EN IT

The US 10-year towards 4.8% flips the rate outlook: a hike is now priced for 16 September. Energy first, healthcare last, with Amgen stock down 10.08% and Baidu down 6.96% on 96.6% buy-side volume. And the FTSE MIB has turned to sell, closing at 12.20% after twenty-one weeks.

Technical analysis: the FTSE MIB turns to sell after 21 weeks, energy leads, US healthcare collapses
Economic Observatory · The session

Technical analysis: the FTSE MIB turns to sell after 21 weeks, energy leads, US healthcare collapses

8 September 2026 AiTrading67 · Trade Desk Observatory Markets

Wall Street reopened on the sell side after the long weekend, and there was a single cause: the US 10-year yield climbed towards 4.8%, its highest in almost two years. On top of that came a jobs report at roughly three times expectations, which flipped the rate outlook — markets now price a 60% chance of a hike on 16 September, where in late August they were betting on a cut.

In the cards below you will find the technical analysis of the names that moved the session, with our model's position on each one, long or short, and the date the signal fired. On US stocks: Micron stock and SanDisk stocks led the memory names higher while Shopify stock fell 7.60%, and healthcare took the worst of it with Amgen stock down 10.08%. In Europe healthcare is the weakest sector across day, week and month. On the Milan market, Telecom Italia gained 2.89% on Poste's revised offer while STMicroelectronics was the worst name on the index, and semiconductor stocks split by product on both sides of the Atlantic.

Then there is the technical fact that concerns the Italian index: at Friday 4 September's close our model turned the FTSE MIB to sell, closing a buy signal that had run for twenty-one weeks with a 12.20% gain. Since Monday the Italian basket has been working from the other side. Everything below comes from a systematic model — ai trading applied to roughly 450 instruments — and remains a reference point: the decision always belongs to whoever takes it.

The news that moved our instruments
🌐
GLOBAL
The rate reversal

Markets now price a US hike on 16 September, not a cut

The American reopening had a single theme, and it was the cost of money. The US 10-year yield rose towards 4.8%, its highest since October 2023, inside a global bond sell-off. The jobs report did the rest, with 162,000 new positions against expectations near 56,000 and upward revisions to prior months: futures now put roughly a 60% probability on a rate hike at the 16 September meeting. In the background, the European Central Bank is expected on Thursday to take its main rate to 2.50%.

Indices paid unevenly. The S&P 500 ETF S&P 500Long shed 0.55% and the small-cap ETF Russell 2000 ETFShort 0.45%, while the Nasdaq 100 ETF Invesco QQQ TrustLong stopped at 0.08%. Europe was flat: Frankfurt unchanged, Paris up 0.16%, Milan down 0.11%, London down 0.04%.

How I read itA hike priced at 60% is not a nuance. It reverses an assumption the market had been holding for months, and it shifts the ground beneath the sectors. Businesses valued on tomorrow's earnings — software, biotech, anything with distant revenue — lose more when money costs more; businesses that produce and collect today gain. Tuesday's rotation is precisely that, and it shows up more clearly in the sectors than in the indices. On our dashboard four stress measures are rising together: equity volatility, bond volatility, tail risk and the Italy-Germany spread. When tail risk climbs alongside ordinary volatility, the market is not merely more nervous: it is paying more to insure against extremes. For risk management the practical consequence is that this week has two dates, Thursday and the 16th, and between them positions are better kept smaller than usual.
S&P 500 long (G / P) · Invesco QQQ Trust long (G / P) · Russell 2000 ETF short (G / P) · DAX ETF long (G / P) · FTSE MIB ETF short (G / P) *
↑ back to menu
📈
ROTATION
Energy first, healthcare last

The clearest rotation in weeks: buy what produces today

Tuesday's sector grid says one thing, and it says it on both sides of the Atlantic. In the United States energy XLELong led with 1.11%, and it carries 12.64% over the month, the highest figure on the whole grid; utilities XLUShort followed at 0.86%. At the bottom, healthcare XLVLong lost 2.52% and financials XLFLong 1.38%.

Europe drew the same picture, more sharply at the top: basic resources EXV6Long gained 1.83% and oil and gas EXH1Long 0.76%, while European healthcare EXV4Short shed 2.33%. That last one is the only sector on the entire grid, either side of the Atlantic, losing across all three horizons: day, week and month.

