EN IT

Technical analysis after the US inflation print: CoreWeave and Cisco stock light up AI, Nebius stocks jump 34%, Brazil falls again

US July inflation lands in line at 3.4% annual and September hike odds drop to 38%: Wall Street closes mixed with the Nasdaq ahead. Technical analysis of CoreWeave up 19.28%, Nebius stock up 34.14%, Cisco stocks near highs, and Eni leading the Milan market on crude.

Technical analysis after the US inflation print: CoreWeave and Cisco stock light up AI, Nebius stocks jump 34%, Brazil falls again
Economic Observatory · The session

Technical analysis after the US inflation print: CoreWeave and Cisco stock light up AI, Nebius stocks jump 34%, Brazil falls again

13 August 2026Fabio Gentili ObservatoryTechnical analysis

The number that was supposed to decide the day arrived and decided very little, which was the least-considered outcome. US July inflation rose 0.1% on the month and 0.2% at the core, taking the annual rates to 3.4% and 2.5%, both a tenth lower: the odds of a Federal Reserve hike in September fell to 38% from 48% the day before. Wall Street closed mixed with the Nasdaq ahead, lifted by AI earnings and held back by the large names that sit outside that theme.

In the cards below you'll find the technical analysis of the names that moved the session — CoreWeave and Cisco stock on results, Nebius stock up 34% and Super Micro up 19% on compute infrastructure, Dell stocks up almost 10%, NVIDIA stock back up 3% among semiconductor stocks, Alphabet stock still weak on capex, and Apple stock drifting with our sell signal. Nebius stocks and Cisco stocks were the two ends of the same theme. On the Milan market Eni followed crude while Leonardo rose on an RBC upgrade. For each name you'll find our model's position, long or short, with the date the signal fired.

The session cards
📈
OUR MODEL
View as of today

This week's new buy signals

There are 27 new buy signals from the week that closed on Friday, and this is their third day of life: the point where the difference between those that have started working and those that have not stops being noise. Below are 12 of them, ranked from best to worst by return from the signal, and the horizon throughout is weekly swing trading. Three families hold them together: gold and its miners, European and American defence, and large-cap technology.

