Europe started winter 18 points below its own gas target

I turned the heat on last night.

It was the first cold evening of the fall. You know the one. You walk through the house in socks, you feel the floor, and you stop pretending summer will come back. You turn the dial and hear the furnace kick on and you think… fine. Whatever it costs.

Most people don't think about where their heat comes from. I do. And right now, Europe is walking into winter with its gas tanks at 72% full. The EU's own target is 90%.

That gap is 18 percentage points. It is the lowest storage level at the start of winter since the EU began tracking. And it matters far beyond Europe.

ENTSOG, Europe's gas network body, published its Winter Supply Outlook on October 8. EU underground storage held 72% of capacity as of October 1, down from 83% a year earlier and 94% two years before that. In a cold winter, the agency warns of a supply shortfall of up to 15%.

The Dutch TTF benchmark — the main European gas price — traded at €88 per megawatt-hour on October 19. That is up 214% since January. The highest level since December 2022, when Europe was still replacing Russian pipeline gas.

But this time, the cause is different.

When the US-Israel-Iran conflict began on February 28, Iranian strikes damaged roughly 17% of Qatar's liquefied natural gas capacity at the Ras Laffan complex. QatarEnergy's chief executive confirmed the figure. LNG tanker traffic through the Strait of Hormuz — which carried about a fifth of the world's LNG — fell sharply.

Europe depends on LNG for roughly half its gas now. So the injection season — the months when countries refill storage before winter — started weak and stayed weak. By mid-August, storage reached only 60%. By mid-October, 74%. That is the lowest for that date in the 2017–2025 range. About 9 points below 2025. About 14 points below the long-run average.

And as of October 26, injections still exceeded withdrawals. The heating season is here.

I keep coming back to the trajectory. Four years of storage at winter's start: 96%, 94%, 83%, 72%. Each year lower. Each year closer to a problem that cannot be solved in a week.

The obvious risk is a cold winter and not enough gas. But the less obvious risk is what the price does to everything else. Energy costs flow through supply chains. They push up factory input prices. They force central banks to hold rates higher for longer. They eat household budgets.

This week is about that second-order damage. The energy cost that most markets are not measuring.

Energy costs do not stay in one place. They travel. In Europe, they show up in gas bills and factory inputs. In Japan, they showed up in inflation — and inflation forced the Bank of Japan to raise rates. Those rate hikes are now destroying the value of bonds that Japanese insurers hold.

Japan's four largest life insurers — Nippon Life, Dai-ichi Life, Sumitomo Life, and Meiji Yasuda — reported combined unrealized losses of ¥15.13 trillion ($96 billion) on domestic government bonds as of the end of June 2026. That is up 7% from the prior quarter. It is a record. And it is the seventh consecutive quarter of increases.

The math is simple. These firms bought JGBs when the 10-year yield was near zero. The BOJ has since raised its policy rate to 1.25%, the highest since 1995. The 10-year JGB yield has risen to 2.83%. When rates rise, old bonds with lower coupons lose market value. That loss sits on insurer balance sheets as an unrealized number — paper losses that matter in every way except on the income statement.

The insurers can hold these bonds to maturity and never take the loss. That is the plan. But the plan depends on policyholders not surrendering their policies early. It depends on no liquidity pressure forcing a sale. It depends on rates not going higher.

$96 billion in losses. Paper losses. For now.

All four insurers posted increases. All four are in the same position. And all four survived the last Japanese insurance crisis. Between 1997 and 2002, seven of their peers did not.

And here is where the same energy shock shows up on the other side of the world.

China's National Bureau of Statistics reported September producer prices rose 5.0% year over year — the fastest pace since June 2022. Meanwhile, consumer prices rose only 0.9%. Core CPI, stripping out food and energy, held at 1%.

That 4.1-point gap between PPI and CPI is worth paying attention to. It tells you the cost shock is hitting factories, not consumers. Transportation fuels rose 14.9% year over year. Auto prices fell 1.7%. Factories are eating higher input costs because Chinese consumers do not have the spending power to absorb them.

In other words, this is not demand-driven inflation. This is an oil-and-energy cost shock that factories are absorbing as margin compression. Pinpoint Asset Management's Zhiwei Zhang described the CPI reading as evidence that domestic demand remains weak.

