# France’s day of reckoning; this is not 1999

_France has the biggest state and the highest taxes of any large developed economy, its bonds now yield more than Italy’s, and its voters are turning to Marine Le Pen. Plus Ukraine and Iran, two UK jobs measures heading in opposite directions, and why the AI sceptics are a healthy sign._

Neil Woodford · 23 October 2026 · 12 min read

![Protesters in France carry a "Frigo Vide" sign ("empty fridge") as they protest against the cost of living in Paris, France in October 2026.](https://cdn.sanity.io/images/v3acfbvo/production/02834c290c931bb8412c55b6855b105a28f70d2a-2784x1856.jpg?w=1600&fit=max&auto=format)

_Paris, France, October 17 2026, urban riots in the city of Paris as the high cost of living continues to create unrest._

---

This has been a very interesting week, albeit one characterised by [worrying riots in France](https://www.hawaiipublicradio.org/national-international/2026-10-20/france-announces-nightly-curfews-for-as-long-as-necessary-in-paris-and-other-cities) and continued conflict in Ukraine and in the Middle East. The issues that have grabbed my attention are those confronting the French economy, continued and potentially positive developments in the negotiations to end the US/Iran conflict, and the crisis narrative that has developed around the companies most exposed to the AI infrastructure boom. 

_[Watch: We discuss the AI boom and the recent selloff on the Show this week — The AI Boom Is Only in the Foothills](https://www.noisecancelling.co/the-show)_

Given the inclination of the media to catastrophise everything, there has been no let-up in the negative reporting of all of these stories, and whilst one could argue that there is not much to be positive about in relation to the riots that affected a number of European countries, one might hope that their underlying causes might at last trigger a generally complacent and aloof governing class to take them a little more seriously. 

As for the ongoing military conflicts, there have at least been tangible developments this week that some informed commentators have characterised as positive, especially in relation to Iran and the US.

## Ukraine: the diplomacy steps up

There has been a lot of diplomatic engagement this week involving multiple US envoy trips to Europe and meetings with leaders on both sides. Apparently, the US special envoy, Steve Witkoff, [has discussed a resumption of peace talks](https://greekreporter.com/2026/10/21/ukraine-peace-talks-putin-witkoff/) between Russia, Ukraine and the US, to be followed by [a possible summit between Presidents Trump, Xi and Putin](https://news.liga.net/en/amp/politics/news/putin-proposed-meeting-trump-in-china) next month. 

Whilst the intensity of the diplomacy is to be welcomed, it is of course accompanied by appropriate levels of “cautious expectation” given President Putin’s “limited willingness so far to engage seriously in negotiations”. 

Where this all leads is not at all clear, but at the very least it appears that the world’s most powerful leaders are engaging more seriously in an effort to bring this nearly five-year-long conflict to an end. That, in and of itself, is positive, since if the talks do go ahead, all parties will lose political capital if nothing positive emerges. 

The prize for the world economy is, of course, lower energy prices, which would inevitably follow a settlement; the same goes for the conflict between Iran and the US, which, although currently limited in scope and intensity, is ongoing.

## Iran: close, but no deal

Here too, behind the scenes, diplomatic engagement has noticeably stepped up, with a new and interesting development emerging this week. Indeed, such was the level of progress at the start of the week that there was some speculation emanating from Doha that a US/Iran breakthrough was possible within days. 

Omani, Qatari and Pakistani diplomats have all been involved, along with Russia’s Sergei Lavrov, who [has held a number of meetings in Iran](https://www.hngn.com/articles/273557/20261021/iran-says-us-accepted-seven-day-hormuz-plan-wants-more-strait-stays-shut.htm). According to one informed source, the growing involvement of Russia as a key player across every major diplomatic track is a new and potentially positive development. The angle here is that Russia, acting as a trusted ally of Iran, could play a role in [managing Iran’s enriched uranium stockpile](https://www.adaderana.lk/news/121267/russia-renews-offer-to-receive-irans-enriched-uranium). 

However, by the end of the week, the lack of a positive announcement, despite Iran’s foreign minister, Abbas Araghchi, suggesting that the US had accepted the broad conditions of Iran’s proposals while seeking additional clauses, resulted in yet more disappointment. The explanation seems to be that Iran’s hardliners had once again resisted the compromise offer.

## More oil, higher prices

Against this backdrop, and despite further Iranian attacks on shipping in the Gulf, Middle East oil flows have continued to grow. [On Wednesday Kpler reported](https://gcaptain.com/middle-east-oil-flows-near-prewar-levels-defy-hormuz-attacks/) that so far in October an average of 16.7mn barrels a day had left the region, and that in the week to 18 October crude and condensate flows out of the Persian Gulf averaged 24.6mn barrels a day – a new high. 