Instrument
Signal
Weekly confirmed
My read
🇺🇸 NEMLong
Newmont
BUY
07/08
buy confirmed
week closed 07/08
The best on the card for a second morning, up 4.30% from the signal. This is the gold miner that moves more than the metal, and the metal has been rising for two weeks with plus 9.7% over four: the theme is supporting the signal rather than contradicting it. The severe-loss column stays empty over forty-seven cases, and past data does not guarantee future results.
🇮🇹 AZMLong
Azimut
BUY
07/08
buy confirmed
week closed 07/08
Italian asset management, up 2.73% from the signal with both horizons aligned. It remains the only Italian financial among this week's new buys, in days when the listing's large banks move by fractions of a point: the signal comes from the single name rather than from the sector, and that kind of signal ages better.
🇩🇪 RHMLong
Rheinmetall
BUY
07/08
buy confirmed
week closed 07/08
Up 1.90% from the signal. European defence has been the sturdiest theme of these three sessions and this is its German standard-bearer: the European industrials basket added 0.54% on the day and 5.63% over the month. The distance from the Point of Inversion is 12.5% though, a lot for a week-old signal, and it translates into a smaller size.
🇺🇸 GLDLong
SPDR Gold Shares
BUY
07/08
buy confirmed
week closed 07/08
Up 1.62% from the signal, and the calmest way to hold gold among the three instruments on our list: ordinary volatility and a 2.3% typical loss over fifty-three cases. The metal closed at record levels and rises while equity volatility falls 22.5% over the month — it is not being bought out of fear, it is being bought against real rates.
🇺🇸 NOCLong
Northrop Grumman
BUY
07/08
buy confirmed
week closed 07/08
Up 1.00% from the signal, with both horizons in agreement and an empty risk column. It is the American version of the defence theme and it moves less than its German counterpart: anyone looking for the same theme with less volatility finds it here, which is why it appears on this card and not among the outsized moves.
🇺🇸 ZMLong
Zoom
BUY
07/08
buy confirmed
week closed 07/08
Up 0.60% from the signal with results in twelve days and three active clues in the window that precedes them: improving money flow, rising buy volume and lows that have stopped falling. Our measurement of the July season says earnings do not move the average return but double the dispersion, so size stays reduced.
🇺🇸 NVDALong
NVIDIA
BUY
07/08
buy confirmed
week closed 07/08
Essentially flat from the signal at plus 0.06%, but on Wednesday NVIDIA stock gained 3.03% as appetite returned to semiconductor stocks. It is the name where the weekly structure and price behaviour stopped arguing for one session: real confirmation comes with results, which are still some way off.
🇺🇸 NOWLong
ServiceNow
BUY
07/08
buy confirmed
week closed 07/08
Up 0.05% from the signal, flat. Confirmation comes from the slow horizon, though: the monthly sell that weighed on this name appears among those at risk of flipping, because price has crossed the level the other way. Two horizons that were arguing and are settling their difference.
🇺🇸 PLTRLong
Palantir
BUY
07/08
buy confirmed
week closed 07/08
Down 0.56% from the signal after a 2.23% fall on the day. Friday's caveat stands intact: the distance from the Point of Inversion is 13.2%, the widest on the card, and on a week-old signal that means the stop sits a long way off. It also appears among the monthly sells at risk of flipping, so the two horizons are converging.
🇺🇸 SNAPLong
Snap
BUY
07/08
buy confirmed
week closed 07/08
Down 2.44% from the signal, with US communications losing 0.90% on the session and sitting among the four sectors in sharp divergence from Friday's snapshot. Historical potential stays the highest of the block but rests on nineteen cases only: a large number drawn from few cases is a hint rather than a measurement.
🇺🇸 AVGOLong
Broadcom
BUY
07/08
buy confirmed
week closed 07/08
Down 2.74% from the signal, and on Wednesday it stood still while the rest of semiconductor stocks rose. Among the technology block this is where the gap is widest between the weekly structure, which holds, and price behaviour, which is not confirming. A signal that has produced nothing after three sessions stays valid while the structure holds, and is worth watching in the meantime.
🇺🇸 CCLLong
Carnival
BUY
07/08
buy confirmed
week closed 07/08
The worst on the card among those listed, down 4.55% from the signal with the daily horizon selling. The cause remains structural: fuel is the first cost line for anyone selling travel, and crude closed its sixth consecutive up session. While that theme runs, this signal works into the wind, and risk management matters more than direction.
How I read it. The number that best describes these three sessions is not in the table: of the 27 new buys, 9 now have the daily horizon disagreeing with the weekly one, against 7 yesterday and a single case on Sunday. A third of the setups born on Friday is already at odds with its own short-term chart, and it is not confined to one sector. The group holding up best is defence, where structure and movement have pointed the same way for three sessions; the one holding up worst is tied to cost variables that are moving, fuel for anyone selling travel and panel prices for solar. On risk management the rule stands as it did on Friday: a signal born on a candle that has already run wants a pullback before full size.
Our model · Newmont long (G / P) · Azimut long (G / P) · Rheinmetall long (G / P) · SPDR Gold Shares long (G / P) · Northrop Grumman long (G / P) · Zoom long (G / P) · NVIDIA long (G / P) · ServiceNow long (G / P) · Palantir long (G / P) · Snap long (G / P) · Broadcom long (G / P) · Carnival long (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
📊
02 · MACRO
The print that broke nothing

In-line inflation and a Fed stepping back: the technical analysis of a postponed session

Wednesday's session turned on a single number, and that number landed where it had to. US July inflation rose 0.1% on the month and 0.2% at the core, taking annual rates to 3.4% and 2.5%, both a tenth below June and in line with expectations. In a normal cycle it would be a non-event; today it is a relief, because the market does not fear a missed cut but an incoming hike.

The consequences were immediate and orderly: September hike odds fell to 38% from 48% the day before and from 70% a month ago, the US ten-year eased for a second consecutive session towards 4.68%, and the dollar weakened. Across the stress complex the move was collective: equity volatility lost 4.8% on the day and 22.5% over four weeks, bond volatility — the only tension measure that had been rising — gave up 7.5% in a single day, and the Italy-Germany spread tightened 3.1%.

