The Reverse Perestroika thesis draws from Michael Every’s work at Rabobank. Defense and drone landscape data referenced from OptimusDelta’s defense research and VettaFi’s drone industry analysis.

I recently took profits on EWY ~$152 average and escaped the deleveraging/retail flush. Korean retail has piled into the memory trade with the kind of volume and euphoria that tells you a local top was probably forming. I remain mid-term bullish on the AI value chain bottleneck thesis, in particular Japanese optics and HBM, and will look to rebuild positions in SK Hynix and Samsung (expressed in EWY) when the dust settles. But a huge part of capital allocation is knowing when to step away from a crowded table, and the recent Iran developments helped to drive home that point.

This memo is about what comes next, and why I believe drones and UAVs are the single best thematic exposure for the regime shift already underway.

The Reverse Perestroika

Michael Every at Rabobank published a thesis in January that I keep returning to. Just as Gorbachev attempted to restructure the Soviet Union from a centrally planned war economy toward openness and market liberalization, the United States is now executing the inverse (reverse “Perestroika”). A consumption-driven, financialized, globally open economy is being deliberately restructured back toward state guidance, security focus, and a heavy military-industrial model.

For decades the markets and global economy optimized for low-cost, just-in-time supply chains underwritten by globalization, free capital flows, and the assumption that global co-operation would be permanent and order enforced by the world police, the USA. This assumption is now dead. Much of what we observe from Trump’s geopolitics, from Greenland, Iran, Venezuela, tariffs and China follows directly from this understanding. The US is pivoting toward defense readiness, domestic manufacturing, energy independence, and winning the AI arms race. The 2026 National Security Strategy made it explicit: economic security IS national security.

We live in a neo-Cold War world between the two future AI super-states. The US and China are the principals and live players. Everyone else is either picking a side, hedging, or pretending the old order still holds. The catalysts are compounding: US-China rivalry escalating, Taiwan as the strategic chokepoint for chip fabrication, the Middle East fracturing further, and supply-chain fragility that COVID exposed but never resolved.

The game theoretical optimal decision for governments will become the same. Resilience over efficiency. Redundancy over optimization. Hard power over financial engineering. When those same governments are drowning in the late-stage debt dynamics I wrote about in Māyā — sovereign debt at historically unsustainable levels, central banks trapped between crashing asset markets and debasing the currency, then war becomes the oldest mechanism for redirecting popular anger outward and keeping the veil intact a little longer.

My base assumptions for the next few years are as follows: defense budgets structurally increase across NATO, the Pacific, and the Gulf. Industrial policy maintains regardless of who holds office. The AI capex cycle becomes a matter of national security rather than shareholder returns. And I think the money is already moving. Trump is proposing a $500 billion increase in the Pentagon budget for 2027. NATO allies are being pushed toward 5% of GDP on defense by 2035. Germany alone has hit €108.2 billion for 2026, nearly 2.8% of GDP and climbing toward 3.5% by 2029. And the recent developments between the US Department of War and Anthropic only prove that AI itself is now embedded into the fates of the global world order at hand.

Why Drones

If Reverse Perestroika is the reallocation of capital from consumption and financial engineering back toward hard power and industrial capacity, then UAVs are where I believe is one of the highest convexity trades that are available to take at the moment.

The numbers back it up. A $700 drone knocked out a multimillion-dollar tank in Ukraine. Iran launched waves of cheap drones to exhaust $2 million Patriot interceptors across the Gulf this week, bleeding US inventories dry through pure arithmetic. Debris from kinetic intercepts set fire to the facade of the Burj Al Arab and killed someone at Zayed International Airport. Russia is churning out 100,000 long-range strike drones per year while the combined Western output is a fraction of that.

“We can’t do business the same way we have in the past. We cannot afford to shoot down cheap drones with $2 million missiles.”— Pete Hegseth, US Secretary of War, Unleashing U.S. Military Drone Dominance (July 2025)
“Firing a $2.1 million Patriot missile to intercept a $20,000 drone is not a sustainable defense; it is a recipe for strategic exhaustion.”— OptimusDelta, Drone Dominance and the Attritable Math War

The Western defense posture — built over decades around a handful of multi-billion-dollar weapons systems — has become vulnerable to drone swarms. The 21st drone gets through because the magazine is empty.

This is why the US Department of War launched the $1.1 billion Drone Dominance Program and why the Gauntlet evaluation at Fort Moore began last month — hardware, not PowerPoint roadmaps, in the hands of warfighters who write the end-user requirements. This is why Germany’s 2026 defense budget hit €108.2 billion with over €1 billion ring-fenced specifically for counter-UAS systems. Berlin even passed a law, the BwPBBG, mandating speed over perfection in procurement. This is why Poland signed a $3.8 billion contract for an integrated drone wall along NATO’s eastern flank, the leading edge of what is estimated to be a $60 billion+ total market for European layered air defense.

The global commercial drone market is projected to reach $57.8 billion by 2030. The military drone market, $88 billion at a 14% CAGR. These are not just speculative projections but are indicators of procurement decisions already made.

But the numbers alone miss the deeper point. Drones exist in the intersection of every critical technology stack in the current regime: AI for perception and autonomy, advanced optics and thermal sensors, semiconductors, secure communications, energy density, and precision manufacturing. They are not just a defense product anymore but are a politically durable demand signal inside a rising defense budget. Unlike legacy prime weapon systems locked into decade-long procurement cycles, UAV ecosystems evolve at software speed. They iterate in months. They scale the way a Cold War defined by technological acceleration demands.

This is the Atoms vs Bits thesis coming to life. The attack side: mass-replaceable strike drones built cheap and shipped fast, hundreds of thousands of units, designed to be lost. The defense side: directed energy systems with infinite magazines and near-zero cost per shot. Ten cents of electricity to vaporize what a $2 million Patriot used to intercept. And the infrastructure layer underneath both: sovereign industrial platforms that localize production inside each market because the old US-exporter model is dead and every country now demands its drones be built on domestic soil.

All three layers are being funded now, at scale, across multiple continents simultaneously.

Positioning

I have been building a small position across the drone/UAV/counter-UAS value chain over the past few days.

The defense allocation gap I flagged in memo-001 is now the priority. I still have my same portion of cash, gold and copper as a core part of my portfolio, and I still believe in the hypergamblification and privacy thesis. But the Reverse Perestroika thesis makes a lot of sense to me, and I believe drones are the best expression and my skin-in-the-game for aligning with this idea.

More on specific names and the deeper supply chain mapping in a future memo... perhaps.

And as always, not financial advice :)

-fin-