It’s a little odd to see the crude price keep climbing to [over $113 a barrel](https://tradingeconomics.com/commodity/brent-crude-oil). Further Houthi attacks on Saudi infrastructure haven’t helped sentiment, and, of course, higher shipping and insurance costs are boosting the price, but I’m a little perplexed because these additional costs aren’t new and have been elevated for months. 

Could part of the explanation be that the industry is extracting maximum economic rent whilst it can? Either way, if this continues and Iranian shipments remain at zero, the downward pressure on the crude price and the increasing economic pressure on Iran will tell. Rebuilding inventory may well be another factor, though the data don’t yet show it, and against this are significant production increases from across the globe, especially from North and South America.

## France’s day of reckoning

France’s economic and social problems have been front-page news this week, ironically coming only days after [Andy Burnham reignited the idea of Britain rejoining the EU](https://www.expressandstar.com/uk-news/burnham-not-ruling-out-going-all-the-way-and-rejoining-european-union-9230143). 

For me, the social and economic stress are inextricably linked, although the riots themselves appear to be as much a product of simmering tension in urban communities where there has been little successful integration in _les banlieues_, which have been ravaged by violent crime and unemployment for years. 

France has seen much finger-pointing, but the radical left appears to bear a disproportionate share of the blame. Either way, the French economic model appears to be breaking down. [Government spending in France](https://www.insee.fr/fr/statistiques/8997691) takes a higher share of GDP than in any other large developed economy, at about 57%, and so does [tax](https://www.oecd.org/en/publications/revenue-statistics-2025_3a264267-en/full-report/tax-levels-and-tax-structures-1965-2024_2033f3ea.html), at close to 44%. 

France also has a ridiculous [retirement age](https://www.service-public.gouv.fr/particuliers/actualites/A18825?lang=en) and pension bill (the age is 62 years and nine months, now that the 2023 reform has been suspended), a [debt-to-GDP ratio of 119%](https://www.insee.fr/fr/statistiques/9053525), which [economists commissioned by the Prime Minister’s office](https://www.publicsenat.fr/actualites/economie/la-france-devra-realiser-126-milliards-deuros-deffort-budgetaire-dici-la-fin-du-prochain-quinquennat-pour-stopper-la-progression-de-la-dette) say could reach 130% by 2030 if nothing changes, and a deficit of over 5% of GDP. Its ten-year bond yield has now become detached from Germany’s: [on Thursday it closed above 5%](https://en.ilsole24ore.com/art/btp-spread-rises-by-126-basis-points-day-of-selling-yield-459-per-cent-AJWbCgrB), 174bp more than the equivalent German Bund and significantly more than Italy’s and Greece’s ten-year yields. 

And yet more protests are planned this weekend, and the [main unions have called a national strike](https://lebonbon.fr/paris/news/apres-17-octobre-deux-nouvelles-manifestations-24-31-octobre/) for 5 November, demanding yet more money for public services and pensions.

![France’s debt has gone from a fifth of GDP to more than all of it](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/chart-weekly-22oct26-france-debt-75582fad892c-light.png)

_The 2008 crisis and the pandemic each lifted France’s debt to a new level, and neither was unwound. It was 115.5% of GDP in 2025; INSEE puts it at 119% at the end of June, already above the IMF’s figure for this year._

From a distance, and not a very distant one, this looks like an incredibly challenging set of problems for France’s new president, who will be elected [next April and May](https://www.service-public.gouv.fr/particuliers/actualites/A15053?lang=fr), and of course for the EU. 

The economy urgently needs austerity, but its population, or at least parts of it, appears totally opposed to any cuts to welfare and benefits. If Andy Burnham and his backbenchers are not paying close attention, they should be, because this appears to be the model they are wedded to, but one which will inevitably lead, as France is about to learn, to a day of reckoning where its economic and spending decisions are made for it and not by it.

## UK jobs: two measures, two stories

I should also mention the important UK macro data released this week. On Tuesday, the ONS published [August’s labour market numbers](https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentandemployeetypes/bulletins/uklabourmarket/october2026). Before summarising what they showed, it’s important to note again that the “official” measure of employment, and the one the OBR has consistently used to measure the economy’s productivity, the Labour Force Survey, continues to move in a completely different direction from payrolled employment. 

