Having tracked the FTSE 100 for years alongside sterling, I have made one habit non-negotiable: before crediting any rally to the companies, check what the pound did. The index closed at 10,608.92 on 10 September 2026, according to the delayed exchange feed carried by CNBC, after falling 1.31% on 9 September and another 0.57% the next day. That two-session drop of 1.88% took it to its lowest close since 21 July. Yet the benchmark is still 15.0% higher than it was on 10 September 2025 and only 2.8% below its record close of 10,910.55, set on 27 February. British blue chips have plainly had a good year. The fourth-quarter question is who, exactly, had it, and whether those same companies can carry the index another leg higher with a Bank of England decision six days away.
Start with the currency, because it clears away the most common explanation. GBP/USD fixed at 1.3520 on 10 September, against 1.3534 on the same date a year earlier. Sterling did nothing, so none of the index's gain came from a weak pound inflating overseas profits. The gain came from a narrow group. Using FTSE Russell's end-August weights and each group's price move over the twelve months to 28 August, I estimate that banks, miners, the two oil majors and Rolls-Royce, together 43.1% of the index, delivered roughly 15.6 to 16.2 of its 17.8 points. The other 56.9%, which holds AstraZeneca, Unilever, RELX, Diageo and the London Stock Exchange Group, added about two points between them. The FTSE 100 is not a broad bet on Britain. It is a concentrated bet on rates and raw materials, with a defensive half that has been dead weight.
Key facts
- The FTSE 100 closed at 10,608.92 on 10 September 2026, down 1.88% over two sessions and at its lowest close since 21 July — CNBC delayed exchange feed (FTSE International), retrieved 11 September 2026.
- Banks make up 19.28% of the index and HSBC alone 10.19%, while technology is 2.53% — FTSE Russell factsheet, data as at 31 August 2026.
- Total return was 21.6% over the twelve months to 31 August, against a price gain of about 17.8%, implying roughly 3.2% came from dividends — FTSE Russell factsheet and CNBC closes, our calculation, 11 September 2026.
- Bank Rate is 3.75%; the July Monetary Policy Committee voted 6–3 to hold, with three members voting for 4%, and the next decision is due on 17 September — Bank of England, 30 July 2026.
- UK CPI inflation was 2.9% in July, up from 2.6% in June, with the August figure due on 16 September — Office for National Statistics, 19 August 2026.
- UK GDP grew 0.4% in July, the eighth consecutive three-month period of growth — ONS monthly GDP estimate, 11 September 2026.
- GBP/USD fixed at 1.3520 on 10 September 2026 against 1.3534 a year earlier, inside a twelve-month range of 1.3044 to 1.3817 — ECB reference rates via frankfurter.dev, 11 September 2026.
Forty-three per cent of the index did the lifting
The factsheet FTSE Russell published for the end of August is the cleanest place to see how lopsided the benchmark has become. The ten largest holdings account for 48.36% of its £2.57 trillion net market value. HSBC, at 10.19%, is heavier than the entire technology, telecoms and media sectors combined. Banks as a supersector weigh 19.28%, energy 10.43% and basic resources 8.41%. Compare that with Frankfurt, where our DAX forecast argued the German benchmark is trading a fiscal promise through defence and infrastructure names. London's index trades something older: net interest margins and commodity prices. Our FTSE 100 market page collects this forecast alongside the site's other index coverage.
The chart below sets out the path. The index climbed from 9,196 at the start of September 2025 to its February record, lost just over 1,000 points into the March low, and spent the summer rebuilding.
The table breaks the twelve-month gain into its sources. The price moves are closes from 29 August 2025 to 28 August 2026, chosen to match the date of the weights, and the contribution column is my own arithmetic: each group's end-August weight, scaled back to its starting weight, multiplied by its return. For basic resources I used Rio Tinto alone as the low end and the average of Rio Tinto, Glencore and Anglo American as the high end, because the smaller members of that sector have not all kept pace.
| Group | Index weight, 31 Aug 2026 | 12-month price move to 28 Aug | Estimated contribution to the 17.8% gain |
|---|---|---|---|
| HSBC | 10.19% | +61.4% | about 4.6 points |
| Other banks (six members) | 9.09% | Barclays +37.8%, Lloyds +38.3%, NatWest +33.1%, Standard Chartered +57.1% | about 3.1 points |
| Basic resources (six members) | 8.41% | Rio Tinto +65.3%, Anglo American +87.4%, Glencore +103.8% | 3.9 to 4.6 points |
| Energy (Shell, BP) | 10.43% | Shell +22.8%, BP +19.1% | about 2.2 points |
| Rolls-Royce | 4.99% | +43.0% | about 1.8 points |
| Everything else | 56.89% | AstraZeneca +1.5%, RELX -22.5%, Diageo -16.6%, LSEG -1.8%, BAT -1.3%, GSK +27.6% | 1.6 to 2.2 points (residual) |
Two things follow. First, the index's recent calm hides a lot of stock-level violence. Measured against their highest closes of the past twelve months, AstraZeneca stood 24.3% lower on 10 September, RELX 29.9% lower and Diageo 21.5% lower. AstraZeneca is the third-largest weight at 6.98%, so this is no side-show. Second, the leaders are near their highs. On the same yardstick Standard Chartered closed 1.0% under its peak and HSBC 4.3% under. A leadership group that has already run hard is exactly where a forecast should look for fragility.
