NextFin News - Global military spending has risen for 11 straight years, reaching a record $2.887 trillion in 2025, yet the stocks of the companies that build the weapons are behaving like anything but a defensive haven. When the U.S.-Iran war escalated in March 2026, the iShares U.S. Aerospace & Defense ETF (ITA) fell about 12% over the following eight weeks while the S&P 500 gained 3.5%. Lockheed Martin dropped from roughly $676 to $580. RTX Corp fell from $130 to $115. The uncomfortable truth for investors who bought defence equities as a hedge against a dangerous world: the arms supercycle is real, but the market has already paid for it.
The numbers on the spending side are unambiguous. World military expenditure reached $2,887 billion in 2025, a 2.9% increase in real terms over 2024, according to the Stockholm International Peace Research Institute. That is the 11th consecutive year of growth and the highest level the institute has ever recorded, pushing the global military burden to 2.5% of GDP, the highest since 2009.
Europe is rearming at a pace not seen since the Cold War. European military spending rose 14% in 2025 to $864 billion, while spending in Asia and Oceania climbed 8.1% to $681 billion. Germany's defence budget grew 24% to $114 billion, crossing the 2% of GDP threshold for the first time since 1990. Of the 32 NATO members, 23 now spend at least 2% of GDP on their militaries. The top three spenders — the United States, China and Russia — accounted for a combined $1,480 billion, or 51% of the global total.
Yet equity investors are not being rewarded in kind. ITA's one-year total return of 29.85% as of August 14, 2026 looks respectable in isolation, but the broader Industrials category returned 45.64% over the same period. The S&P 500 is up 13.95% year to date as of August 17. Defence stocks, in other words, are participating in the bull market but not leading it. Over six months the gap is even clearer: ITA returned 9.07% while the S&P 500 returned 11.68%.
The historical record shows this is not a new phenomenon. ITA lost 38.65% in 2008 and 13.57% in 2020 — both years in which government defence spending rose. A portfolio that falls harder than the market during recessions and wars is not defensive by any accepted definition. The sector's best years, including a 48.64% gain in 2025, have come from multiple expansion, not from the steady, low-volatility compounding that investors associate with true defensive holdings.
This is the central tension: defence spending is structural, but defence equity returns have become cyclical. The market has treated the sector as a growth story, re-rating valuations to levels that leave little margin for error.
The Spending Is Real. The Multiple Expansion Has Already Happened.
The first reason defence stocks do not behave defensively is the simplest: investors have already prepaid for the good news. The S&P 500 Aerospace & Defense sub-sector now carries a forward price-to-earnings ratio of 32.5, well above its historical range of 15 to 20, according to Yardeni Research. The forward profit margin sits at just 8.6%.
Compare that with the broader market. The S&P 500's forward P/E was approximately 21.5 times as of late February 2026, above its 10-year average of 18.8 times but far below defence's premium. Individual contractors are no cheaper. Lockheed Martin traded at an enterprise-value-to-sales ratio of 1.82 in early 2026, against a 2003-2023 average of 1.30. RTX Corp sat at 3.32. Boeing, despite its well-documented production troubles, commanded 2.58 times sales versus a historical 1.36. General Dynamics, Northrop Grumman, and L3Harris all trade at EV/S multiples 30% to 100% above their long-term averages.
This is not a market that is pricing in risk. It is a market that is pricing in a decade of uninterrupted growth. When a sector's valuation implies perfection, any stumble becomes a de-rating event rather than a buying opportunity. Lockheed Martin proved the point in April 2026: a flat first-quarter revenue print of $18.0 billion and earnings that missed consensus sent the shares down more than 13% in a week, one of the stock's worst weeks since 2020.
Bank of America analyst Ronald Epstein captured the dynamic precisely. After the first-quarter earnings disappointments, he wrote that expectations for the sector were "skewed too high." The question he raised — whether "peak defence" has been reached — is the right one. Not peak demand. Peak expectations.
The reported numbers were generally ok. But there were company specific execution issues, concerns about margins, cash flow and a perceived lack of short-term positive catalysts.
