The Russian problem is this: The West is bankrolling, arming, and providing tactical intelligence to strike the Russian homeland. This is eroding morale at home, even more than the slow progress in Ukraine. Meanwhile, Russian is not able to impose comparable costs on the West with hybrid warfare.

Moscow understands that a window of opportunity has opened up because Russian defense production has ramped up much faster than Western defense production. They have a window in which they have the West outgunned, particularly when it comes to a missile war.

I wanted to document the evidence for my claim that the Russians have the West outgunned because their missile production has been ramped up well beyond the West’s current production.

Let me preface this documentation with a couple of observations about where we are and how we got here. The Russians have partly reconstituted the great defense production capacity of the Soviet Union. Consistent with the competitive learning theory of war, they have also made great strides in electronic warfare, missile guidance, and above all, in defense production.

Meanwhile, whether because of the proliferation of veto points or the general decay of Western industry, the rate of production of munitions in the West has fallen behind. One strategic error by the West was to rely almost exclusively on interceptors for defense against the missile weapon. Iran’s precision-mass innovation has destroyed the theory of the case for an interceptor-centric defense. The cost-exchange ratio for one or more $4mm Patriot interceptors used against a missile that costs $1mm may be tolerable. But they are simply prohibitive against one-way attack drones that cost tens of thousands of dollars.

The tables that follow were sourced from reliable mainstream sources. We systematically searched, scraped and compiled quantitative estimates of production rates of precision weapons by Russia and the Western powers. The sources with direct links to the estimates are presented at the end of this note.

The objective is to create a standoff weapons class-specific balance of production rates. This is a decisive factor in a missile war. Put simply, if Russia can fire hundreds of missiles and thousands of drones at Europe and the West can only fire back dozens of missiles, then the balance of attrition will decisively favor Russia. What the production rates cannot tell us how a war would play out. That would require a full military campaign analysis. But what it can tell us is who has whom outgunned ex ante.

We start with theater ballistic missiles of the sort that are devastating Kyiv on a weekly basis. Russia is producing 75 theater missiles per month. The US expected to take delivery of 54 PrSM missiles in 2025, for a monthly rate of 4.5. This yields a factor of 16.7 in Russia’s favor. The US is producing about 62 Patriot and THAAD interceptors per month. The ratio of Russian theater ballistic missile production and US interceptor production is 1.21 in Russia’s favor.

Both cruise missiles and loitering munitions or one-way attack drones, are air-breathing weapons that are meant to be expended. What distinguishes the two is that cruise missiles do not use wide wingspans to generate lift. However, the speed gap is now closing with Russia’s new jet-powered Geran drones that feature powerful Chinese engines. We distinguish between these two classes here because they have different use cases in Russian military operations: cruise missiles are being used for counter-force missions while Geran drones are being used for counter-value ones.

Russia is producing 167 long-range cruise missiles per month, while the US is producing 58; a factor of 2.86 in Russia’s favor. But the US is producing 175 air-to-air weapons that can that intercept these cruise missiles. Note, however, that standoff air-to-air weapons like the AMRAAM are also needed for air combat.

The most decisive Russian advantage is in one-way attack drones. Russia is producing 5,500 Geran drones per month. The monthly production rate of the American knockoff of the Shahed is minor and likely in the dozens so far.

While many research teams are working on a counter-drone capability and Ukraine has developed counter-drone drones, no counter-drone weapons are yet being produced in the West at any scale.

The Europeans depend for the most part of US defense production. So, we can see clearly just how badly the West is outgunned at this time. People are desperately trying to ramp up production. But even present plans only envision eventually closing the gap to some degree. The West will continue to lag behind in missile and drone production for some time. The idea that the West can quickly ramp up production is fanciful. The Russians really have the West outgunned.

Recall my scenario where the Russians occupy the Baltics.