Index-level volatility has been flattered by that split, too. When miners rise while pharmaceuticals fall, the moves cancel out in the aggregate. FTSE Russell puts twelve-month volatility at 11.2%, and the twenty-session figure I calculate from daily closes is only 7.4% annualised. Quiet, in other words, is partly an accounting effect of two halves pulling in opposite directions.
Why a hike on 17 September cuts both ways
Most equity commentary treats a rate rise as bad news for shares. For this index the arithmetic is less obvious, and it is where I think the consensus framing is weakest.
The Monetary Policy Committee held Bank Rate at 3.75% in July, the level it has sat at since 18 December 2025. The vote was 6–3, and the three dissenters, Megan Greene, Catherine L Mann and Huw Pill, wanted a rise to 4%. Since then, the Office for National Statistics has reported CPI inflation climbing from 2.6% to 2.9%. The August print lands on 16 September, the day before the decision.
A hike would help the largest block in the index directly. HSBC's 2026 interim results, filed with the SEC on 4 August, said net interest income rose partly on "the benefit of reinvestment of our structural hedge at higher yields", and the bank firmed its guidance for banking net interest income to at least $46bn for 2026, from around $46bn, "reflecting a continued favourable interest rate outlook". Georges Elhedery, Group Chief Executive at HSBC, put the tone in one line: "HSBC is becoming the stronger bank we set out to build." The UK-focused lenders, whose earnings lean even harder on domestic rates, stand on the same side of that trade.
The same hike would hurt through the other channel. A firmer pound shrinks the sterling value of profits earned in dollars, and plenty of this index keeps its books in dollars. SEC filings show HSBC, Shell, BP, AstraZeneca, Rio Tinto and Diageo all report in US dollars. The first four alone carry 27.6% of the index's weight, before Rio Tinto and Diageo are counted. Every 1% rise in GBP/USD trims roughly 1% from the sterling value of those companies' dollar profits.
That is why I treat 17 September as a two-sided event for the FTSE 100. A hawkish surprise that lifts sterling towards the top of its twelve-month range, 1.3817 in January, would cost the dollar earners more than it gives the banks. A hold with a hawkish vote split would probably be the friendliest outcome of all. Our GBP/USD forecast works through the currency side of that meeting in detail, and the GBP/JPY forecast covers the Bank of Japan decision that lands the next day.
The copper session that broke the uptrend
The two-day slide had two distinct engines. On 9 September the selling was broad: of the twenty largest members I priced, seventeen closed lower, led by Anglo American, Rolls-Royce, the London Stock Exchange Group and BAE Systems. It was the same session in which the DAX lost 1.66%, which points to a continental risk-off move rather than anything British.
The next day was different and far more specific. Front-month COMEX copper futures fell 4.95%, from $6.8885 to $6.5475 a pound, according to CNBC's futures data. Anglo American dropped 4.86%, Glencore 4.11% and Rio Tinto 3.12%. Shell and BP rose as oil prices climbed, which softened the blow, but the index still finished 0.57% lower. For a benchmark that owes perhaps a quarter of its twelve-month gain to miners, a single bad copper day moves the whole benchmark. Our copper price prediction sets out why the metal's range matters for exactly these three stocks.
Technically, the damage is contained so far. The index closed below its 50-session average of 10,738 but remains above the 100-session average of 10,562 and the 200-session average of 10,378. I would not read too much into moving averages on a benchmark this concentrated, but the 200-session average is a useful marker. A close below it would mean the leaders themselves are being sold.
Buybacks run out on 23 October, and the Budget lands on the 28th
Flows are the part of the FTSE 100 story that rarely makes the headlines. Shell announced on 30 July a buyback of $4.232bn, made up of $3bn of new purchases plus $1.232bn carried over from a programme it suspended while agreeing to acquire ARC Resources, to run until 23 October, according to its SEC filing. HSBC started a buyback of up to $1bn on 6 August under agreements that end no later than 23 October. Together that is about £3.9bn of steady, price-insensitive demand from two members, roughly 0.15% of the index's net value, and both programmes are scheduled to finish by the same Friday.
Five days later, on Wednesday 28 October, the Chancellor delivers his first Budget; the Treasury confirmed the date on 31 July. John Healey, Chancellor of the Exchequer, told a growth audience on 7 September that "the UK stock market is hitting all-time highs" in his growth speech. On the FTSE 100's own numbers, that needs a footnote. The index closed at 10,822.13 that day, 0.8% under its February record close, and its all-time intraday high of 10,989.45 dates from 31 July. The gap is small. It still matters for anyone pricing Budget risk: an index that has not made a new closing high in six months is not yet priced for good fiscal news.