Epstein's assessment identified the mechanism behind the selloff: expectations had run ahead of what the companies could deliver. He also noted that investors were growing concerned that a defence reconciliation budget might not pass before November's midterm elections, in which case it could be dead in the water.
Defence Equities Are Long-Duration Cyclicals, Not Bond Proxies
The "defensive" label rests on a category error. It assumes that because defence revenues come from government contracts rather than consumer wallets, the stocks should behave like utilities or consumer staples — steady, low-beta, reliable in downturns. That logic confuses the customer with the cash flow.
A defence contractor's revenue may be insulated from a recession, but its equity value is not insulated from the discount rate. These are long-duration assets: their value derives from cash flows stretched across multi-year, sometimes multi-decade programs. When interest rates rise, the present value of those distant cash flows falls, just as it does for any long-duration growth stock. A forward P/E of 32.5 times on an 8.6% margin is a growth-stock valuation, and growth stocks do not act defensively when the cost of capital moves against them.
There is a second, subtler problem: the sector is not purely defence. ITA's three largest holdings are GE Aerospace at 21.92%, RTX Corp at 17.08%, and Boeing at 9.26% — nearly half the fund is exposed to the commercial aerospace cycle, where order books depend on airline profitability, travel demand, and interest rates, not on Pentagon appropriations. When the commercial aerospace cycle turns, or when Boeing's production problems dominate headlines, a "defence" ETF moves with the industrial cycle, not with the security environment. The label on the fund is misleading; the exposures underneath are not.
The budget process adds a third layer of cyclicality. The U.S. fiscal year 2026 begins on October 1, and Congress has not passed full-year appropriations. A continuing resolution is the likeliest outcome, and a government shutdown remains a live risk. The Trump administration's FY2026 request of $892.6 billion in discretionary national defence funding is essentially flat with FY2025's enacted level — and FY2025 itself was funded under a full-year continuing resolution, the first in the Pentagon's history. Flat funding against a 32.5-times multiple is not a recipe for outperformance.
War News Sells Newspapers, Not Stocks
The clearest evidence that defence is not defensive is its behaviour when conflict actually escalates. In March 2026, as the U.S.-Iran war broke out, ITA dropped about 12% over the following eight weeks while the S&P 500 added 3.5%. Lockheed Martin fell from roughly $676 to $580. RTX Corp fell from $130 to $115. The European defence sector fell 11% in March, its biggest monthly loss since the pandemic, as investors worried about energy shocks and budget uncertainty.
The pattern is perverse but consistent: investors buy defence stocks in anticipation of conflict, then sell them when the conflict arrives, because war brings budget uncertainty, political scrutiny, and the risk that earnings-growth expectations get cut. Epstein noted in early April that 2026 earnings-growth expectations for General Dynamics, Lockheed Martin, Northrop Grumman, L3Harris and RTX had already been trimmed to around 12%, down from about 15% at the start of the year. The inverse of his observation is equally true — war is not great for defence stocks either, once the initial headline spike fades.
Northrop Grumman's trading in July 2026 illustrates the volatility that persists even in calmer periods. The stock touched an intraday low of $479.02 on July 21, the day of its second-quarter earnings release, before recovering to $585.87 by August 14 — a swing of more than 20% in under a month. The company beat estimates, earning $7.68 per share versus $6.82 expected, and raised full-year guidance. Even a clean beat produced a violent reversal. This is not the price action of a defensive holding.
The Counter-Thesis: This Time, the Supercycle Is Structural
The bull case deserves a full hearing. Defence spending is not a temporary spike; it is a regime shift. Europe has abandoned its post-Cold War peace dividend. NATO's 2% floor is becoming a norm rather than an aspiration, with 23 of 32 members now meeting it. China has increased military spending for 31 consecutive years. The United States is discussing a defence budget that could approach $1.5 trillion by 2027. These are not cyclical fluctuations. They are structural changes in the global security order, and they will sustain defence revenues for years.