Indeed, the gap between the two is getting wider, with the LFS showing employment up 0.3% in the three months to August and 1% higher than a year ago, whilst payrolled employment fell by 0.1% in the three months to August and, on September’s flash estimate, is down by 180,000, or around 0.6%, since October 2024 (when Rachel Reeves’s first Budget resulted in huge increases in the cost of employing people).

![Since the October 2024 Budget, the two jobs measures have gone opposite ways](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/chart-weekly-22oct26-uk-jobs-98821f3e0709-light.png)

_HMRC counts 194,000 fewer payrolled employees in August than in October 2024 (181,000 on September’s flash). Over the same period the Labour Force Survey has employment up by almost 800,000._

In summary, August’s data showed that the labour market is stabilising after a period of weakness. Unemployment fell, but then so did the participation rate, and vacancies fell again, by 21,000, the biggest quarterly fall since the start of the year. Whole-economy earnings rose 3.7% in the three months to August, which compares with 5% a year ago, although public sector pay growth is still running at 6.3%. Private sector regular pay growth remained steady at 3%, which compares with 4.4% a year ago. 

Unlike Megan Greene, I think these data are an indication of moderating wage pressures and an overall slowing labour market and, going forward, are consistent with inflation falling to the MPC’s 2% target, assuming, conservatively, that productivity growth is only 1%. (I think it could be more like 2% right now.)

Finally, next Wednesday we will hear what the OBR thinks the economic future will look like. I wonder if it will be brave enough to admit that the LFS data it’s been using for years is now totally unreliable?

## UK inflation: fuel does the damage

The second important release this week was [the September inflation data](https://www.ons.gov.uk/economy/inflationandpriceindices/bulletins/consumerpriceinflation/september2026) on Wednesday. CPI came in slightly ahead of expectations at 3.5%, up from August’s 3.1% but still lower than a year ago and lower than in the US and the euro area. Unsurprisingly, the damage was done by transport fuel costs, despite fuel being less than 3% of the CPI basket. Having increased by 7% in August, motor fuel prices rose 5.4% in September and are 30% higher than a year ago, adding roughly three-quarters of a percentage point to the annual inflation rate. Including gas and electricity, overall energy price inflation was 18% in September, highlighting again why there will be a policy response from the government.

**3.5%** — UK CPI inflation in September, up from 3.1% in August. Core inflation was 2.8%

These data were somewhat depressing, frankly, but more encouraging was that core inflation came in at 2.8% which, although higher than in August, was significantly lower than last September’s 3.5%, illustrating once again the absence of the much-discussed second-round effects.

Looking forward, energy prices are the key variable, and so the economy’s inflation rate and interest rate path depend almost entirely on the conflict in the Middle East. If we assume some kind of settlement on a relatively near-term time horizon and factor in oil prices falling back to $80 a barrel, then UK inflation should be close to its target by the end of next year, which should in turn allow for rate cuts. The fact that the “market” is factoring in [up to four interest rate increases](https://www.euronews.com/2026/10/21/uk-inflation-jumps-more-than-expected-fuelling-rate-hike-expectations) over the same time period seems extraordinarily odd to me and, I am sure, will be completely wrong once again.

_Why market pricing for Bank Rate keeps getting it wrong:_ [When all you have is a hammer](https://www.noisecancelling.co/read/when-all-you-have-is-a-hammer) — The OBR builds its forecasts, and with them the government’s fiscal choices, on futures and other market prices. Those markets exist to manage today’s risk, not to predict where interest rates or energy prices will be in five years, and their record shows it.

## AI: this is not 1999

The final subject I wanted to address this week is the AI-bubble-bursting narrative that appears to be preoccupying financial markets right now. 

We address this in this week’s Show, so I won’t spend too much time on it here. However, it’s an issue where I find myself once again at odds with the emerging consensus which, at least this week, appears to have decided that everything AI-related is about to go bang. 

_[Embedded media](https://youtu.be/gWIIfuzP0hk)_

The problems have seemingly come all at once and include rising interest rates and bond yields, intense competition for capital in debt markets, fears over the financeability of AI infrastructure spending commitments, doubts about the returns on this massive investment programme, competition from China in semiconductors and LLMs, regulation, [community resistance to data centre builds](https://www.datacenterfrontier.com/site-selection/article/55359925/community-opposition-emerges-as-new-gatekeeper-for-ai-data-center-expansion) and power bottlenecks, which are leading to project cancellations across the world, and, I suppose, fears about the downside of more capable AI in relation to cyber security.

This is a very long list, and I do not intend to address all these points here, but I do want to make a few general points. The first is that the fact these issues are being aired so prominently is a very healthy sign. Readers will be familiar with the parallels being drawn with the tech bubble, which inflated and then burst so spectacularly in 2000. 

