Lasers Are the Hot New Thing in Defense: You Can Win Every Fight and Still Lose on the Books

In the operations of war, where there are a thousand swift chariots, a thousand heavy chariots, and a hundred thousand mail-clad soldiers, with provisions enough to carry them a thousand li… the daily cost will amount to a thousand pieces of gold. Only then can an army of a hundred thousand be raised.

—— Sun Tzu, The Art of War, “Waging War” (Spring and Autumn period; translation adapted by the author)

Sun Tzu doesn’t open by telling you how to win. He opens with the daily burn rate. Twenty-five centuries later, Bloomberg’s Odd Lots spent an episode on laser weapons and landed in the same place: you can win every engagement and still lose on the books.

What the episode covers

Tracy Alloway and Joe Weisenthal talk with Wayne Sanders, senior defense research analyst at Bloomberg Intelligence — West Point 2002, a career in military intelligence and offensive cyber operations, finishing as chief of staff to the Under Secretary of Defense for Research and Engineering. So he comes to the market question having sat on the technology side of it.

There’s a good bit up front: his name is Sanders, his rank was colonel, so he is Colonel Sanders. Tracy says she has sympathy, because her father was in the Air Force and his name is Tom — he never got past major, and always insisted he was under-promoted because people found “Major Tom” funny. Tracy points out it might have run the other way: maybe he was over-promoted because it was funny.

Under the jokes the episode answers a hard question. Movies have had laser guns for sixty years. Why aren’t they on the battlefield? What’s the constraint? And if they do arrive, where does the money go?

The main points

Lasers and microwaves are two different animals, and Sanders has a comparison that stuck with me: a laser is a sniper rifle, a high-powered microwave is an electronic shotgun. The laser concentrates energy into a narrow beam, hits one target at a time, and works by heat — it burns through. The microwave attacks electronics over a wide beam and can degrade many systems at once. Drones hit by it don’t explode; they stop working and fall. He adds a caution useful to anyone reading spec sheets: the GAO notes microwave peak powers up to 100 megawatts, but comparing that number to a laser’s continuous-wave kilowatt rating misleads. A laser scales power; a microwave scales effect. Two different rulers.

The military saw the potential as soon as the laser arrived in the 1960s. It then stalled for six decades, and the obstacle was never the principle — it was SWaP: size, weight and power. Early chemical lasers drew their energy straight from a chemical reaction rather than electricity. They worked, but the facilities were the size of buildings. The Star Wars era produced a version mounted in a Boeing 747, meant to loiter near where you expected a launch and chase the missile down. The joint US-Israel Tactical High Energy Laser did destroy rockets and artillery rounds in flight, and DARPA’s MIRACLE knocked down live short-range rockets. Too big and too logistics-heavy; the program ended in 2006.

What changed is the shift to electrically driven systems. The Army now fields 50-kilowatt lasers on Stryker vehicles, running off generator power the vehicle already carries. Israel’s Iron Beam is around 150 kilowatts and heading to deployment. The Pentagon’s research shop wants a one-megawatt laser by 2030. Fifty kilowatts brings down a drone flying at you; a megawatt is aimed at something travelling four or five thousand miles an hour.

The line that matters most comes mid-episode: you can win every engagement defensively and still lose economically. Firing a four-million-dollar interceptor at a thirty-thousand-dollar drone is a tactical win and a balance-sheet loss. The US optimised for decades against a peer adversary and then found itself shooting at drones made of styrofoam and plywood. What’s happening around Iran now is the slow drawdown of the expensive interceptors.

But lasers don’t replace missiles, and Sanders is firm about it: you cannot give up the exquisite interceptors, because adversaries have hypersonics, and you don’t want to bet on placing a beam precisely on something moving at Mach 5. The real architecture is layered. Israel runs the most complete version: Arrow 2 and Arrow 3 exo-atmospheric for the long shot, David’s Sling closer in, Iron Dome closer still, now Iron Beam added underneath. Each layer gives the commander another option so the most expensive round isn’t the default. Filling the gap between them is a whole class of cheaper interceptors and guns — Coyote, Vampire, APKWS, Skynex — for the targets not worth four million but not yet microwave work.

That splits the market in two. The primes win in the near term because the lines already exist: PAC-3 interceptor output going from 620 a year toward 2,000, THAAD (twelve to fifteen million dollars per interceptor) targeting a fourfold increase. But the middle tier isn’t ordered in those quantities — the demand signal is tens of thousands, and the five firms currently contracted for it are Coaspire, Leidos, Anduril, Zone 5 and Castilian. Sanders puts the disruptors’ window at three to five years, and notes the primes aren’t fighting hard for that middle ground. It’s the innovator’s dilemma playing out in public.