How I read itOur model has been long US energy since 13 July and long European oil and gas since 20 July, so the session went our way on both; and it has been short European healthcare since 20 July, which is the third confirmation from the same day. On US healthcare, however, we are long since 26 May, and there the 2.52% drop went against us — we write it because the picture belongs whole. The operational point lies elsewhere and concerns risk management for anyone building a list: on Tuesday there were two themes, and each dragged a dozen names with it. In a rotation this compact the risk is not picking the wrong stock, it is picking five that are the same bet without noticing. Sector price action should be read as a correlation map, not as a list of independent opportunities.
US energy ETF long (G / P) · US utilities ETF short (G / P) · US health care ETF long (G / P) · US financials ETF long (G / P) · Basic Resources Europe long (G / P) · European oil and gas ETF long (G / P) · European health care ETF short (G / P) *
↑ back to menu
🇮🇹
ITALY
Milan

Telecom Italia up 2.89% on Poste's revised bid, STM the worst name on the index

Milan closed almost flat at 0.11% lower, after being down close to 1% earlier in the session, and the recovery has a single name. Telecom ItaliaLong gained 2.89% on the revised takeover and exchange offer from Poste ItalianeLong: the board raised the cash component by 30 cents, taking the consideration to 1.97 euros plus 0.218 new Poste shares per TIM share, an implied valuation of roughly 7.83 euros at early-September prices. Poste called the move final, dropped the 66.67% minimum acceptance condition and reopened the offer window from 21 to 25 September. Poste shares closed 0.53% higher.

The rest of the index split cleanly. On the way up: INWITShort at 3.12% and PrysmianShort at 2.39%, defensives and network infrastructure, with EniLong up 1.25% on crude and TenarisShort up 1.35%. On the way down: STMicroelectronicsShort worst on the index at 3.41% lower, Stellantis MilanoShort down 2.80%, FerrariLong down 2.20% and NexiLong down 1.40%.

How I read itOn STMicroelectronics there is a clarification worth more than the percentage, because it overturns the common explanation. The widespread reading is that the stock followed a general weakness in semiconductors; our data says otherwise. In the very same session Micron Technology gained 6.10%, Sandisk 11.90% and Intel 9.10%. American memory was running while STM lost 3.41%: the sector is not weak, it is Europe's analogue and automotive chipmakers, who share the label but not the cycle. A stock falling while its peers climb is paying for something that concerns it alone. Our model had a good day here: we are short STM since 13 July and short Stellantis since 8 June, and long Telecom Italia since 24 August. Eni remains the Italian name with the highest Signal Strength on our list at 83%, where Strength measures the conviction and breadth of a move and never its safety. On the other side, the day went against us on Inwit and Prysmian, where we are short.
Telecom Italia long (G / P) · Poste Italiane long (G / P) · INWIT short (G / P) · Prysmian short (G / P) · Eni long (G / P) · Tenaris short (G / P) · STMicroelectronics short (G / P) · Stellantis Milano short (G / P) · Ferrari long (G / P) · Nexi long (G / P) · UniCredit long (G / P) · Monte dei Paschi long (G / P) *
↑ back to menu
📊
OUR MODEL
The FTSE MIB flip

The buy signal closed at 12.20% after twenty-one weeks: from Monday we work the other side

The technical fact of the week does not show up in the index's daily change. On the FTSE MIB ETFShort the buy signal had run since 13 April, twenty-one weeks, carrying 12.20% with all three profit-taking windows already reached — the first in week six, the second in week eleven, the third in week seventeen. At Friday 4 September's close the price settled below its own exit level and the weekly signal turned: since Monday the Italian basket has been working from the other side.

The Inversion Point has repositioned to 54.75, or 2.40% above the current price, which is where a protective level sits on a newly opened short. The two sessions since the flip have left the price essentially unchanged, down 0.11% on Tuesday: the signal is new and has not yet produced a move.

How I read itThis is the best way a trade can end, which is why it is worth telling. The distance was covered in full, all three targets were banked between week six and week seventeen, and at the end the exit level did its job and closed the position at 12.20%. Nothing went wrong: a cycle simply ran its course. The numbers had been saying so for weeks. Signal Strength on this basket had fallen to zero out of one hundred, the floor of the scale, against 7 for the S&P 500 ETF and 17 for the Nasdaq one. Weekly money flow is the only negative one among the three baskets. And the exit level had stopped rising for two consecutive weeks, which on a rally is the first warning. One number points the other way and keeps the reading delicate: the weekly directional index sits at 33, against 13 for the S&P and 15 for the Nasdaq. There is a trend here, and it is the only real one of the three. For anyone doing swing trading on the Italian index the question is no longer whether the rally holds: it is whether the newborn downtrend has room ahead of it.
FTSE MIB ETF short (G / P) · S&P 500 long (G / P) · Invesco QQQ Trust long (G / P) *
↑ back to menu
🇪🇺
EUROPE
Frankfurt, Paris and London

European healthcare loses across day, week and month: the weakest sector on the grid

The three main venues finished essentially flat: Frankfurt unchanged, Paris up 0.16%, London down 0.04%. Beneath the surface, though, the move followed the American template. The worst sector was healthcare EXV4Short at 2.33% lower, also negative on the week and the month; the best were basic resources EXV6Long at 1.83% and oil and gas EXH1Long at 0.76%.