Comparisons can be helpful, but there are some very important distinctions to be drawn too. For example, I can recall very well the two years that led up to the tech bubble bursting in 2000. It was characterised by the rapid recruitment of an overwhelming consensus that believed utterly in the riches that would flow from the new companies being created and listed at the time to exploit the opportunities presented by the internet and telecoms infrastructure. I was a sceptic and extremely lonely. It was very hard to find anyone who didn’t believe utterly in what I saw as a bubble, and indeed such was the isolation that I only just managed to hang on to my job. 

Now, you do not have to try too hard to find legions of sceptics of the AI industrial revolution. This week it was [Ken Griffin](https://decrypt.co/381578/ken-griffin-ai-bubble-gpus-airline-seats), who said that AI was almost certainly in a bubble. Ray Dalio, Howard Marks, Steve Eisman and Michael Burry have all said the same in recent months. No one of equivalent stature and profile was saying that back in 1999, as far as I can remember. I do recall one famous statement from the economist Paul Krugman, who said the internet’s impact on the economy would be no greater than the fax machine’s, but that was about it. He was, of course, spectacularly wrong.

The other point to make is that some of these challenges are the product of forces that naturally act as a brake on economic and financial market excess: the higher cost of capital, for example, or the power bottlenecks, or indeed regulation. 

As long as these forces don’t fundamentally change the direction of travel or disable the ability of companies at the frontier to drive the technology’s development, they should be seen as the pause that refreshes, in my opinion. 

Finally on this point, it’s important to remember how well some of the leading companies in this space are performing. Operationally, the revenues and earnings (where relevant) of companies like Tesla, Meta, Micron and even Anthropic (and others) are typically beating recently elevated expectations, even though their share prices may not have responded well to the results announcements.

## What to look out for next week

There’s a lot of macro data in the calendar next week, especially from the US. The highlights will be [US inflation, GDP and labour market data on Thursday](https://www.bea.gov/news/schedule/full), all of which will be followed closely by the bond market. There is hardly any data coming next week from the UK, but the highlight will be the Budget on Wednesday. I do not expect the Chancellor to say anything of substance – after all, there is not much he can do. I do expect a series of small measures, one of which will be another tax raid on the UK banks, which has been well trailed in the media and will, I suspect, have no impact on the stocks. I am told that the planned initiative to relieve the pressure on household energy bills will be held back until later in the year. Frankly, I am expecting about an hour or so of utter guff, unfortunately.

The UK results diary is getting a little busier, with results from Lloyds and Shell on Thursday and BP on Friday that will attract my attention. The corporate diary in the US remains very busy.

## Three charts to finish

### The French appear to be making their minds up

[Ifop’s latest poll](https://www.sudradio.fr/politique/exclu-presidentielle-2027-apres-les-violences-du-17-octobre-marine-le-pen-senvole), taken just after last weekend’s riots, has Marine Le Pen winning every second-round match-up it tested, by margins of between 12 and 36 points.

![Le Pen wins every run-off Ifop tested, the closest by 12 points](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/chart-weekly-22oct26-france-runoff-459303219b11-light.png)

_Fieldwork began on the evening of 17 October, as the riots started. Ifop has Le Pen ahead in every second-round pairing; even Édouard Philippe, her strongest opponent, loses 56–44._

### AI: it’s really, really early

Fidelity’s chart puts AI on the same scale as every general-purpose technology of the last two centuries. Most of them took a decade or more to go from early adopters to near-universal use; AI, by Fidelity’s measure of the companies its analysts cover, has barely started.

![Adoption rates of different technologies](https://cdn.sanity.io/images/v3acfbvo/production/6ab214f4b65232cbe4dddeabca0f07b918c8a8ac-1225x627.png?w=1600&fit=max&auto=format)

_Fidelity Investments, Quarterly Market Update, Q4 2026_

### Bristol’s ketamine problem

It appears as if Bristol City Council, which is run by a minority Green administration, should be paying quite a lot of attention to this horrifying statistic. [In the EU drugs agency’s wastewater study](https://www.euda.europa.eu/publications/html/pods/waste-water-analysis_en), Bristol’s sewage carried more than twice as much ketamine per head as Amsterdam’s, the next highest in Europe.

![Bristol’s ketamine load is more than twice Amsterdam’s](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/chart-weekly-22oct26-ketamine-025df5db2141-light.png)

_Outside Europe, only Whitehorse in Canada, at 253, came within 100 of Bristol. Ketamine loads rose by about 41% across the cities measured in both 2024 and 2025._