Russia is tempted by a fait accompli strategy akin to the Egyptian strategy in the Yom Kippur War. Given the massive advantage of prepared defenses, the most promising strategy is to rapidly occupy the Baltics and immediately move to a defense-in-depth. They can then hold the line at the Suwalki Gap in the ground war. Meanwhile, they can use their upper hand in the missile balance to coerce the West to back down; and impose comparable costs if it does not.

Here is the range map of these Russian theater weapons in that scenario. But regardless of whether Russia chooses to occupy the Baltics, Europe is badly exposed to Russian missile and drone attack and there is not much the West can do about it.

As I said on X, just because the Russians have the upper hand in the missile balance does not mean that they have a theory of victory.

What Moscow must determine is whether an occupation of the Baltics and bringing the war to Europe will be enough to break Nato, to defeat the West, or at least get it to come to terms. It is not obvious that calling Nato's bluff in the Baltics will be sufficient to break Nato. Nato may decide to fight on, if for no other reason than the iron grip of the Atlanticist elites. It may then become an extremely painful and protracted war against the combined might of the West.

However, just because Putin does not have a theory of victory through military escalation does not mean we can rule out military escalation.

Recall the problem that no analyst assessed both that (1) Russia would attack Ukraine and (2) that it would be a protracted war of attrition. The underlying reason is the natural analytical desire for epistemic closure.

Another analytical risk today is simple fear of world war. It is hard to imagine nuclear-armed great powers going to war against each other. The outcomes are too horrific to contemplate.

But just because it would potentially be a catastrophic error for Putin, does not mean he cannot persuade himself to do it. In my assessment, the risk of a world war is real. The West has cornered Russia into a difficult position. Escalation may seem more attractive to Moscow than continuing the slow grind.

Sources

1. Russia · Iskander + Kinzhal · CSIS ASAP · 2026-03-23 · 800–1,000 / year

Russia is now estimated to produce up to 2,000 cruise missiles, 800 to 1,000 9M723 and Kh-47M2 Kinzhal short- and medium-range ballistic missiles, and over 30,000 Geran-2 drones annually

2. US · PrSM (2025 deliveries) · Defense News (Lozano) · 2025-10-13 · 54 planned in 2025

Lockheed delivered the first 26 missiles throughout 2024 and the Army is set to receive another 54 this year, totaling 100 missiles.

Latest Army delivery plan (Oct 2025). Lozano: 400/year only in Q4 FY 2027 into 2028.

3. US · PAC-3 MSE · CSIS (Cancian & Park) · 2026-05-27 · around 650 / year

Current production of the latest variant—PAC-3 Missile Segment Enhancement (MSE)—is around the baseline rate of 650 interceptors per year, with half the deliveries going to the United States and the rest to allies and partners.

4. US · THAAD · CSIS ASAP · 2026-03-23 · 96 / year

The situation is equally dire for THAAD, as Lockheed currently produces 96 interceptors per year, with aims to increase annual production to 400 units.

5. Russia · All cruise · CSIS ASAP · 2026-03-23 · up to 2,000 / year

Russia is now estimated to produce up to 2,000 cruise missiles, 800 to 1,000 9M723 and Kh-47M2 Kinzhal short- and medium-range ballistic missiles, and over 30,000 Geran-2 drones annually

6. US · JASSM · CSIS (Cancian & Park) · 2026-05-27 · nearly 500 / year

The Air Force has procured large quantities of these long-range cruise missiles since the 2000s—on average, nearly 500 a year for the past decade. To deliver these orders, current production appears to be already at the surge rate unlike the other munitions discussed in this article.

Procurement average used as production stand-in because CSIS says the line is already at surge.

7. US · Tomahawk · CSIS (Cancian & Park) · 2026-05-27 · less than 200 / year

While Raytheon’s goal is to have the capacity to make more than 1,000 Tomahawks per year, the recent annual production rate is less than 200 because of small past orders.

