Being Exponential | Luke Lango

🔊 New Being Exponential podcast - US seeks Iran ceasefire - Swing trade short timeframes & embrace volatile chop - Buy AI infrastructure dips $AVGO $MRVL - Private credit blowups threaten consumer spending & $AMZN - Hedge PE risks with physical gold and silver - AI deeply disrupts legacy software $IGV $ADBE $CRM Link below.

$NVDA v $XOM. Liberation Day to now. Same result, different path.

Young people can't afford a home, can't find love, aren't starting families, are straddled with tons of student debt, are paying an arm and a leg for rent, aren't saving money, and now can't find jobs. As a society, we need to address this. We need to change. Because if we don't... the next 10 years won't be pretty.

Well, would ya look at that? Maybe the biggest winner of the Iran War is Russia, who is earning the most from its oil exports since just after its troops marched into Ukraine in 2022.

So the 82nd Airborne Division was just ordered to deploy to the Middle East. But we're in the middle of peace talks. While energy infrastructure all across the Gulf gets bombed. No wonder the $SPX $NDX $DJI $RTY can't sustain a rally.

I'm bullish on AI stocks because there is a lot of growth momentum in the AI industry to support more spending, revenue growth acceleration, and higher earnings across the AI infrastructure complex $SMH But even if the Iran War ends next week, I'm still worried about the rest of the economy. The March S&P Global US Flash PMI today provided the first look at the impact of the Iran War on the real U.S. economy, and it wasn't pretty. It showed that composite business output across the manufacturing and services sectors dropped to an 11-month low while selling price inflation hit its highest level in nearly 4 years. The data is broadly consistent with ~1% GDP growth and ~4% inflation. That's not a pretty combination...

New research urges everyone to use AI with their brain, not instead of it. A new paper, from University of Toronto, warns that "frictionless" AI -- tools that instantly produce high-quality answers with zero effort -- may be bypassing the cognitive struggle that helps us learn, build skills, and find meaning in work. Similar to how gyms have replaced "necessary" exercise (hunting, farming, etc.) for many people, it's important to continue exercising your brain, even if you don't "have to." Here's the key bit from @IEEEorg' write-up (read the full article if you have time, or just ask AI to summarize it if you don't 😉): > Past technologies have mostly focused on reducing physical effort. We don’t have to go down to the lake to wash our laundry anymore. [Past technologies] took away the mundane tasks that weren’t driving our learning and growth, they were just adding unneeded obstacles and taking away time from more important tasks. > But AI is taking away effort from creative and cognitive processes that drive meaning, motivation, and learning. That’s a key difference, because it’s not taking away friction from tasks that don’t serve us. It’s taking away friction from experiences that are really important and integral to our development. Just something to keep in mind. 🔗 spectrum.ieee.org/frictionless-a…

AI is raising the floor for standards of living in real time. Kansas just committed $3.9 million to AI and smart transportation projects. AI-driven security monitoring for public transit. Intelligent intersection management. Drone delivery systems for rural healthcare. Part of a $10 BILLION infrastructure program. 12 projects. All targeting safety, mobility, and access in underserved areas. A few years ago, rural healthcare access meant driving hours to a clinic. Now it means a drone lands in your backyard with medication. AI makes things possible that weren't before. Better emergency response, safer traffic intersections, and healthcare where there was none. The standard of living at the top is rising. But it's rising at the bottom, too. Will it rise faster at the bottom than the top? Unclear. But if AI can close that gap even a little, it'll be one of the most consequential technologies in history.