The technical analysis I take from this concerns price rather than the print. SPYLong, the S&P 500 ETF, moved less than half a percentage point on the most anticipated day of the month and closed exactly on the lower edge of its congestion band, re-entering it after leaving the day before. An index that breathes less than half its normal range on inflation day has decided nothing: it has postponed, and price action trading reads that as congestion rather than direction. Buy volume recovered to 33.43% from 25.14%, so selling pressure eased without reversing.

How I read it. Our model is long SPY since 6 April, up 13.69% from the signal and with all three profit-taking windows already banked: a mature trade where the useful work is protecting the result rather than adding to it. The concern we wrote about yesterday on the bond side unwound within a single session, and that is worth saying: anyone reading rising rate volatility as the warning of something coming was wrong for twenty-four hours. For swing trading the read is unchanged — with price on the edge of congestion, waiting costs less than a position taken for its own sake.
Our model · SPY long (G / P) · QQQ long (G / P) · US small-cap ETF long (G / P) · SPDR Gold Shares long (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
🧠
03 · AI INFRASTRUCTURE
Two sets of results, a whole sector

CoreWeave and Cisco: results light up AI infrastructure

If the macro removed a risk, results provided the push. CoreWeaveLong posted quarterly revenue of $2.58bn, up 112% year on year, with an adjusted loss of $1.03 per share better than expected, full-year guidance raised to $12.4-13.2bn and a backlog of $104bn: the stock had gained over 13% after Tuesday's close and added 19.28% in Wednesday's session. CiscoLong followed with a record quarter — revenue at $17.3bn, up 18%, adjusted earnings at $1.22, up 23%, AI-related orders up roughly four and a half times — closing 2.86% higher near its highs.

The coda is a reminder that the market is demanding: despite the beat, Cisco stock shed around 4.6% after the close, penalised by a margin outlook weighed down by a more hardware-heavy product mix. It is the pattern that has separated winners from losers in this theme for months: guidance counts, yesterday's numbers do not.

The pull spread across the whole scale of the sector: NebiusShort up 34.14%, Super MicroLong up 19.02%, Dell TechnologiesLong up 9.87%, IRENShort up 9.86%, Credo TechnologyLong up 8.26%, Arista NetworksLong up 6.39%, OracleLong up 5.36%. Three of those moves ended up among the day's outsized ones, meaning they travelled more than twice their own typical swing.

Our model comes into this session with the right positions already open for months. On Dell it is long since February with a 227.19% gain from the signal, on Arista Networks since June with plus 24.06%: two trades the session rewarded without any need to chase. We stay short on Nebius stock and IREN, both of which joined the rally. A sector jumping as a block after results is not noise, it is money coming back: telling a return of money from a rebound is the work an ai trading system does by scanning the whole universe before picking a side, and it is where a trading strategy earns its keep.

How I read it. The technical distinction that matters is between names where the weekly signal was already in place and names where it has only just fired. On CoreWeave, Super Micro, Credo, Oracle and Cisco the buy appeared on the forming candle, so it is provisional and only confirms at Friday's close: treating it as settled means running two sessions ahead of the system. On Dell and Arista the signal had been there for months and the session rewarded it, which is the case where the work was done weeks ago. On Nebius the short is one to review for exit.
Our model · CoreWeave long (G / P) · Cisco long (G / P) · Nebius short (G / P) · Super Micro long (G / P) · Dell Technologies long (G / P) · IREN short (G / P) · Credo Technology long (G / P) · Arista Networks long (G / P) · Oracle long (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
💻
04 · TECHNOLOGY
Half the index up, half down

Technical analysis of big tech: who rises on results and who stays behind

In an index pulled higher by artificial intelligence, the large names outside that theme lost ground, which is why the Nasdaq added less than one per cent on a day like this. MetaShort shed 3.38%, MicrosoftLong 2.26%, AmazonLong 1.83%, and TeslaShort 1.59% despite the calmer mood.