The pattern matters globally. China is the world's largest exporter. When its factory input costs rise by 5% and it cannot raise output prices, either Chinese margins compress or Chinese export prices eventually rise. Either way, someone pays.

Same energy shock. Different continent. Same problem: costs rising faster than the economy can absorb them.

Three more things worth keeping track of this week.

The gold-to-copper ratio sat at 632 as of October 19, per spot market data. Gold at $4,145 per ounce, copper at $6.55 per pound. The ratio measures how much copper one ounce of gold can buy. When it rises, investors are choosing safety over industrial growth. At 632, the ratio is deep in territory that has preceded past recessions. It hit 914 earlier this year before falling back. Copper trades on factory output and construction. Gold trades on fear. Right now, fear is winning. We are tracking this.

In the UK, pension funds received collateral calls this month after the 30-year gilt yield reached 6.034% on October 7 — the highest since 1998. British defined-benefit pensions use liability-driven investment strategies that hedge future pension payments with leveraged gilt and swap positions. When gilt prices fall, the LDI funds need more cash to back those positions. Reuters reported that BlackRock told clients it would not allow them to replenish collateral. The situation calmed after the October 28 Budget (30-year gilts settled near 5.7%), but those LDI strategies hedge roughly £880 billion in pension liabilities. The 2022 episode — when similar moves forced the Bank of England to buy £19.3 billion of gilts in two weeks — showed how fast pension margin calls can spread. We'll see.

Volkswagen plans to close four German factories and cut up to 100,000 jobs globally, according to leaked restructuring plans reported by Manager Magazin in June. The company has already signed departure agreements covering more than 28,000 workers at VW, Audi, Porsche, and its software unit CARIAD. The plants at Emden, Hanover, Zwickau, and Audi's Neckarsulm facility are on the list. Higher energy costs, weaker Chinese demand for German cars, and the shift to electric vehicles are all part of the problem. Europe's largest automaker is restructuring at the same time Europe's energy costs are rising again. We'll see.

I want to go back to Tokyo. Not today's Tokyo. 1997.

In April of that year, Nissan Mutual Life Insurance Company went bankrupt. It was the first Japanese life insurer to fail since the end of World War II. Fifty years of stability, gone in a single filing.

The cause was a promise. During Japan's bubble years of the 1980s, life insurers sold policies with guaranteed returns of 5.5% or more. Policyholders paid in, and the insurers promised to pay out at those high rates for decades. The policies were long-dated. The guarantees were locked in.

Then the bubble collapsed. Japanese interest rates fell for years. By the late 1990s, insurers were earning near-zero returns on their bond portfolios while still owing 5.5% to policyholders. The gap was called the "negative spread." It ate them alive, one by one.

Seven insurers failed between 1997 and 2002. Together they left ¥2.68 trillion in negative equity.

After Nissan Mutual came Toho Mutual, Dai-Hyaku Mutual, Taisho Mutual, Chiyoda Mutual, Kyoei Life, and Tokyo Mutual. Seven companies. The combined assets of those seven firms represented about 8.6% of total Japanese life insurance assets. When Chiyoda Life failed in October 2000, Tokai Bank lost ¥74 billion because banks and insurers provided each other capital — a practice called double-gearing.

The survivors were the big four: Nippon, Dai-ichi, Sumitomo, Meiji Yasuda. They made it through because they were larger, better diversified, and held enough assets to absorb the negative spread.

Today, those same four survivors sit on ¥15.13 trillion in unrealized losses on domestic bonds. The direction of the problem has flipped. In the 1990s, low rates made their liabilities too expensive. In the 2020s, rising rates are destroying the market value of their assets.

Different direction. Same vulnerability: a mismatch between what they promised and what their portfolio can deliver under current rates.

I have no idea if this leads to failures. These firms are far larger and better capitalized than the seven that fell. But the structure rhymes. A promise made in one rate environment. A portfolio stuck in another. And a gap that grows every quarter the BOJ keeps hiking.

Seven quarters of growing losses. Seven funerals in five years, twenty-five years ago. I am not predicting the next part. But I am watching it. We'll see.