The Economics of Wars

For most of the twentieth century, the economics of war was a simple question: whose factories could out-produce whose. Tanks, aircraft, and the industrial base behind them decided outcomes, and the cost was counted in steel, fuel, and lives. That war still exists — it is being fought in Ukraine as this letter is written — but it is no longer the only war, or even the most economically consequential one for most nations most of the time. Six distinct weapons have layered on top of the battlefield since, each cheaper to deploy than the last, each harder to see coming, and each landing its bill on a different part of the economy.

Weapon One: The Battlefield — Still Real, Increasingly Rare

Kinetic war remains the most destructive and the least frequently used of the four, because it is the most expensive in every sense — in lives, in international standing, in the sheer industrial base required to sustain it. NATO's 5% GDP defence spending target, discussed in Letter 46 (Security Goes Big), reflects a genuine re-arming already underway, and the private security market alone — $278 billion in 2026, heading toward $394 billion by 2034 — is itself evidence that even peacetime economies are pricing in a higher baseline of physical conflict risk than they did a decade ago. But battlefield war is now the last resort, not the first move, because three cheaper weapons came first.

Weapon Two: Currency Wars — The Oldest Quiet Weapon

Before sanctions became fashionable, nations already had a way to fight without soldiers: deliberately weaken your own currency to make exports cheaper and imports more expensive, forcing trading partners to absorb the competitive disadvantage. The 1930s remain the textbook case — more than 70 countries engaged in competitive devaluation as a response to the Depression, and the resulting "beggar-thy-neighbour" spiral contributed to a 25% decline in global trade volume, a self-inflicted wound with no external enemy at all. The modern version is quieter but no less real: the US-China dispute over the yuan's valuation, running since at least 2010, rests on the same mechanism — allegations that Beijing keeps its currency undervalued to preserve an export advantage, met with the implicit threat of tariffs and countervailing currency actions. Letter 36 (The De-Dollarization Reality) already documented the newest front in this weapon's evolution: central banks buying more than 1,000 tonnes of gold a year for three consecutive years, and parallel settlement infrastructure like mBridge, both quiet hedges against a dollar-denominated financial system that sanctioning nations have learned they can be locked out of instantly.

The honest accounting problem with currency wars is that, unlike a tariff or a frozen reserve account, the bill rarely arrives as a single named number. It shows up as suppressed trade volume, as inflation in the currency-weakening nation's own import bill, and as the slow accumulation of gold and alternative settlement systems by nations hedging against being on the wrong end of the next one. That diffuseness is not a reason to leave this weapon off the list — it is the reason it has been used continuously for a century without ever generating the kind of single "$2.44 trillion" headline number a shooting war produces, and diffuse, headline-resistant costs are exactly the kind this letter series exists to name.

Weapon Three: Sanctions — War Without Soldiers

Sanctions let a nation impose real economic damage on an adversary without a single shot fired, and the past four years have been the largest live demonstration of both their power and their limits. Russia's central bank had roughly $300 billion in reserves frozen at Euroclear following the invasion of Ukraine — an indefinite freeze, with Russia's own central bank now suing for $230 billion in damages, per Letter 22. Russian equity market capitalisation fell to 23% of GDP from 47% pre-2022. And yet the same letter's honest accounting showed sector-level Russian profits — Rushydro, fertiliser exporters — still growing, because sanctions rarely achieve total economic isolation; they achieve expensive, distorted, partially-rerouted trade. The cost lands on both sides: the sanctioned nation loses market access and reserve mobility, and the sanctioning nations lose a trading partner and absorb the second-order effects on energy and commodity prices that follow.

Weapon Four: Trade Wars — Tariffs as Domestic Policy Wearing a Foreign Policy Costume

Letter 40 (State of Trade in a Tariff World) already delivered the honest number: the US effective tariff rate moved from 2.3% to 15.8% in twelve months — the sharpest reversal since Smoot-Hawley in 1930 — with a stated goal of reducing deficits and rebuilding domestic manufacturing. The same letter's honest accounting of the actual outcome: a flat trade deficit, declining factory employment even as tariffs rose, and a $1,500 average household cost increase. This is the weapon that most resembles a war fought against one's own consumers as a side effect of being fought against a trading partner — the bill is real, it is just paid domestically, in higher prices, before it is ever paid by the intended foreign target.