The case that keeps dragging is AlphabetShort, still among the weakest of the mega caps. The real catalyst dates from 5 August — the departure of its chief scientist after twenty-seven years, alongside other senior researchers — but the market continues to discount uncertainty over AI research leadership, concerns about rising data-centre capex, and a regulatory front that will not close, with the distribution-payments question in the background.

Among semiconductor stocks the session marked a return instead: NVIDIALong up 3.03% after weeks of indecision. The rotation rewarded silicon and infrastructure and left advertising and services behind. Worth noting that European technology did the opposite at minus 0.76%: the move came from two American companies' results, so it stays a corporate fact before a sector one.

How I read it. On Alphabet stock our model is short since June with a 4.49% gain from the signal, and it is the best-working position in this block: the stock has been losing ground for two weeks and the sell has tracked it. We are short on Meta too, and Wednesday's 3.38% decline runs with the signal. On Microsoft and Amazon we stay long since late July, with a negative session inside a longer move. On Apple stock the sell fired on the forming candle, so it remains provisional.
Our model · Alphabet short (G / P) · Meta short (G / P) · Microsoft long (G / P) · Amazon long (G / P) · Apple short (G / P) · Tesla short (G / P) · NVIDIA long (G / P) · Broadcom long (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
📉
05 · BRAZIL
A sixth losing session

Brazil's sixth losing session: the technical analysis of a market falling as a block

The day's broadest move came from São Paulo, where the main index fell 2.5% in a sixth consecutive down session. The trigger was the central bank minutes, which pointed to rates staying higher for longer and cooled hopes of a faster cutting cycle, all in an election year. Banks paid the heaviest bill, but the commodity pillars were not spared either: ValeShort lost around 2% and PetrobrasLong 1.3% despite higher crude, both on heavy volume from institutional rebalancing.

There is something on our model worth isolating and stating precisely. The weekly signal on Itaú UnibancoShort, Banco do BrasilShort, ItausaShort, BradescoShort, CosanShort and AmbevShort turned to sell on this week's candle, that is, while the market was falling: those six signals are still forming and only confirm at Friday's close. On ValeShort the sell has instead been active and confirmed since May, up 10.2% from the signal.

One technical detail matters more than the chronicle. Among the day's outsized moves sits MercadoLibreLong at minus 5.76%, listed in the United States but earning most of its revenue in Brazil: when the information is the country, individual names stop telling different stories and geography counts for more than the exchange the stock trades on.

How I read it. Six sell signals forming in the week the market falls for a sixth session is the best outcome this card could deliver, and precisely for that reason it deserves the right caution: the candle closes on Friday and until then those six remain provisional. The technical read is that the model responded to the price structure of preceding sessions rather than to Wednesday's headline. Three of those six report results before that candle closes, and that is the condition where size is decided in advance rather than afterwards.
Our model · Vale short (G / P) · Petrobras long (G / P) · Itaú Unibanco short (G / P) · Banco do Brasil short (G / P) · Itausa short (G / P) · Bradesco short (G / P) · Cosan short (G / P) · Ambev short (G / P) · MercadoLibre long (G / P) · Gerdau long (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
🇮🇹
06 · MILAN
Energy and defence hold

The Milan market: Eni follows crude, Leonardo rises on an upgrade

Milan held up better than the other European markets, and the usual trio did the pushing. EniLong led the board, hooked to crude and to the geopolitical chessboard, with our buy signal active since mid-July and a 3.92% gain from the signal. On defence, LeonardoLong rose after a favourable RBC opinion, also supported by the European rearmament theme: the signal has been a buy since late July and carries plus 8.38%. FincantieriLong also did well, continuing on the subsea consolidation and naval defence theme.

On the banks moves were minimal but the direction is clear: UniCreditLong shaved a gain while confirming itself as the listing's largest by market value, while Intesa SanpaoloLong gave up a few tenths with the MPS deal now expected in the September-October window. Banco BPMLong remains the most advanced trade of the group.

A technical note on the banks, because that is where our position is most exposed. The weekly signal on Intesa Sanpaolo, UniCredit and Banco BPM has been a buy since April, and the three positions are up 21.18%, 18.02% and 33.93% respectively from the signal. Relative strength on the weekly, though, sits above 70 on all three: these are mature trades, where the useful work is protecting the result rather than adding to it. EnelLong is instead the only one of the group still below its entry level, with price just above its own weekly Point of Inversion.