The earnings bear this out. ITA's largest constituents — GE Aerospace, RTX, TransDigm, Howmet Aerospace, and Lockheed Martin — have beaten consensus, raised full-year guidance, and delivered double-digit earnings and cash-flow growth. Northrop Grumman's backlog reached a record $104.7 billion after $20 billion in net awards in the second quarter, up from $95.7 billion at the end of the first quarter. The industry is not struggling to find demand.
The problem is not demand. The problem is price. A structural revenue story does not guarantee a structural equity return when the entry multiple is 32.5 times forward earnings. The supercycle is real; the question is whether it has been pre-paid. History suggests that sectors which re-rate to extreme multiples on a structural narrative tend to deliver mediocre returns for the next several years, even when the narrative proves correct. The 1990s telecommunications buildout and the 2021 semiconductor boom both delivered on their structural promises and still produced poor equity returns for investors who bought at peak multiples.
There is also a positioning argument. After a year in which defence was one of the most-crowded thematic trades, much of the incremental demand has already been expressed. When everyone who wanted to own the rearmament story already owns it, the marginal buyer disappears — and prices stop going up even when the news stays good.
The Second-Order Consequence: The Fear Hedge Has Failed
The deeper failure is not just one of valuation but of function. Investors did not buy defence stocks merely for returns; they bought them as insurance — a fear hedge that would pay out when the world became more dangerous. The March 2026 episode demonstrated that the policy does not pay. When geopolitical risk spiked, the hedge lost money precisely when it was supposed to work.
This has a corrosive second-order effect. Once investors learn that the fear hedge does not hedge, they stop allocating to it as insurance and start judging it purely on valuation and earnings. That shifts the sector's investor base from long-horizon holders who tolerate volatility to momentum traders who exit on the first disappointment. The result is higher volatility, tighter correlation with the broader market, and a lower tolerance for execution misses. The "defensive" label, in other words, is not just wrong — it is self-defeating. The more investors believe it, the faster it stops being true.
Outlook: Three Scenarios for the Sector
The defence sector sits at a crossroads between a structural spending story and a cyclical valuation story. The spending is structural. The valuation is cyclical. And the market is currently pricing the former as if it eliminates the latter.
In the short term, the sector's direction will be set by rates and budget headlines, not by order books. A continuing resolution or a government shutdown in October would likely pressure multiples even as backlogs grow. In the medium term, earnings growth can grow into the valuation — but only if margins expand from their current 8.6% and if the multiple compresses gradually rather than abruptly. In the long term, the structural rearmament of Europe and the Indo-Pacific provides a floor under defence revenues that did not exist a decade ago.
Three scenarios frame the path ahead. The base case is mean reversion: defence spending keeps rising, earnings grow into the multiple, and the sector delivers mid-single-digit annual returns over the next three years — respectable, but below the index. The upside case requires a catalyst the market has not priced: a sustained drop in interest rates that re-inflates long-duration multiples, or a budget settlement that lifts the FY2026 topline materially above the administration's flat request. The downside case is a de-rating: margins disappoint, a shutdown drags on, or peace breaks out in Ukraine or the Middle East, and the forward P/E compresses back toward 20 times. In that scenario, a stock up 29% over the past year can give back half its gains without any deterioration in the underlying business.
For investors, the implication is asymmetrical. The upside from current levels is capped by the multiple the market has already granted. The downside is opened by any disappointment in rates, budgets, or execution. The defensive label is a mirage: these are high-multiple, long-duration cyclicals that happen to sell to governments.
The falsifying signal is straightforward. If the Aerospace & Defense forward P/E re-rates back toward its historical 15-to-20 range while order growth holds — or if the sector underperforms the S&P 500 by more than 10 percentage points over the next 12 months despite rising defence budgets — then the "defensive supercycle equity" thesis is wrong, and defence belongs in the growth bucket where its valuation already places it.
Defence spending may be the one certainty in a dangerous world. Defence stocks, priced for perfection, are the one thing they are not: a safe place to hide.
Explore more exclusive insights at nextfin.ai.
