Citrini’s 26 Trades for 2026 | Citrini on BS Jobs, AI Materials, Advanced Packaging, & More

Episode Overview

• This episode provides a forward-looking analysis of investment themes for 2026, with a strong focus on the second-order effects of artificial intelligence on the economy, labor markets, and specific industries.
• A core investment framework called "Post-Traumatic Supply Disorder" (PTSD) is introduced to explain how capital discipline in cyclical industries like energy and technology is creating supply constraints and investment opportunities.
• The discussion breaks down several actionable trade ideas, including a bullish case for natural gas, a long/short strategy for on-device AI chips, and event-driven plays based on the World Cup and tax refund cycles.
• The conversation also explores the broader societal impacts of AI, such as the potential for rising unemployment alongside a strong stock market, and analyzes the full lifecycle and long-term value of AI hardware.

Key Concepts

• Thematic Investing: An approach where investment returns are driven by a company's exposure to broad, overarching themes like AI, energy transition, and consumer behavior.
• AI's Economic Impact: The potential for AI to create a disconnect where rising unemployment coexists with a rising stock market, driven by massive corporate productivity gains, with historical parallels to the post-WWII economy.
• Post-Traumatic Supply Disorder (PTSD): An investment framework describing how companies in capital-intensive, cyclical industries (e.g., energy, memory chips, power producers) exhibit extreme capital discipline due to trauma from past boom-bust cycles, leading to supply squeezes when new demand emerges.
• Natural Gas "Twin Demand Shocks": A bullish thesis for natural gas based on two massive, concurrent new sources of demand: the power requirements for AI data centers and the construction of LNG export terminals.
• AI Chip "Cascade" Lifecycle: The theory that AI chips retain value over a long period. The newest chips are used for intensive model training, and as they become outdated, they "cascade" down to be repurposed for less demanding but high-volume inference tasks.
• AI Capex as Deflationary: The argument that massive capital expenditure on AI hardware is not just a cost but a long-term deflationary force, as it is spent to replace future, recurring operational expenses like human labor.
• Inference on Device: The thesis that AI processing will increasingly move from the cloud to personal devices ("the edge") to reduce latency and cost. The primary bottleneck for this shift is identified as memory (RAM), not processing power.
• Event-Driven Trades: A strategy that focuses on predictable, calendar-based events. Examples for 2026 include plays based on the North American World Cup's impact on hotels and the consumer spending surge following a projected record tax refund season.
• The "Bread and Circuses" Trade: An investment idea based on the expectation that increased tax refunds will fuel a surge in spending on deferred purchases, particularly consumer durables (like mattresses) and healthcare.