How I read it. On the Italian board our model is long on eight of the nine names mentioned and short only on Stellantis and Prysmian. The two shorts are up 21.44% and 2.70% respectively from the signal, so the Italian picture is where the model is working best of all. What needs watching is not direction but maturity: four of the long names have had their signal active since April with relative strength above 70, and on a trade in that condition every addition buys the part of the move that has already happened.
Our model · Eni long (G / P) · Leonardo long (G / P) · Fincantieri long (G / P) · Intesa Sanpaolo long (G / P) · UniCredit long (G / P) · Banco BPM long (G / P) · Enel long (G / P) · Unipol long (G / P) · Stellantis short (G / P) · Prysmian short (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
🇬🇧
07 · UNITED KINGDOM
Renewables change hands

United Kingdom: Shell sells renewables to TotalEnergies, Rolls-Royce keeps running

The UK basket closed lower for a third consecutive session, weighed down by commodities retreating intraday and by weakness among the heavyweights. The corporate news of the day was a change of hands in renewables: ShellLong sold its European onshore renewables division to France's TotalEnergiesLong, a move that realigns both around their core business. Both closed lower with the oil complex, and on both our weekly signal is a buy: this is the case where the decline runs against the signal rather than confirming it, and the lens flags without concluding.

BPLong fared worse in the same London crude pullback, despite Brent closing higher. Against the trend, Rolls-RoyceLong gained around 2%, extending its very strong 2026 run on defence and aerospace: our signal has been a buy since mid-April with plus 17.96% from the signal, and the stock sits 2.47% below its record with relative strength still under 65 on both horizons — meaning room ahead, which is rare in this phase.

HSBCLong closed slightly higher after results, though the market judged the new share buyback programme too small. In the background, without a catalyst of its own, sits the weakness of AstraZenecaShort, still under pressure after merger speculation and a failed clinical trial earlier this month: there our model has been short since mid-July.

How I read it. Rolls-Royce is the best name in the group and carries plus 17.96% from the signal since mid-April: it has structure, movement and headroom at the same time, and finding all three together is rare in this phase. On AstraZeneca the sell keeps working and is the block's only short. On the three oil names we stay long through a day of crude pulling back: names to keep at the right size until the sector picks a direction again.
Our model · Shell long (G / P) · TotalEnergies long (G / P) · BP long (G / P) · Rolls-Royce long (G / P) · HSBC long (G / P) · AstraZeneca short (G / P) · FTSE 100 long (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
📈
08 · ETFs
The instruments you use to be there

The ETFs that moved the session: the semiconductor ETF, gold ETFs and US technology

Six baskets tell the session. The semiconductor ETFShort semiconductor ETF was the strongest at plus 2.32%, ahead of the broad technology sector, and carries plus 3.00% on the week: it is how you hold silicon without depending on a single name, and tonight it faces its own test with results from the largest equipment maker. There our weekly signal has been a sell since mid-July.

The US technology ETFLong on US technology led the sector map at plus 1.49% and is the most direct way to take the earnings theme without choosing among seven names that rose together; its weekly signal turned to buy on this very candle, so it is provisional. The European technology basket did the opposite at minus 0.76%, and the gap between the two says the move started in America.

Gold ETFs remain the calmest way to take the metal theme: SPDR Gold SharesLong sits among this week's new buy signals, carries plus 1.62% from the signal and has the lowest typical loss of the group at 2.3%; over the past month it has gained 8.81%. The metal closed at record levels, above $4,400 an ounce in the morning, and rises while equity volatility falls: anyone looking for the exposure without the miners' volatility finds it here.

Two Europe-listed baskets complete the picture, and they deserve naming for what they are. The US technology UCITS ETFLong tracks the S&P 500 technology sector in UCITS form and lists in Frankfurt: up 0.79% on the day and 2.96% over the month, it is the European way to hold the very theme that moved Wall Street on Wednesday, without taking a dollar instrument. The European basic resources ETFLong on European basic resources added 0.37% and carries plus 6.73% over the month, and it is the basket closest to the industrial metals theme: copper closed near its record highs on a squeeze between the two main metal exchanges, and this is how we get to be there.

