitative change. In the past, customers would ask: "Help me send this item." In the future: "Help me design how to ship it." Further in the future: "Help me determine when to ship, from where to ship, and what method to use." Even further: "Directly help me automatically optimize the entire supply chain." This is a transition from: Execution Vendor → Decision Partner → Operating System. ──────────────── Eighty-nine, this logic is especially important for AI entrepreneurs. Many entrepreneurs are currently thinking: "I want to create an AI chatbot." The truly significant AI business opportunities are often not chatbots. But rather: Embedding AI into high-value workflows in the real world. For example: Logistics scheduling, Healthcare supply chains, Manufacturing planning, Financial risk control, Industrial maintenance, Energy management. The reason is: Each decision in these scenarios is related to real money. ──────────────── Ninety, truly valuable AI products should ask three questions: First: What workflow do I control? Second: Do I have proprietary data? Third: Can the decisions made by AI truly change real-world outcomes? If all three answers are Yes, The value is usually far greater than: A simple chat interface. FedEx is a great case in point. ──────────────── Ninety-one, from an investor's perspective, what FedEx should focus on in the future is not "AI news." What it should really focus on is: 1. Operating Margin Can it approach 8%? 2. FCF Can it get close to $6B? 3. ROIC Can it increase to about 11%? 4. Revenue Mix Is the proportion of Healthcare / Aerospace / B2B increasing? 5. International Margin Especially in Europe. 6. Network 2.0 Savings Are cost reductions sustainable? 7. CapEx Intensity Can it stabilize around 4%? 8. Digital Revenue Can Dataworks / orchestration truly monetize? These are the real investment indicators. ──────────────── Ninety-two, the key risks are also very clear. FedEx is not without risks. The biggest ones include: Decline in global trade. Competition from Amazon. Fuel prices. Labor costs. Global recession. Geopolitical conflicts. Failures in network integration execution. AI investments failing to generate revenue. Long-term low returns from European operations. Growth of high-value customers falling short of expectations. So we cannot automatically assign tech company valuations just because of: "AI + logistics." In the end, it still comes down to: Cash Flow. ──────────────── Ninety-three, from a capital allocation perspective, what I like most about this transformation is: Not AI. Not even One FedEx. But rather: Raj has begun to acknowledge that "FedEx does not need to own everything anymore." Use what should be used from partner airlines. Split off Freight where necessary. Merge networks where applicable. Close redundant facilities where needed. Own planes only on the highest value routes. This illustrates: Asset Ownership ≠ Strategic Control. ──────────────── Ninety-four, the best industrial companies in the future may increasingly follow this logic. What truly matters is not: How many assets I own. But rather: How much value flow I control. Apple does not own all factories. Uber does not own all cars. Airbnb does not own hotels. FedEx cannot become completely asset-light, But it can increasingly shift from: Asset Ownership To: Asset Orchestration. This is a very important change. ──────────────── Ninety-five, from the perspective of billionaires and capital allocators, I would learn three things from Raj: First: Never treat the successful organizational structure in a company's history as gospel. What was the most correct structure yesterday May be the biggest burden today. Second: Revenue growth is not the goal; capital returns are. $100B Revenue, If ROIC is poor, It is meaningless. Third: A truly mature CEO must dare to shrink the empire to increase value. Splitting off FedEx Freight is a typical case. ──────────────── Ninety-six, if we refine it further from the entrepreneur's perspective, FedEx's history is actually a very beautiful entrepreneurial formula: Fred Smith created in the first phase: Infrastructure Moat. Fifty years later, it formed: Data Moat. Raj then transforms it into: Intelligence Moat. If successful in the future: It will become: Workflow Moat. These four layers are increasingly difficult for competitors to replace. ──────────────── Ninety-seven, this is also the "moat upgrade diagram" that I think is most worth remembering: First stage: Assets Planes, trucks, warehouses. ↓ Second stage: Network Routes, density, global coverage. ↓ Third stage: Data 2PB of logistics data daily. ↓ Fourth stage: AI Prediction, optimization, automatic scheduling. ↓ Fifth stage: Workflow Entering customer supply chain decisions. ↓ Sixth stage: Platform Becoming global trade infrastructure. This is what Raj truly wants to achieve. ──────────────── Finally, I would condense the entire FedEx case into one sentence: Fred Smith solved the question of "how to move goods faster" with planes, hubs, and a Hub-and-Spoke network; Raj Subramaniam is trying to solve a bigger problem with One FedEx, Network 2.0, digital twins, and AI—how to let global goods know when, how, and along which path they should move. Therefore, FedEx's greatest future opportunity is not: To ship more packages. But rather: To participate in deciding how nearly $20 trillion worth of goods should flow globally. If this step is truly realized, FedEx's identity will shift from: Transportation Company To: Global Supply Chain Intelligence Infrastructure. And this is what investors, entrepreneurs, and AI practitioners should truly learn from this interview.