The space data center race is heating up. Blue Origin filed plans for a 51,000-satellite constellation called Project Sunrise. $TSLA/SpaceX is an obvious competitor here. So is $GOOGL with Suncatcher. And $NVDA-backed Starcloud. The economics don't favor orbital compute -- yet -- but there are real benefits to putting chips in space: - 24/7 compute powered by abundant solar energy (for sun-synchronous orbits). - No adjusting output for weather, grid constraints, or consumer demand fluctuations that could slow or pause compute. - Most adversaries can't attack orbital data centers and won't be able to for a long time. Those who can (China, Russia) face serious escalation risk -- think MAD-lite for commercial space assets. - Lower latency for space-based applications. Data doesn't need to bounce from space to Earth and back. - Better data sovereignty and resilience for the same reason. - Bypasses terrestrial NIMBY issues. Big hurdles remain. Cooling chips in space is harder than it sounds. Vacuum doesn't conduct heat away from chips, making heat transfer more difficult. Radiation randomly flips bits. Useful orbital real estate isn't infinite. But solving these problems doesn't just unlock economically viable space data centers. The technologies developed here will matter for deep-space exploration and extraterrestrial habitation down the line. Every billionaire throwing money at space data centers is accelerating humanity's path off-planet. And investors should be mapping the crowd-in effects and infrastructure plays that benefit from this. Ad astra per aspera.

The $SPX briefly dropped below its 200-day moving average last week but is commandingly retaking it today. If this rebound continues, that would be a very bullish indicator for the next 12 months. Going back to 1950, the historical analog for the S&P 500 briefly dropping below its 200-day and then retaking it within days is noisy. Sometimes, the bounce led to a monster recovery rally. Sometimes, the bounce failed and led to a deeper bear market (like 2007). But the differentiating factor oftentimes was what the S&P 500 did over the subsequent 3 weeks. When the market retook its 200-day and then continued the rally over the next 3 weeks and didn't look back, that almost always led to a really strong 12-month rally for stocks. But when the market retook its 200-day and then immediately failed and fell back below its 200-day within the next 3 weeks, that almost always led to underperformance over the next 12 months. So the next 2-3 weeks will be crucial. If we can sustain this rally in a V-shaped manner and stay above the 200-day during that time, this is likely the start of a much bigger stock market rally into early 2027. And I think that's exactly what will happen.

This is the TACO. Here's what happens next... Iran will eat the taco. Quasi-diplomatic resolution likely within 2 weeks, which will probably include cessation of all hostile activities. Oil retreats to just above pre-war levels and stabilizes around $75. We get one, maybe two, bad CPI reports, then it washes through. Fed looks through that. Warsh leads us to 2-3 cuts this year. Economy restrengthens. Earnings pick back up. Estimates rise. Stocks blast to new highs. Maybe that's a bit Panglossian, but I've worked through all the possible outcomes here (including the really bad ones) and this feels like the most likely path forward to me. Color me bullish.

The $MU numbers are fantastic for a number of reasons but this is my biggest takeaway: Based on the results, it is clear that memory is no longer a commodity input that gets priced at marginal cost in the semiconductor stack. It is a strategic bottleneck in the AI stack that commands a premium because the demand for intelligence is growing faster than the capacity to produce the memory that intelligence requires. And that means the typical cyclical risk here — memory suppliers overbuild, supply gets commoditized, margins crash, profits erode — may not rear its ugly head. Demand and value and margins will likely all stay high for a lot, longer. And that means memory stocks MU $SNDK $STX $WDC deserve growth stock multiples of 20X+, not cyclical stock multiples of sub-10X.

NEW: SEC APPROVES NASDAQ'S RULE CHANGE TO ALLOW TRADING OF TOKENIZED SECURITIES (VIA DTC PILOT) ON ITS EXCHANGE, EFFECTIVE DURING THE PILOT PROGRAM SOURCE: sec.gov/files/rules/sr…

