Weekly Note
The S&P rose. Most of the market did not.
Narrow AI leadership, rising yields and oil risk will make a technology-led pullback the most likely outcome for 13 to 17 July, with AMD carrying the clearest single-stock risk.
Technical setup · daily
AMD is testing a crowded rally
Daily price action from 2 January through 10 July 2026. The forecast window contains no outcome data.What the market showed on 10 July
The S&P 500 ended 10 July at 7,575.39 after gaining 1.2% for the week. The Nasdaq Composite gained 1.7%. Those headline numbers looked constructive.
The rest of the market did not confirm them.
| W28 observation | Weekly change | Forecast implication |
|---|---|---|
| Nasdaq Composite | +1.7% | Leadership concentrated in technology |
| S&P 500 | +1.2% | Cap-weighted index remained supported |
| Dow Jones | −0.5% | Broader cyclicals did not participate |
| Russell 2000 | −0.6% | Smaller companies did not confirm risk-on sentiment |
| 10-year Treasury yield | 4.49% → 4.56% | Valuation pressure was increasing, not easing |
| Brent crude | Approximately +5% | Inflation and geopolitical risk remained live |
The divergence was visible inside individual sessions. On 6 July, the S&P 500 rose 0.7% even though most companies in the index declined. By the end of the week, technology and energy had done most of the work while the equal-weight market had lagged.
That divergence will define the following scenario distribution for 13 to 17 July:
| Scenario | Probability | Falsifiable definition |
|---|---|---|
| Technology-led pullback | 55% | S&P 500 falls 1 to 3%; Nasdaq underperforms; Russell 2000 outperforms the S&P relatively |
| Volatile consolidation | 30% | S&P 500 finishes between −1% and +1%; daily leadership continues to rotate |
| Broadening rally | 15% | S&P 500 gains more than 1% and the Russell 2000 outperforms |
Our base case will therefore be a 1 to 3% S&P 500 decline, with a larger loss for the Nasdaq. The 10-year Treasury yield will likely remain inside a wide 4.45 to 4.65% range as potentially softer June inflation data competes with oil risk. A renewed oil shock will be the clearest route to a larger equity decline.
The forecast is not “stocks must fall.” It is that the reward-to-risk balance has turned asymmetric: a small group may still lift the index, but failure by that same group will transmit quickly to the whole benchmark.
Why the outlook is bearish
Three independent signals point in the same direction.
Breadth stopped confirming price
A healthy continuation would normally bring the Dow, small caps and the equal-weight S&P into the advance. Instead, both the Dow and Russell 2000 declined during a positive S&P week.
This matters because the concentration is not passive. AI-linked shares are experiencing unusually large two-way moves. The market will demand more evidence for capital expenditure while relying on those same companies to hold the indexes near their highs.
The implication is simple: if the Nasdaq loses leadership, the S&P will have difficulty replacing it with small-cap or cyclical strength.
AI spending was strong, but expectations were stronger
The fundamental announcements during W28 were substantial. Apple committed more than US$30 billion to Broadcom-related US chip production. Micron increased its planned US investment to more than US$250 billion through 2035. SK hynix completed the second-largest US share sale on record.
These announcements validate AI infrastructure demand. They will not validate every price paid for that demand.
The distinction between a strong industry and a safe entry price is central to the outlook. When positive announcements are required merely to keep a trade stable, the market will have little tolerance for disappointment. The SK hynix listing will also create an immediate test of global demand for an AI leader after an exceptionally strong first half.
AMD carries the most concentrated risk
AMD is where the broader market risk will be most concentrated. The shares closed at US$557.89 on 10 July, up 2.0% on Friday and 7.7% from the previous week’s close. But the weekly gain concealed an unstable path: AMD rose 6.6% on Monday, fell 6.5% on Tuesday and jumped another 5.7% on Thursday. Three moves larger than 5% in four sessions show price discovery under stress rather than calm accumulation.
The fundamentals will remain strong. First-quarter revenue reached US$10.3 billion, up 38% year on year, and AMD guided second-quarter revenue to approximately US$11.2 billion, plus or minus US$300 million. Demand for EPYC processors and Instinct accelerators will continue to support the long-term AI story.
That strength will not guarantee a safe one-week entry. AMD’s next major company catalysts sit beyond the forecast window: Advancing AI 2026 is scheduled for 23 July, while second-quarter results will arrive on 4 August. From 13 to 17 July, the share price will therefore have to defend a near-record valuation without a scheduled fundamental event to reset expectations.
The AMD forecast for 13 to 17 July
| Scenario | Probability | Price definition by 17 July |
|---|---|---|
| Momentum reversal | 55% | AMD falls 8 to 13% and closes between US$485 and US$515 |
| Volatile consolidation | 30% | AMD closes between US$515 and US$570 after wide daily ranges |
| Catalyst anticipation | 15% | AMD closes above US$570 as buyers look through to the 23 July event |
The base case will be an 8 to 13% AMD decline, materially worse than the projected 1 to 3% loss for the S&P 500. A break below US$540 will expose the US$515 area first; failure there will open a path toward US$500. The forecast will be invalidated by a close above US$570 accompanied by broader semiconductor strength rather than an AMD-only squeeze.
AMD’s business can keep growing while its share price falls. With expectations this high, any cooling in AI enthusiasm will hit the stock before it changes the earnings outlook.
Rates and oil were not supporting the rally
The 10-year Treasury yield rose seven basis points during W28 to 4.56%. The June FOMC minutes showed that a few participants saw a case for raising rates, several did not consider the current stance restrictive, and many assessed that the appropriate year-end rate would be above the existing range.
The Fed is not promising an imminent increase. But the minutes will remove the assumption that the next move has to be a cut.
Oil will complicate the picture. Renewed conflict has already pushed Brent higher during W28. Even if the scheduled June inflation data proves soft, another oil increase could raise forward inflation expectations and keep pressure on long-duration technology valuations.
This is why yields will likely be volatile rather than move in one direction. Inflation data will describe the previous month; oil prices will shape the next one.
What will confirm or reject the forecast
The useful prediction will be relative, not clairvoyant:
- Nasdaq risk will be greater than S&P 500 risk because W28 leadership is concentrated.
- AMD will fall 8 to 13% if the AI trade reverses, making it a higher-beta expression of the index forecast.
- Small caps may outperform during a decline because they did not participate in the preceding AI-led gain.
- Soft backward-looking inflation may coexist with high forward-looking oil risk.
- Strong earnings outside technology will not necessarily repair index-level weakness.
The forecast will be confirmed if the S&P 500 falls 1 to 3%, the Nasdaq underperforms and the Russell 2000 performs better on a relative basis. It will be rejected if the S&P gains more than 1% and participation broadens decisively beyond technology.
The forecast will not attempt to predict the precise sequence of geopolitical events, the exact inflation releases or the size of individual semiconductor moves. It will identify the transmission mechanism that could make the market vulnerable when catalysts arrive.
The signal for the week ahead
The W28 headline was bullish, but its structure was fragile. A 1.2% S&P 500 gain accompanied by declines in the Dow and Russell 2000 was not evidence that every part of the market had become expensive or weak. It was evidence that the index had become unusually dependent on one theme.
That distinction will produce a falsifiable forecast. If technology leadership fails, the index will have little support from the parts of the market that already weakened during W28.
The signal should not be treated as infallible. But when the S&P rises while equal-weight, small caps and most sectors fall, and yields rise at the same time, the probability of a technology-led reversal will be materially higher than the headline index suggests.