8. Europe · IRIS-T · CSIS ASAP · 2026-03-23 · 800–1,000 / year

and pushing IRIS-T production well beyond the current 800 to 1,000 missiles per year

ASAP treats 800–1,000 as current; Hoffmann (Jul 2025) treated that band as the 2026 projection.

9. US · AMRAAM · Missile Matters (Hoffmann) · 2025-07-15 · around 1,200 / year

Raytheon’s total annual AMRAAM production is around 1,200 missiles, but this includes multiple variants, particularly the AIM-120D-3.

10. Russia · Geran · CSIS Beyond Parallel · 2026-03-09 · more than 5,500 / month

This circle of cooperation has allowed Russia to quickly industrialize and scale production systems at Yelabuga ahead of schedule, which is now reportedly producing more than 5,500 units per month to support the war in Ukraine.

11. US · PrSM (400/yr target) · CSIS (Cancian & Park, Last Rounds) · 2026-04-21 · annual target of 400 last year

Lockheed Martin has been scaling up PrSM production, setting an annual target of 400 units last year and announcing further increases under the framework agreement with the Trump administration.

Target, not confirmed output.

12. US · PAC-3 MSE · CSIS (McGinn & Dilts) · 2026-08-25 · 2,000 / year by 2030

Across framework agreements with Lockheed Martin of up to seven years, the DOD anticipates growing the production capacity of PAC-3 MSE interceptors by 233 percent, from roughly 600 to 2,000 interceptors per year by 2030

13. US · THAAD · CSIS (McGinn & Dilts) · 2026-08-25 · 400 / year

THAAD interceptor output capacity by 317 percent, from 96 to 400 units annually

14. US · Tomahawk · CSIS (McGinn & Dilts) · 2026-08-25 · more than 1,000 / year

RTX also plans to raise annual production of the Tomahawk missile to more than 1,000

A second reason for yields to rise is that bond investors are anticipating higher inflation. This is the inflation compensation channel. This is also not the case. Inflation compensation as measured by the difference between inflation-indexed Treasuries (TIPS) and nominal Treasuries is modest and largely stable. Both the 5-year and the 10-year breakeven are at 2.3%. Inflation expectations remain well-anchored.

A third mechanical reason, which must necessarily be true because we have ruled out all other possibilities, is that the bond risk premium is going up. The bond risk premium is also called the duration risk premium or the term premium. This the compensation that investors demand for bearing the risk that their bond portfolio holdings will decline in value. This is not an observable. It is estimated by stripping out the expected path priced into yields. The gold standard is the NYFed’s ACM term premia. Sure enough, ACM term premia have risen.

We can also compute real bond risk premium by further stripping out breakeven inflation at the same maturity. This shows that indeed, it is not inflation expectations that have risen, it is the real bond risk premium.1

Treasuries are still treated as default-remote. The risk being priced is not that the USG will default outright. So what is the risk?

Read more

The US has a policy of deterrence regarding nuclear attack on the US homeland. This policy of deterrence works because the US has a secure second-strike capability in the form of the nuclear triad.1 The undersea leg of the triad is the key.

Stealthy nuclear submarines armed with strategic nuclear missiles (SSBNs) quietly patrol the high seas on the deterrence mission. Even if a nuclear-armed adversary launches a highly successful first-strike that takes out all the ICBMs, all the bombers, and everyone in the National Command Authority, weapons release authority is pre-delegated to SSBN commanders to retaliate with devastating force in the event of a first-strike.2

A single SSBN can fire up to two dozen strategic nuclear missiles with MIRV’d warheads that can deliver a total of 30 megatons. This is enough to basically destroy the enemy’s largest cities. So no one with a fixed address would ever think of attacking the US homeland with nuclear missiles.