Quotes

• At 0:04 - "That's why thematic equity research is so valuable." - The host, Jack Farley, explains that investment returns are increasingly influenced by the broad themes a company is exposed to.
• At 1:06 - "How are you thinking about markets in 2026?" - Farley frames the discussion around the guest's forward-looking analysis from the "26 Trades for 2026" report.
• At 23:00 - "Could we see an economy in 2026 where the unemployment rate continues going up, but stocks also continue going up?" - The speaker questions if AI will create a disconnect between the labor market and the stock market.
• At 23:29 - "When you're an investor, you have to separate what you want to happen versus what's in front of your face and likely going to happen." - Emphasizing the need for objective analysis, even when dealing with personally or politically charged topics.
• At 24:30 - "I think deep down everyone knows that this is going to happen. Nobody that's using AI with any regularity doesn't witness the improvements, doesn't notice that it's doing more of the things that they would be doing themselves." - Arguing that the rapid and observable improvements in AI make its widespread impact on labor inevitable.
• At 51:40 - "it's going to start having a competition with this other huge mega trend in infrastructure construction, which is LNG export terminals." - Highlighting the two massive, concurrent sources of new demand that will compete for natural gas supply.
• At 52:20 - "the excuse of the Permian will just bring on capacity forever... I think that's definitely how it's priced." - Characterizing the market's current assumption that natural gas supply is effectively unlimited, which he believes is a mispricing opportunity.
• At 58:41 - "...this thing you call post-traumatic supply disorder, PTSD, where these cyclical markets have suffered for a long time of building up inventory... only to lead to a cyclical crash, that they are avoiding investing in the capital expenditure." - Introducing and defining the central investment framework for cyclical industries showing new capital discipline.
• At 1:01:31 - "They listened to a bullish forecast, they built a new factory... demand fell off a cliff and now they got absolutely screwed by doing that. And now they're reticent." - Describing the "trauma" that cyclical companies experienced in past cycles, which now informs their cautious approach to increasing supply.
• At 80:43 - "The bears are essentially saying these chips will be obsolete in 18 months when the next Blackwell generation arrives and then they're capitalizing assets that will soon be worthless." - The speaker is summarizing the core bear argument against the hyperscalers' massive spending on AI hardware, focusing on rapid depreciation.
• At 81:45 - "And then the new chip comes out and that chip moves to like inference... because inference is much less demanding." - The speaker explains the "cascade" theory, where chips transition from intensive training of new AI models to the more common task of running existing models (inference).
• At 82:35 - "That capex is deflationary. When you look at high expenses in isolation, I get that theory, but I would argue there is a certain degree that capex is replacing future operating expenses." - This quote frames the massive AI investment not just as a cost, but as a long-term strategy to reduce future labor and operational costs.
• At 94:48 - "The pretty elegant trade you suggested there is going long a lot of the custom ASIC players... and actually short the companies that are the net buyers of memory." - The speaker outlines a long/short trade idea based on the thesis that AI inference will move to devices, benefiting chip enablers and hurting device makers who will face rising memory costs.
• At 97:54 - "Because the bottleneck for running AI on your phone isn't the [processor], it's the RAM." - Identifying Random Access Memory (RAM) as the key constraint and cost driver for enabling powerful AI on mobile devices.
• At 99:50 - "STRD, the most junior of the preferred structure, has perhaps one of the most asymmetric downsides of any financial product we have seen. While (misleadingly) marketed as 'high yielding fixed income,' these instruments are not bonds." - Highlighting a specific trade idea involving a mispriced MicroStrategy preferred security that carries significant hidden risk.
• At 107:50 - "One of the other areas that we try to focus on for the year ahead outlook is less so what do we think is going to happen and much more what do we know is going to happen and how do we trade it." - Explaining the philosophy behind the "26 Trades for 2026," which focuses on predictable, calendar-based events.
• At 109:19 - "It's mostly consumer durables and then some in deferred medical or healthcare." - Explaining the primary spending categories for households after receiving a tax refund, which forms the basis of the "Bread and Circuses" trade idea.
• At 110:10 - "You kind of get the added optionality of maybe he actually does this to try to lock up the house for the midterms, which would flow through to the same exact kinds of things: these deferred consumer durables." - Connecting the political incentive for a potential "tariff refund" directly to the same consumer segments that would benefit from a tax refund.
• At 110:53 - "A lot of the critiques of the newsletter business... 'Oh, you throw out 200 ideas and some of them stick and then you say, oh, pound the table, this worked well.' This Citrinex is accountable." - The host positions the Citrinex platform as a transparent tool that holds research accountable by tracking the performance of all proposed themes.
• At 111:15 - "It's a, in my opinion, a really useful tool. I use it when I'm making decisions. I think it's a great, centralized place to look at all of our themes, our macro trades." - The guest explains how he personally uses the Citrinex platform to monitor investment themes.

Takeaways

1. Practice objective analysis by separating personal desires for societal outcomes, like low unemployment, from the market realities being driven by technological disruption.
2. Prepare for a potential market environment where AI-driven productivity fuels corporate profits and stock prices, even if the broader labor market weakens.
3. Seek investment opportunities in capital-intensive sectors where "Post-Traumatic Supply Disorder" has led to severe underinvestment, creating the potential for supply shortages and sustained high prices.
4. Apply the PTSD framework not just to energy, but to other cyclical sectors like memory chips and independent power producers that are showing similar capital discipline.
5. View large-scale AI capital expenditure as a deflationary force that substitutes one-time hardware costs for ongoing, and often rising, operational costs like labor.
6. Understand that older AI chips are not worthless; they are repurposed for less demanding inference tasks, extending their useful economic life well beyond the cutting edge.
7. Anticipate the shift of AI processing to "the edge" (personal devices) and identify the key bottlenecks, such as memory (RAM), to find investment opportunities.
8. Structure trades around the "Inference on Device" trend by favoring component suppliers and chip designers over the device assemblers who will bear the rising memory costs.
9. Build a portion of your portfolio around high-probability, event-driven trades that are based on predictable events like major sporting tournaments or changes in tax law.
10. Monitor macroeconomic indicators, like tax refund amounts, to anticipate waves of consumer spending on deferred big-ticket items.
11. Always consider the "optionality" in a trade, such as how a political event like a "tariff refund" could amplify an existing thesis based on consumer liquidity.
12. Use transparent, real-time tracking tools to hold investment ideas accountable and measure the performance of thematic baskets against the market.