At the other end of the map, the US materials ETFLong on US materials lost 1.24% and was the day's worst, followed by consumer discretionary at minus 1.13% and communications at minus 0.90%. Those are three baskets Friday's report had in good health and the following sessions have contradicted: the badge reads sharp divergence on all three. The semiconductor ETF remains the way to hold silicon without depending on a single name, and it is the sector facing its own test tonight with results from the largest equipment maker.

How I read it. You pick a basket for the opposite reason you pick a stock: on the stock you want the specific, on the basket you buy the theme and pay for dilution. Wednesday is the textbook case — seven technology names rose together with moves between 3% and 34%, and whoever chose the basket took 1.49% without having to guess which one. On gold the reverse applies, as we wrote in the signals card: the basket returns less than the miner, and in exchange it does not ask you to sit through the miner's volatility. On gold the signal has been working for two weeks and on European materials for three, with plus 6.73% over the month: those are the two baskets where theme and signal point the same way.
Our model · semiconductor ETF short (G / P) · US technology ETF long (G / P) · SPDR Gold Shares long (G / P) · US technology UCITS ETF long (G / P) · European basic resources ETF long (G / P) · US materials ETF long (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
📅
09 · EARNINGS
Reporting over the next seven days

This week's earnings: from semiconductors to Walmart

The calendar for the coming days is concentrated at two points. The first lands tonight, after the US close, with Applied MaterialsShort: it is the test of sentiment towards semiconductor equipment makers, the sector that returned to rising on Wednesday, and our model has held it short since late July. The second is Brazilian and arrives in the worst possible week, with utilities and financials reporting while the index falls for a sixth session.

Thursday 13 · Applied MaterialsShort · RWEShort · OndasLong · CPFL EnergiaShort · LightLong · BradesparShort

Friday 14 · CosanShort · PDG RealtyLong

Thursday 20 · WalmartShort

Tuesday 25 · ZoomLong · IntuitLong · AutodeskLong

The name worth pausing on is JD.comLong, reporting today. Our pre-earnings lens assigns it four active clues — improving money flow, rising buy volume, compressing volatility and a pattern of higher lows — and the weekly signal has been a buy since mid-July. Four favourable clues are not a forecast about the result: they say the structure was preparing itself, and structure does not know what the accounts contain.

One point of method we make every time. Our measurement of the July season says entering with an active signal ahead of results returns 2.12% on average against 2.03% for the control group: earnings do not move the average return, they widen the dispersion. They remain a reason to enter smaller, and it is worth remembering the measurement covers a single season.

How I read it. The Brazilian concentration remains the calendar's real risk element, and not because of the individual results. Six of our Brazilian instruments have their weekly signal turning to sell right now and three of them report before that candle closes on Friday: signal and corporate fact arrive together. On the three American names in the window — Zoom, Intuit and Autodesk, all twelve days out — size stays reduced for calendar reasons rather than on merit.
Our model · Applied Materials short (G / P) · JD.com long (G / P) · Walmart short (G / P) · Zoom long (G / P) · Intuit long (G / P) · Autodesk long (G / P) · Cosan short (G / P) · RWE short (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
🇪🇺
10 · EUROPEAN INDICES
Four baskets, one divergence

European indices: three baskets at record, one left behind

The four European baskets we follow all have an active weekly buy signal dating from the first half of April, so eighteen weeks, and all four sit above their own weekly Point of Inversion. That is where the similarity ends. The difference lies in what they have produced and in how far they are from their highs.

FTSE MIBLong remains the best of the four at plus 15.73% from the signal and stands 0.61% below its record. CAC 40Long follows with plus 7.22% and minus 0.89% from its high, DAXLong carries plus 6.32% and sits 0.79% below. Three baskets glued to their peaks, with distance from the weekly Point of Inversion between 2.39% and 2.99%: a thin, even cushion that has narrowed by a few tenths since yesterday, because the European session was negative while America rose.