There is also the dividend cushion. FTSE Russell's twelve-month total return of 21.6% set against a price gain of about 17.8% implies roughly 3.2% of the year's return came from dividends. At today's level that is about 340 index points a year that the price index quietly hands back on ex-dividend dates. For anyone holding the benchmark through a UK 100 CFD, those payments arrive as cash adjustments rather than as index points, which is why the price chart understates what holders earned. It also means the price index needs to climb just to stand still in total-return terms.
Where the bear case gets its fuel
The honest objection to any constructive call is that concentration works in reverse. If copper extends the 10 September drop, if a hawkish Bank of England pushes sterling back towards 1.38, and if HSBC's rate tailwind is priced out by a global slowdown, the 43% that did the lifting becomes the 43% doing the selling. The defensive half would then have to rise to hold the index, and nothing in its recent record suggests it is ready. RELX has spent almost a year falling; AstraZeneca is still a quarter below February.
There is a precedent inside the chart. Between the 27 February record and the 23 March close at 9,894.15, the index fell 9.3% in under four weeks, and the intraday low that day was 9,670.46. FTSE Russell's one-year maximum drawdown on a total-return basis was 8.9%. My bear level of 9,700 simply returns the index to that March floor.
The call: 11,500 bull, 10,900 base, 9,700 bear to 31 December
Base case, 10,900 (50% probability), 2.7% above the 10 September close of 10,608.92. The Bank of England holds or hikes by a quarter-point without a sterling spike, copper stabilises, and the index grinds back to its February record by year-end as bank earnings and buybacks carry it. The 23 October end of the Shell and HSBC programmes and the 28 October Budget cause a wobble in late October that I would expect to fade.
RelatedUSD/BRL Forecast: 5.60 Bull Case vs 4.75 Bear Case for Q1 2027
Bull case, 11,500 (25% probability), 8.4% above the last close. This needs the laggards to join in. If AstraZeneca and RELX recover even half of their drawdowns while the banks hold their gains, the index clears its 10,989.45 intraday high and extends into a new range. Third-quarter results from HSBC and Shell that renew buybacks at similar size would supply the flow.
Bear case, 9,700 (25% probability), 8.6% below the last close. Copper keeps falling, a hawkish MPC lifts GBP/USD towards 1.38, and the Budget disappoints. That revisits the March low and gives back roughly two-thirds of the past year's gain.
Weighting the three gives an expected level of about 10,750, roughly 1.3% above the 10 September close, before dividends. That is why the stance is neutral, with modest upside and a conviction of three out of five.
What would change my mind? A daily close below the 200-session average near 10,378 while copper makes fresh lows would move me towards the bear case, because it would mean the leadership is breaking rather than resting. A close above 10,910.55 with AstraZeneca participating would move me towards the bull case, because it would show the rally broadening beyond rates and metals for the first time this year.
FAQ
What is the FTSE 100 forecast for the end of 2026?
Our base case is 10,900 by 31 December 2026, about 2.7% above the 10 September close of 10,608.92, with a 50% probability. The bull case is 11,500 (25%) if defensive heavyweights such as AstraZeneca recover, and the bear case is 9,700 (25%) if copper and a stronger pound pull the banks and miners lower. These are analytical scenarios, not advice.
Why did the FTSE 100 fall on 9 and 10 September 2026?
The first day was a broad European sell-off: seventeen of the twenty largest members fell and the DAX lost 1.66% in the same session. On 10 September the move was specific to miners. Front-month COMEX copper fell 4.95%, and Anglo American, Glencore and Rio Tinto dropped between 3.1% and 4.9%, while Shell and BP rose.
Does a weaker pound help the FTSE 100?
Usually, because many members earn and report in dollars. HSBC, Shell, BP and AstraZeneca report in US dollars and carry 27.6% of the index weight between them. A falling pound raises the sterling value of those profits. Over the past year, though, GBP/USD barely moved, from 1.3534 to 1.3520, so currency explains almost none of the index's 15% rise.
How would a Bank of England rate hike affect the FTSE 100?
In two opposite ways. Banks, 19.28% of the index, benefit from higher rates; HSBC has cited reinvesting its structural hedge at higher yields. But a hike tends to lift sterling, which reduces the sterling value of dollar profits at the big exporters and commodity producers. Three of nine MPC members voted for a hike in July, and the next decision is on 17 September.
What is the FTSE 100's dividend yield?
FTSE Russell does not print a yield on its factsheet, but the gap between the index's 21.6% twelve-month total return and its roughly 17.8% price gain implies dividends contributed about 3.2% over the year to 31 August. At current levels that is worth around 340 index points a year, which the price index gives up on ex-dividend dates.
Disclaimer
This article is analysis and commentary, not investment advice or a recommendation to trade any instrument. Index CFDs and other leveraged products carry a high risk of losing money rapidly, and your capital is at risk. Scenario probabilities are the author's judgement and can be wrong. Past performance is not a reliable indicator of future results. Figures are sourced and dated in the text; check current prices before making any decision.