Weapon Five: Supply Chains — The Quiet War Over Who Controls the Chokepoint

This weapon does not touch tariffs or reserves at all. It targets the physical chokepoint a rival economy cannot function without. US export controls on advanced semiconductors and the chip-manufacturing equipment behind them are the clearest current example — restrictive enough that they arguably accelerated China's own semiconductor self-sufficiency drive (YMTC, CXMT, the DeepSeek cost-efficiency story already covered in this letter series) rather than simply denying China the technology. Indonesia's 2020 ban on raw nickel ore exports, covered in Letter 73, is the same weapon used defensively by a resource-rich nation instead of offensively by a technology-rich one: force the value-add to happen inside your own borders or not at all. Letter 97 (The Great Return) put a number on the reshoring response this uncertainty has triggered: 244,000 manufacturing jobs reshored to the US in 2024 alone, and a persistent bottleneck of 2.1 million unfilled manufacturing jobs by 2030 — the cost of rebuilding a supply chain on purpose, rather than the cost of losing one by surprise.

Weapon Six: Cyber Warfare — War With No Uniform, No Border, and No Bill Anyone Agrees On

The newest weapon requires no ships, no tariffs, and no physical chokepoint at all — just code, and it is now the weapon with the least agreed-upon price tag of any on this list, which is itself revealing. Estimates for the narrower category of state-sponsored cyber warfare specifically — as distinct from the far larger universe of ordinary cybercrime — put 2025's global damages at roughly $13.1 billion, a 21% increase on the year before, with 39% of major attacks now confirmed state-sponsored and 76 countries affected by nation-state cyber operations. The United States, China, and Russia together accounted for 61% of all observed cyber warfare activity. Critical infrastructure attacks — energy, water, transport — rose 34% in a single year. Zoom out to cybercrime broadly, including the criminal activity states quietly tolerate or launder through, and the more credible current estimates run to $1.2-1.5 trillion annually — Americans alone lost nearly $21 billion in 2025, and North Korea's state-linked hacking has become a documented funding mechanism for its weapons programme, laundering stolen cryptocurrency around the sanctions this same letter already covered as Weapon Three.

The honest caveat belongs here more than anywhere else on this list: cybercrime cost estimates have a well-documented history of wild disagreement, with older projections of $10.5 trillion annually now considered exaggerated by the researchers who study this seriously, precisely because most of the cost is indirect — lost productivity, reputational damage, downstream disruption — and very hard to verify against a receipt. That does not make the weapon smaller. It makes it the hardest of the six to hold anyone accountable for, which is exactly why state actors increasingly prefer it to the five weapons named above.

The Costs Nations Actually Pay, Side by Side

Laid next to each other, the honest accounting looks like this. Battlefield war costs lives and GDP directly, concentrated and visible, borne overwhelmingly by the combatants. Currency wars cost trade volume and purchasing power, diffusely enough that a century of continuous use has never produced a single headline number — the 1930s episode alone cost 25% of global trade. Sanctions cost the target nation market access and reserve mobility — Russia's ~$300B freeze — while quietly costing the sanctioning nations energy and commodity price volatility they then have to explain to their own voters. Trade wars cost the tariff-imposing nation's own households first — America's $1,500 average household hit — before they cost the intended target anything at all. Supply-chain warfare costs whoever discovers their dependency last: 2.1 million unfilled manufacturing jobs is not a cost imposed by an enemy, it is the bill for decades of assuming a chokepoint would always stay open. Cyber warfare costs whoever gets breached, in a currency — reputational damage, downstream disruption, indirect productivity loss — that resists being priced at all, which is precisely its appeal to the states using it.

Why This Progression Is Not Random

Each weapon emerged because the previous one became too expensive, too visible, or too easily countered. Battlefield war became too costly in a nuclear-armed, economically interlinked world. Currency wars invited direct retaliation the moment a trading partner noticed. Sanctions became partially counterable through exactly the kind of parallel infrastructure Letter 36 (The De-Dollarization Reality) documented — mBridge, local-currency trade settlement, gold accumulation by central banks. Trade wars proved to cost the imposing nation almost as much as the target. Supply-chain control is hard to counter quickly, because building an alternative chip fab, an alternative rare-earth refining base, or an alternative undersea cable network takes years, not a policy announcement. And cyber warfare is hardest of all to counter, because it is hardest of all to even attribute with certainty — a nation can rarely be completely sure who hit it, let alone retaliate proportionately. That escalating difficulty of attribution and response is precisely why the newest weapons are also the quietest, even in years with no declared war at all.