Among individual names, Infineon TechnologiesShort remains the most closely watched German stock after the C2i Semiconductors acquisition announced last week, while SAPLong and Dassault SystemesLong stay on the weak side of European software. On AirbusShort we flag a divergence from the day's press coverage: some sources report the stock lower, our prices have it up 1.03%, and in this card our own feed prevails. Deutsche Bank reiterated its buy rating with a 232-euro target after 57 August deliveries.

How I read itEuropean healthcare is the only sector on the grid, either side of the Atlantic, losing across all three time horizons, and our model has been short that basket since 20 July: it is one of the positions working best right now. On European software the picture fits the rate theme — distant revenue is worth less when money costs more. The methodological note on Airbus deserves a line because it describes how we work: when an outside source and our own price feed contradict each other, the feed wins, and we check it name by name. Of eight European stocks for which yesterday's coverage carried a percentage, seven match ours to the cent and one does not. That one we report as we see it, and we say that we did.
European health care ETF short (G / P) · Basic Resources Europe long (G / P) · European oil and gas ETF long (G / P) · Infineon Technologies short (G / P) · SAP long (G / P) · Dassault Systemes long (G / P) · Airbus short (G / P) *
↑ back to menu
🇺🇸
UNITED STATES
Semiconductor stocks run, healthcare collapses

Memory runs, software falls: Amgen stock and Amgen stocks down 10.08%, Baidu down 6.96% with 96.6% of volume on the buy side

The American session split technology by product. On the way up: SandiskLong 11.90%, CoreWeaveLong 11.70%, IntelShort 9.10%, Micron TechnologyShort 6.10% and Advanced Micro DevicesShort 4.69%. On the way down: ShopifyLong 7.60% and BaiduShort 6.96%. Among the heavyweights, AppleShort lost 2.51%, MicrosoftLong 2.04% and NVIDIALong 2.01% ahead of a 0.25-dollar dividend going ex on the 10th, the same day as a conference presentation.

Outside technology the blow came from healthcare: AmgenLong fell 10.08%, Beam TherapeuticsShort 9.54% and Intuitive SurgicalLong 4.51%. In consumer discretionary ExpediaLong lost 7.90% and BookingLong 6.70% — the two large travel platforms on the same day and by a similar margin. On the China front, the US Department of Defense designated AlibabaShort and Baidu as entities tied to the Chinese military, barring them from federal contracts.

How I read itThe most interesting figure of the day appears on no price chart. On Baidu, down almost 7% on the American news, 96.6% of traded volume was classified on the buy side: underneath that collapse someone was buying all of it, plausibly the mainland channel that opened on 7 September when its class A shares entered Stock Connect. A drop with that volume signature is a change of hands, and it deserves to be told differently from a liquidation. Our model is short Baidu, so the session went our way, but this number keeps us cautious about our own position. On the others: we are short Apple since 10 August and the day agreed with us; we are short Intel while the stock rose 9.10%, which is the most uncomfortable position we hold right now. On Amgen we are long and the 10.08% fall went against us — three and a half times its typical range in a single session. On Booking and Expedia the price closed beyond the stop we had published, and when two direct competitors move together like that the cause is almost always sector-wide.
Sandisk long (G / P) · CoreWeave long (G / P) · Intel short (G / P) · Micron Technology short (G / P) · Advanced Micro Devices short (G / P) · Shopify long (G / P) · Baidu short (G / P) · Apple short (G / P) · Microsoft long (G / P) · NVIDIA long (G / P) · Amgen long (G / P) · Beam Therapeutics short (G / P) · Intuitive Surgical long (G / P) · Expedia long (G / P) · Booking long (G / P) · Alibaba short (G / P) *
↑ back to menu
📈
INDICES
The four European baskets

European indices: Milan is the only one that has already turned

DAX ETFLong closed unchanged, CAC 40 ETFShort up 0.16%, FTSE MIB ETFShort down 0.11% and FTSE 100 ETFLong down 0.04%. Distances from their records tell four different stories: the German basket is closest to its high at 2.21% below, followed by Milan at 3.49%, Paris at 4.95% and London at 7.68%.