The specific nature of strategic weapons and the mutual vulnerability of the powers to nuclear devastation means that we can expect this policy of deterrence to work with high confidence. This is not true of other policies of deterrence. In particular, extended deterrence—where the deterrent threat is meant to dissuade attacks on allies and forward-deployed US forces—is considerably harder. It depends sensitively on the credibility and formidability of the deterrent threat.

If the enemy does not believe our deterrent threat, then he may choose to call our bluff even if we ourselves mean it. So deterrence is in part about convincing the enemy that we really mean it. That when push comes to shove, we will actually deliver the promised threat.

Deterrence may not work even if the deterrent threat is credible, however, if it does not inspire sufficient fear in the adversary’s mind. We may promise sanctions if China attacks Taiwan, for instance, but that is not going to persuade the Chinese to back off if they are otherwise convinced that they need to wrap up the unfinished business of the renegade province.

The US has inherited two major policies of deterrence. The first is our commitment to defend NATO Europe against a Russian attack. The second is our commitment to defend Taiwan against a Chinese attack. Both of these are now past their sell-by date.

The Iran war has revealed the US military to be a paper tiger. Defense analysts perched at think-tanks are not allowed to say we will lose any war. So you’re not going to see it stated by so-called defense experts plainly. What we are seeing instead is projection of the panic onto the half-empty magazine of interceptors and stand-off munitions. That this is mathematically-challenged nonsense is clear from a basic back-of-the-envelope calculation.

The monthly production rate of Russian missiles is roughly the annual production rate of the same in the US, while China’s defense industrial base is an order of magnitude larger than Russia’s. Iran has fired about 1,000 theater missiles3 and maybe 10,000 Shahed drones since the war began. Over the same period, Russia has fired 1,700 theater missiles and over 52,000 Geran drones.4 If the prewar magazine was insufficient to come within shouting distance of defeating Iran, how could it have been sufficient for a war with Russia, much less China?

Read more

Bessent pretended that the yen intervention was to help out an ally in trouble. But the precise weapon he used for the job belied the claim. Instead of buying yen dollars, he chose to use euros. Why? Because the objective wasn’t to stabilize the yen. It was to prevent the Japanese from dumping Treasuries, which would’ve driven up long-term US yields further up.

Read more

FORTY-TWO PERCENT OF THE BOMBS dropped on Germany by American air forces in 1944-1945 landed more than five miles or eight kilometers away from their targets.1 Those which landed within an eight kilometer radius on average landed four kilometers from target. This implies a circular error probable (CEP) of roughly six kilometers. Recall that CEP is the radius of the circle that contains half the hits on average. It is the fundamental measure of weapon accuracy.

To give you a sense of what this number means, here is a simulation of 100 weapons aimed at Times Square with a CEP of six kilometers. We show the five boroughs to scale. As you can see, many fall outside Manhattan. The odds of hitting a missile launcher, a tank, or even a building located at Times Square with a weapon of this accuracy are remote.

During World War II then, the bombing accuracy was so poor that all bombing was area bombing. No point targets could be attacked directly. If one wanted to hit a rail yard, say, one basically had to carpet bomb it. With enough sorties, the rail yard could perhaps be put out of commission until it was repaired by the enemy.

‘A decision from the air’, the holy grail of the proponents of strategic air war, was simply not feasible during World War II. Eventually, the Americans gave up the pretense of precision bombing and joined the British in simple area attacks for ‘moral effect’. They just unloaded their weapons on cities, often missing them by miles. So, strategic bombing during World War II was a pure terror weapon. One that failed comprehensively to achieve any strategic effects whatsoever.

By the Vietnam war, bombing accuracy had improved considerably. Over the course of the 1965-1968 air coercion campaign against North Vietnam called Rolling Thunder, average CEPs declined from 230m to 115m. The following shows a simulation of a bomber with a CEP of 200m.

This is vastly more accurate than World War II. Attacking point targets was becoming possible in a way that was simply not on the table at midcentury. Still not good enough to destroy a tank-sized target, but with a few sorties one could destroy a bridge or a ball-bearing plant located there.