Drone stocks are ripping. $SWMR IPO'd and rallied 700%. That level of first-day demand signals something bigger than typical new-issue enthusiasm. Contract and partnership activity is accelerating across the sector. In just the last few weeks (and this isn't a complete list)... - $KTOS is working with Airbus to bolster Germany's + Europe's autonomous defense capabilities. - $DPRO is collaborating with the Canadian Armed Forces. - $SPAI just completed field tests with the U.S. Army. - $UAVS invested in an Israeli loitering munitions company. - $ONDS is on an acquisition spree, grabbing "unmanned aerial systems and long-range autonomous platforms" Rotron Aerospace earlier this week - $RCAT reports earnings later today. Meanwhile, the geopolitical tailwinds keep building. Ukraine is sending drone experts to help the U.S. in Iran, as Russia shares drone intel with Iran. Ukraine and the UK are signing a drone deal. Similar agreements with the U.S. and Gulf States are reportedly in the works. Governments are racing to lock in drone capacity, and the companies supplying that capacity are getting locked into multi-year revenue streams. The drone arms race is accelerating.

🚀Announcing a strategic partnership with NVIDIA to co-develop frontier open-source AI models, combining Mistral AI’s frontier model architecture and full-stack AI offering with NVIDIA’s leading compute infrastructure and development tools.

“Nebius will take care of you,” — Jensen Huang, @NVIDIA’s Founder and CEO, in conversation with our CRO Marc Boroditsky at the Nebius booth here at #NVIDIAGTC. We will.

In past tech waves, hardware bottlenecks forced software to optimize. Or software bottlenecks forced hardware to carry the load. The AI wave is different. Hardware and software are advancing in lockstep. AI software improvements accelerate AI chip design. Better chips accelerate software development. The cycle compounds. That's why AI is improving so fast. $MU just started volume production of HBM4 36GB for NVIDIA Vera Rubin. 2.3x the bandwidth of HBM3E. 20% better power efficiency. They're also shipping the industry's first PCIe Gen6 SSD -- 2x the read performance of Gen5 at 100% higher performance per watt, optimized for agentic AI workloads. Micron's doing its part.

$TDY started full production of radiation-hardened memory chips built for AI-enabled satellites. Two mega-trends we're very bullish on -- space infrastructure and AI -- are gradually merging. There are innumerable ways to invest in one or the other. But far fewer great ways to invest in the intersection. TDY is one such option. The chips enable AI workloads in orbit -- inference, sensor fusion, autonomous operations -- while surviving radiation levels that would fry your laptop. They're also building NASA-grade versions for long-duration missions. Teledyne is quietly helping build the hardware backbone for intelligent satellites.

Goldman Sachs now sees AI investment shifting toward companies that own and operate large-scale data center infrastructure. We've been saying this for months. The numbers back it up. AI workloads will hit 30% of data center capacity in the next two years. Global data center power demand will rise 175% by 2030 vs. 2023 levels. That's equivalent to adding another top-10 power-consuming country to the grid. Building that infrastructure takes years. Land, power contracts, grid connections, cooling systems. You can't spin it up overnight. The market is finally waking up to what we already knew: in a messy market, the winners aren't the storytellers. They're the suppliers with bricks, mortar, and power already locked in.

$NVDA launched the Vera CPU, purpose-built for agentic AI and reinforcement learning. Claims 2x efficiency and 50% faster performance vs. traditional CPUs for these workloads. Shipping H2 2026 via $DELL, $HPE, $SMCI, etc. The infrastructure arms race continues.

Lots of bearish sentiment running around Wall Street these days... and, to be sure, I'm not particularly enthused about the current environment... but if we're arguing the basis for "selling everything" is the war = oil price spike = death of stocks... the historical data argues directly against that... Short-term massive oil price spikes = ALWAYS bullish for equities since 1985. Oil has spiked >40% over the last 20 trading sessions. There have only been six similarly large short-term oil price spikes since 1985 wherein oil rose >40% in 20 days: -August 1986 (OPEC production cuts) -August 1990 (Gulf War initial spike) -September 1990 (Gulf War second spike) -March 2009 (GFC recovery) -March 2016 (Oil recovery) -May 2020 (Covid rebound) All six times, the S&P 500 was higher a year later, with a jaw-dropping average forward 12-month return of... wait for it... 31%. Counterintuitive, yes. But markets tend to be counterintuitive.