On our signals the position is this: the German basket has been long since 13 April with the price 1.07% above its exit level; London has been long since 7 April with 1.63% of headroom; Paris has been short since 24 August, with the price 2.12% below its inversion level.

How I read itThe one that diverges today is Milan, and that is the information this card exists to provide. The Italian basket is the only one of the four to have crossed its exit level — it sits 2.40% below, while Frankfurt and London hold positive margins between 1% and 1.6% — and with Friday's close its signal turned. That the other three are holding is itself the news: this is not Europe turning, it is the Italian basket reaching the end of a cycle after twenty-one weeks and 12.20%. The French basket is worth adding: short since 24 August, and on Tuesday it closed higher anyway. A sell signal watching its own index rise is the case where risk management matters more than the thesis.
DAX ETF long (G / P) · CAC 40 ETF short (G / P) · FTSE MIB ETF short (G / P) · FTSE 100 ETF long (G / P) *
↑ back to menu
📊
INSTRUMENTS
The baskets that moved

The ETFs behind the session: the gold ETF, the semiconductor ETF and the energy ETF

GLD — SPDR Gold Shares, the most heavily traded gold ETF in the world — shed 1.73%, the widest move in the group. SOXX — the iShares Semiconductor ETF — gained 1.64% on the back of memory names, while XLK, the Technology Select Sector SPDR and the main US technology ETF, stopped at 0.32%.

Among the rotating sectors, XLE — the Energy Select Sector SPDR, the US energy ETF — gained 1.11% and carries 12.64% over the month; XLU, the US utilities ETF, added 0.86%. In Europe EXV6, the basic resources ETF, rose 1.83% while EXV1, the European banks ETF, was flat at 0.08% lower.

How I read itThe gold line explains the day best. The gold ETF has been ours since 3 August and it lost 1.73% on a day of strikes against energy facilities and an expired truce — exactly the backdrop that should send it higher. The reason sits in the rates card: when the real cost of money rises, holding an asset that pays no coupon costs more, and that cost beats safe-haven demand. It is a working reminder of how gold should be read in this regime. On the semiconductor ETF we are short since 13 July and the 1.64% gain went against us; we keep the distinction between the basket and its parts, though, because we are long the US technology ETF and short the chip-specific one, which is not a contradiction but a picture of rotation inside the sector. On the US energy ETF we are long since 13 July, and that is the strongest confirmation in the group. A note on data transparency: our series for crude as a commodity is frozen at 4 September, so the oil move is understated in this card.
gold ETF long (G / P) · semiconductor ETF short (G / P) · US technology ETF long (G / P) · US energy ETF long (G / P) · US utilities ETF short (G / P) · Basic Resources Europe long (G / P) · Banks Europe long (G / P) *
↑ back to menu
📈
OUR MODEL
The 4 September cohort

New buy signals: yesterday's changes, and one that matters

The weekly cohort was born at Friday 4 September's close: 21 new buy signals, 14 tradable. From Tuesday to Friday this card publishes only what changes — who crosses above or below their entry level, who flips side on the daily chart, who reports earnings, who leaves the cohort.

Yesterday the group finally moved in full, because Wall Street and São Paulo reopened: eleven of the twenty-one names are Brazilian and Tuesday was their first complete session of the week. The change that matters, though, involves an American name: Recursion PharmaceuticalsLong fell 5.23% inside the day's worst sector while remaining among the tradable signals with Signal Strength at 76%. Among the Brazilians, CemigLong gained 3.91% and leads the cohort.

How I read itRecursion is worth pausing on, because it is the teaching case of the week: a weekly buy signal born on a bar pointing the opposite way, inside a sector that lost 2.52% that day. We do not remove it from the list, since for us risk is an attribute and not a veto, but it is the kind of signal to revisit next week rather than chase now. There is also a structural observation about the cohort that concerns anyone building a list: among the tradable Brazilian names there are three banks plus a holding company that controls one of them — four signals that amount to two bets. Risk management here does not run through the stop, it runs through the count. The caveat always applies and we repeat it: past data does not guarantee future results.
Recursion Pharmaceuticals long (G / P) · Cemig long (G / P) · Volkswagen AG Pref long (G / P) · Sandisk long (G / P) · Iveco long (G / P) *
↑ back to menu
📅
CALENDAR
Earnings in the next seven days

This week's earnings: Adobe and Oracle tomorrow night, the thermometer on AI spending

Four companies in our universe report over the next seven days, and the first two arrive together tomorrow after the American close.