US pilots flew a million missions during Rolling Thunder, dropping three-quarters of a million tons of weapons. It was an abject failure. In the end, Johnson threw up his hands and shut down the whole campaign just to get Hanoi to the table.

Nixon’s Linebacker campaign is sometimes rated by air coercion enthusiasts — and even by Robert Pape — as a success. But as John A. Gentry notes, this is incorrect. Linebacker was also an unambiguous strategic failure.2 Linebacker could only bring Hanoi to the table but not coerce it into making any concessions whatsoever. The US still suffered a humiliating defeat. North Vietnam got US forces to leave South Vietnam without withdrawing its own forces—and then promptly conquered the country without much further ado.

Gentry does not tell us the CEP of US air-to-ground weapons in Linebacker. But we know that laser-guided bombs were being used at this time. These were the first precision-guided weapons to be used in combat. By the Gulf War, the CEP of air-to-ground weapons had fallen to just 3m or 10ft.

Read more

That is about 8,654 battlefield deaths per month. Russia had 1,134,000 men in the armed forces on the eve of the war and a popula…

Read more

Nearly one percent of Israel’s population has been leaving each year. Those abandoning the ship are disproportionately liberal professionals. ‘A survey last year found much higher interest in leaving among secular Jews than among religious or ultra-Orthodox Jews.’

Secular liberal professionals are leaving because the country has been increasingly taken over by orthodox Jews and ultranationalist settlers. What is driving the takeover is the war of numbers. Simply put, liberal Ashkenazi professionals are being outbred.

Read more

All channels through escalation are blocked because Iran enjoys escalation dominance. If you hit their infrastructure, they will hit the gulf infrastructure. If you attack civilians, they will kill your military personnel. If you hit their launch sites, they will hit your bases. A ground incursion will require Iwo Jima levels of casualties that are now beyond US political tolerance. There is therefore no path to a US victory through military escalation. Iran is militarily invulnerable.

The war of the sieges is highly unlikely to produce a decision any time soon. Nor will cat and mouse. In fact, as things stand, the US is rigged to lose as the noose eventually tightens in the oil market.

The US is trying to put the war on the back-burner, talk down Brent, and hope for the best. But time is Iran’s great power ally. The US cannot expect the siege to work before the closure of Hormuz starts to bite the world economy, even if this does not happen before the midterms because of ample reserves. The US has a few months before the crisis builds. Given the strategic picture, it is really hard to see how the US can evade defeat.

In what follows, I will propose a strategy to defeat Iran. Specifically, I will argue that the US can solve the problem of outlasting Iran with a market solution. Let me spell out the logic of this strategy.

Read more

But why has the oil market channel of pain delivery been frustrated? Is it because of the Administration jawboning the oil market as many believe? Or is the oil market really resilient against a partial shutdown of flows from the Gulf?

Read more

This changed after July 11, when they officially declared the Strait closed. No vessel will be permitted to transit, they said.

Read more

Military men have long understood the importance of initiative. Initiative comes with all the advantages of the first mover: the side with the initiative forces the enemy into a reactive mode, sets the ‘agenda’, and can control escalation.

But what is Iran trying to achieve? Not a single analyst that the reporters spoke to offered to explain Iranian objectives and grand-strategy. The closest we get is Vali Nasr saying: ‘They escalate with an end result in mind, which is to control American behaviour and force the US to recalculate its options’. That’s hand-waving. Iran has concrete objectives and a compelling theory of victory.

Read more

A little seen tweet from the commander running the IRGC’s missile command may be the reason Trump taco’d. The commander issued a clear counter-personnel deterrent threat, declaring a policy of killing one US soldier for every Iranian civilian killed.

The rule whose operational nature we had previously proven to the enemy, from this moment on we regard as the definitive and announced equation of the battlefield: for the martyrdom of each proud citizen of Islamic Iran, one American force will be sent to hell. We have prepared free, direct tickets to hell for you.