Tracy’s supply chain question lands well. These systems need dysprosium, terbium, gallium, ytterbium, plus the neodymium-iron-boron magnets the military uses in volume. China holds 85 to 90 percent of several of them and added twelve critical minerals to its restricted list for the US last year. The awkward part is dual use: the brushless motor in a hobby drone is the same motor in a military one. Hence the Department of War taking a position in MP Materials to onshore capacity.

One more limit that gets overlooked: weather. A missile doesn’t care about fog or rain. A laser does. Which is why Israel’s operational use matters so much — how far out did it hit, how long did it take to stop the thing. Those numbers only exist if someone is actually shooting at you.

Going further

”The company announced good news — why didn’t the stock move?”

One figure is worth stopping on. Directed energy is a six-to-eight-billion-dollar market today, expected to grow two to three times over the next decade. Cumulative R&D spending to get here is over fifteen billion. The research bill is larger than the market.

My first instinct with that shape is to separate which kind of contract a headline is describing. R&D and procurement are different businesses. R&D is the government paying you to solve a problem that might not be solvable, with a ceiling, and no guarantee anyone buys the answer. Procurement means the thing is designed, the line is running, and units ship every year. “Wins Department of Defense contract” can mean either, and they don’t weigh the same.

Sanders says the transition to procurement is underway — that word, transition, is the whole thing. The way I check is to ask whether the money buys a promise or a batch of goods, and then whether the contract states units per year across how many years. The ones that can state it are structure; the rest is usually noise. In this episode the primes can state it (620 to 2,000, fourfold). The middle-tier five are still at “I need tens of thousands” — that’s an intention, not a production line.

”Something new is coming — should I sell the old thing?”

I’ve made this mistake. A technology sounds like it changes everything, and the reflex is to clear out whatever might get replaced. This episode is a counterexample: lasers are having their moment, and expensive interceptor production is being ramped up, not cut.

Layering is why. As long as one threat class the new tech can’t handle exists — hypersonics, bad weather, anything you need to stop far away — the old layer stays. A new capability’s first role is almost always “one more option,” not “replaces X.” Replacement waits for the old layer’s failure condition to be solved, and solved cheaply.

So I’ve switched to a more concrete question: when does the old thing stop being produced or bought? If you can’t name a date, you’re still in the additive phase. In that phase both sides make money, and make it more comfortably than in the zero-sum phase. It’s the better environment — it just doesn’t sound as exciting as disruption.

“Every call was right and the year still went nowhere”

Tracy closes with a comparison I kept turning over: lasers are the renewable energy of the weapons world — once installed you can keep firing, versus the fossil-fuel economics of a finite missile stock. But plenty of countries and cities decline renewables because the install cost is brutal. Joe adds the sharper version: the scarcity just moves. There’s no silver bullet.

What I thought of wasn’t defense, it was my own statements. A method that’s right every time can still drain you if each use costs too much, and a mediocre-looking method with a low unit cost can run for years. The difference isn’t the hit rate. It’s what you spend per engagement.

That’s what a layered architecture means: not finding the single strongest tool, but matching cost of response to level of threat, and saving the expensive one for when it’s genuinely needed. I only recently started applying this outside markets. Most of what wears people down isn’t bad judgment — it’s using the big weapon on small problems.

Worth a look

• Bloomberg Odd Lots, 2026-09-04, “Why Laser Beams Are the Hottest New Tech in Defense,” with Wayne Sanders (Bloomberg Intelligence)
• Concepts from the episode worth reading up on independently: integrated air and missile defense (IAMD), SWaP constraints, how Iron Beam sits beneath Iron Dome, and the Department of War’s critical-minerals arrangement with MP Materials
• The show’s earlier episode on tungsten with David Fickling, which the hosts bring up themselves: the military had concluded future wars wouldn’t need much of it, and then got wars that did

帶得走的一件事

One idea: winning and lasting are two separate calculations. Track your hit rate and your cost per attempt. Watch only the first and you get the outcome this episode describes — every fight won, the year lost.

One thing I’ve tried that you can do today: pick something you’ve done five or more times this month. Replying to a certain kind of message, a recurring meeting, the same small argument with family, some routine errand on the commute. Write three columns: how many times it happened, what your most expensive resource cost per instance (usually not money — time or patience), and whether a ten-times-cheaper approach would cover eighty percent of the cases.

The third column is empty for most people. Not because no cheap option exists, but because the expensive move is the one closest to hand. Build the cheap layer, keep the expensive move for the remaining twenty percent — that’s what running Iron Dome and Iron Beam together actually means.

Raw Markdown

This article is an educational discussion of investment method. It is not advice to buy or sell any individual security, offers no target prices, and does not analyze any current holding. Investing carries risk; make your own decisions or consult a qualified professional.