Thursday 10 September

Adobe Long

Oracle Long

Avio Short

Wednesday 16 September

MFE-MediaForEurope Short

How I read itOracle and Adobe report on the same day and the market reads them as the thermometer for spending on artificial-intelligence infrastructure: consensus puts Oracle revenue near 19.13 billion dollars and Adobe adjusted earnings at 6.09 dollars per share on roughly 6.7 billion of revenue. Options price implied moves of about 11% and 5.4%. The operational point is not the forecast, it is the stacking: those numbers land on an already-tense sector, in a week that closes with the American central bank meeting on the 16th. Anyone holding technology names with distant revenue faces two events in a row, and the second amplifies the first. The way we look at earnings is measured rather than impressionistic: results do not shift average returns, they double the dispersion. So they are not a reason to enter nor a reason to stay out — they are a reason to enter smaller.
Adobe long (G / P) · Oracle long (G / P) · Avio short (G / P) · MFE-MediaForEurope short (G / P) *
↑ back to menu
📈
UNITED STATES
The two American baskets

Technical analysis of SPY and QQQ: half of Tuesday's gap is sector composition alone

SPY, the S&P 500 ETF S&P 500Long, lost 0.55%; QQQ, the Nasdaq 100 ETF Invesco QQQ TrustLong, stopped at 0.08%. On a risk-off day the more aggressive of the two held up better, and the reason is measurable: healthcare and financials, the session's two worst sectors, weigh 22.06% in the S&P 500 and 5.35% in the Nasdaq 100. That weighting difference alone accounts for 0.28 points of the 0.47 actually observed.

The geometry of the two trades remains very different. The S&P 500 ETF is in the twenty-second week of a buy signal carrying 12.73%, with all three profit-taking windows banked, and sits 1.72% below its record; the Nasdaq 100 ETF is in its fifth week, still 1.74% below its own entry level, with all three windows ahead and the time gates open, and sits 4.05% below its high.

How I read itThe comparison reduces to one sentence: in the first a result is being defended, in the second it has yet to begin. On the broad basket there is a Tuesday figure that deserves attention and does not show up in the price: in a session that closed down half a point, only 17.98% of volume was on the buy side. Price giving up little, volume leaving heavily, a point and a half from an all-time high — that is not distribution driven by weak prices, it is a change of hands near the top. The measure that captures the difference between the two baskets is Signal Strength, which gauges the conviction and breadth of a move and never its safety: 7 out of 100 on the broad basket, 17 on the technology one. Both are low readings. On a mature structure moved by almost no conviction, and with the American central bank meeting on 16 September, this is a phase for risk management on what you hold rather than for building fresh exposure to the broad baskets.
S&P 500 long (G / P) · Invesco QQQ Trust long (G / P) · Russell 2000 ETF short (G / P) *
↑ back to menu

The desk's take

The wider picture: a market rising on fewer and fewer legs

The week is governed by two closely spaced events and one variable almost nobody watches. The events are the European Central Bank on Thursday the 10th, expected at 2.50%, and the Federal Reserve on 16 September, which the market now pictures hiking after a jobs report at three times expectations. The variable is breadth: participation in the rally has fallen below half the American market — fewer than one stock in two sits above its own medium-term average — and it is down 32% over the month. An index that holds while participation thins is an index held up by a handful of names, and that fragility stays invisible until it gives. With crude under pressure and the cost of money rising, picking the wrong stock in a phase like this costs double: there is no tide to make up for it.

What you find here. If you are trying to understand how to manage risk in trading — stops, break-even stops, trailing, and the declaration of stops that have been taken out — or how to read chart patterns (double bottoms, Bollinger bands, the Ichimoku cloud, price action trading), this is the desk's trading journal: published every morning before the US premarket, with our model's position on every instrument mentioned.

* Our system — The rows of names at the foot of each card carry our model's position, long or short. Position from the latest published weekly analysis. Levels, targets and the trading plan sit in each instrument's own page. The two letters in brackets lead to the instrument page (G) and the full analysis (P).

Disclaimer. The content on this page is published for educational and informational purposes and reflects the author's personal opinion and technical analysis. It does not constitute financial advice, an investment solicitation, or a personalised recommendation. Trading in financial instruments carries a high level of risk and may result in substantial loss of the capital invested. Every operating decision rests solely with the reader, who acts independently and in full awareness of that risk. AiTrading67 is not a broker or financial intermediary and is not affiliated with any broker or trading platform.
Sources
Share WhatsApp Telegram Gmail LinkedIn