Commander of the IRGC Aerospace Force, July 22, 2026 on X. Machine translated.

I explained the general logic of the counter-personnel threat before the taco. Returning to the vulnerability matrix I introduced in yesterday’s dispatch, we can see the logic of the direct linkage between Iranian civilian vulnerability and the vulnerability of US military personnel. Iran can deter US attacks on Iranian civilians by threatening to return bodybags to Arlington.

Read more

While Pape is correct about the signal in logistics, there is a stronger signal of the coming escalation. The Israelis are undertaking civil defense measures in preparation for Iranian retaliation. The signal is stronger because civil defense is costly wasting asset. The mayor of Ramat Gan explained that ‘the risk of missile fire from Iran over the weekend has increased’, so he has ordered that the bomb shelters be made open and accessible to the general public ‘despite the cost’.

We can therefore be fairly confident that a massive US assault is brewing. Trump is frustrated with the lack of options to respond to the blockade of Persian Gulf and the Red Sea. He cannot forcibly reopen either. The war of the sieges is unlikely to yield a favorable decision over any horizon he can tolerate. And now Iran has started killing American troops.

Trump has threatened to attack Iranian bridges and power plants again. One would think that the Iranian response to an American counter-value assault would be simply retaliation in kind. But there are good reasons to think that Iranian strategy has evolved in a more aggressive direction.

In order to understand Iranian grand-strategy, it pays to think a bit systematically about vulnerability. Here’s a simple vulnerability matrix to fix ideas. The vulnerability of US civilians to Iranian attack is low. They are located too far for Iranian missiles to reach. Iranian civilians on the other hand, are highly vulnerable to US attack.

Read more

Read more

Read more

That’s a misreading of Iranian objectives. The escalation is not intended to ‘force Washington back to the table’ at all. Rather, the escalation — attacking every ship that disobeys Iranian authority by taking the Oman route — is meant to preserve Iranian leverage. The Iranians believe, quite correctly, and as Vakil herself tells us, that their leverage will erode if the Hormuz weapon does not remain firmly in their grip.

We are told that the Iranian wager is premised on the assessment that Trump does not have the appetite or the stamina for a costly fight: ‘Tehran expects him to want to avoid rising energy prices, American casualties, and another open ended war in the Middle East, particularly with midterm elections approaching.’ This Iranian assessment, she insists, is incorrect: ‘But Tehran may be misreading the source of Trump’s caution. … His reluctance to accept an open-ended war coexists with a deep aversion to humiliation, creating the risk that failed coercion will lead not to restraint but to further escalation.’

But is this quite right? Suppose that we perturb the premise. Let’s say the Iranians think that there is a good chance that Trump will not let Iran keep Hormuz. Would they then be willing to compromise of holding on to Hormuz? Why? What path does Iran have to security against US-Israeli depredations if it does not control Hormuz? Should the Iranians let him neutralize the Hormuz weapon if they think the US will ramp up attacks on Iran? Or even launch a ground invasion as Azizi reckons?

Read more

Read more

Yet, Brent crude is at $70 per barrel. Indeed, even at the peak of the crisis, Brent stayed below the 2008 peak of $144.

By all accounts, the Hormuz dog failed to bark. Why?

Read more

Read more

At the center of world trade sits global manufacturing. The global competition over these high-value trades is the principal driver of innovation and productivity growth in the world economy. Not unreasonably, everyone and their mother wants a bigger piece of the global manufacturing trade. And everyone wants to climb the value ladder. Countries that are able to outcompete others for a larger share of the global manufacturing trade, and especially the high-skill component, win; others lose. This is a zero-sum game. And a brutal one at that. How could it be any other way?

In what follows, we will document some empirical patterns related to who is winning and losing the global manufacturing trade. The data is from the OECD’s TiVa dataset.

Read more