The fourth is the one that says something. FTSE 100Long on the UK basket has produced plus 1.73% in eighteen weeks, a quarter of the French figure and a ninth of the Italian, and sits 7.96% below its record — ten times the distance of the other three. That is the divergence worth watching: when three indices on one continent are at their highs and the fourth is almost 8% below, the fourth is not taking part in the same move. Its distance from the Point of Inversion, 1.81%, is also the thinnest of the group, and it is the basket most exposed to oil and mining names.

How I read it. The read I take from this concerns management. Three baskets at their highs with a cushion of around three points above the level that judges them are healthy mature trades, where the work is raising protection alongside price. The UK basket is a different case: a signal that has produced under 2% in eighteen weeks with its level less than two points below is the definition of a position that will not take long to resolve, and Wednesday added its third consecutive negative session.
Our model · FTSE MIB long (G / P) · CAC 40 long (G / P) · DAX long (G / P) · FTSE 100 long (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu
📈
11 · SPY AND QQQ
Two baskets converging again

Technical analysis of SPY and QQQ: one re-enters congestion, the other changes side

On Wednesday SPYLong, the S&P 500 ETF, and QQQLong, the Nasdaq 100 ETF, parted ways for the first time after a week of moving together: plus 0.25% the first, plus 0.73% the second, almost triple. The reason lies in composition — SPY weighs 32.91% in technology and 12.59% in financials, QQQ 50.54% and 0.24% — and on a day when technology gains 1.49% and financials 0.21%, that difference in weights is worth exactly the half point of gap.

The technical fact is elsewhere, though, and it is the most important of the week. On QQQ our weekly signal flipped to buy on the forming candle, cancelling the sell active since 20 July: that short closes down 5.77%. It is provisional until Friday's close, and it needs saying that way. On SPY the buy signal has not been interrupted since April: the two baskets are converging again after a month of saying opposite things.

The two maturity readings mirror each other. Setup quality — the index that measures which phase the structure is in, from compressed to stretched — sits at 68.7 on SPY and 32.8 on QQQ: the first is stretched and has banked all three profit-taking windows, the second is compressed and its windows do not exist yet, because they are calculated once the signal confirms. One number cools the enthusiasm though: buy volume on QQQ was 18.01% in the session, meaning more than four fifths of trades came from the selling side on a day that closed higher.

How I read it. On SPY the read is pure management: all targets banked, protection rising, no reason to add. On QQQ the signal has only just fired and should be treated as a candidacy rather than a confirmation, for two reasons that compound — the candle closes on Friday, and the session's buy volume did not accompany the rise. Anyone working on the weekly horizon has two sessions to decide with more information than today offers, and in a phase like this those two sessions cost little.
Our model · SPY long (G / P) · QQQ long (G / P) · US small-cap ETF long (G / P) · US technology ETF long (G / P)
Our model's position on the weekly signal, as of Friday's close. Levels, targets and the trading plan are in each instrument's note.
↑ back to menu

The wrap

The print that was meant to decide and did not

US July inflation landed exactly where the market expected it, and that removed pressure from rates rather than adding to it: September hike odds fell to 38% from 48%, the ten-year eased for a second session and every tension gauge lightened together. Wall Street closed mixed, with the Nasdaq ahead and the large names outside the AI theme losing ground. Crude closed its sixth consecutive up session and gold sits at record levels, so price pressure remains where it was.

For our model it was a good day. The long positions on Dell and Arista Networks, open for months, worked inside the theme that moved the index — the first carries plus 227% from the signal, the second plus 24% — and the sells on Alphabet stock and AstraZeneca kept working. On the Italian board the model is long on eight names out of ten and in profit on every position mentioned. We stay short on Nebius and IREN, both of which joined the compute-infrastructure rally. And six sell signals formed across the Brazilian names in the week the market falls for a sixth session, though they stay provisional until Friday.

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 hit — or how to read chart patterns (double bottoms, Bollinger bands, the Ichimoku cloud, price action), this is the desk’s trading journal: published every morning before the US open, with our model’s position on every stock mentioned. If you are after the concepts explained from scratch — what stops are, how to read a double bottom, what an Ichimoku cloud is — the guides live in the Education section.

The desk's other reports

The full analyses behind the cards on this page.

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