Five Points From the Line That Matters

The S&P 500 opened Monday at its high for the week and never traded above it again.

That one sentence covers the whole week. Monday's opening price of 7,791 was the top. Every session after it took place underneath. Tuesday gapped down 45 points and never filled the gap. Thursday broke to 7,639. Friday bounced 33 points and finished at 7,674. The week closed down 111 points, or 1.4 percent, and it ended a three week winning streak.

Here is the detail that matters more than the loss. The high set on August 13 at 7,817 has now stood for eight sessions without being challenged. The market did not fail once at that level. It has not come near it since.

And it now sits five points above the single level that decides what happens next.

One Landed On the Nose. One Was Flat Wrong.

Last issue named 7,669 as the first real shelf underneath the market, and said it mattered because it had been the August ceiling before price passed above it. That call landed almost exactly. The market fell to it inside four sessions, pierced it on Thursday at 7,639, and closed Friday at 7,674, five points above. Tested and held, on the close.

The floor call did not work. We named 7,752 as the expected low for the week. Price went through it by 113 points. When a projected floor breaks that decisively, the break itself is the information, and it argues the direction of travel is down rather than sideways.

The gold call was wrong, and it was wrong in the expensive direction. We said 4,454 was the line that mattered, and that below it the 4,288 to 4,343 area was a zone to buy into. Gold cleared 4,454 instead and ran to 4,690, a gain of more than seven percent in three sessions. That pullback zone was never reached. Anyone waiting for it is still waiting, and missed the strongest move on the board.

The euro call was right, and it was the most important of the four. We said 1.1623 was the last of the lower highs that had capped the currency all year, and that clearing it would break the 2026 downtrend rather than pause it. Price cleared it, then cleared 1.1686 as well, and reached 1.1714. Only one level from that entire year long sequence is left above the market.

The Engine That Drove This Rally Has Stopped

For most of July and August, weak economic data pushed stocks higher. The logic was simple. The Federal Reserve is debating a rate increase, not a cut. Soft data made an increase harder to justify, so soft data was relief. The odds of a September increase fell from 57 percent to about 31 percent, and the market rose the whole way down that path.

Last week those odds sat at roughly 30 percent, essentially unchanged. Think about everything that failed to move them. The Fed released minutes showing that more officials wanted to raise rates than the three who formally voted for it. Oil rose almost seven percent. Neither pushed the odds up. Housing starts fell 12.4 percent in a single month, and Walmart posted its slowest sales growth in six years. Neither pushed the odds down.

The repricing is finished. There is very little rate risk left to remove, which means weak data no longer buys the market anything. Last week both good news and bad news were sold.

Underneath that sits a genuinely split economy. Factory activity in the Philadelphia region came in at 47.4 against expectations of 25.0, a five year high, and national factory output reached a four year high, both driven by spending on artificial intelligence equipment. At the same time housing contracted sharply and the largest retailer in the country said the consumer is the weakest it has been since 2020. Those two halves argue for opposite policies at the same time, which is precisely what leaves a central bank stuck.

Then there is the bond market, which may be the most important thing that happened all week. On Wednesday the Treasury announced it would at least double the size of its bond buybacks, from 2 billion dollars to at least 4 billion, in order to push long term rates down. It worked for one day. The 30 year yield fell to 5.19 percent. By Friday the entire move was gone and the yield finished at 5.28 percent, higher than before the announcement. The government tried to lower long term borrowing costs and the market reversed it in 48 hours.

Everything Now Runs Through 7,669

The expected range for this week is 7,733 on the top and 7,581 on the bottom. Friday's close of 7,674 sits 0.8 percent under the ceiling and 1.2 percent over the floor. Both of those levels came down from last week, and the floor fell further than the ceiling did, which tilts the whole range downward. There is now roughly half again as much room below the market as above it, a straight reversal of the week before.

But the number that decides this week is not in that range. It is 7,669, and the market closed five points above it.

That level was the August ceiling. Price rose through it earlier in the month, which turned it into support. Thursday's low went underneath it and Friday's close recovered back above. So it has already been tested once and it held, by seven hundredths of one percent. If it fails on a weekly close, the next references are 7,616, then 7,581, then 7,541, and there is nothing else in between.

Two separate methods point at the same place on the downside. A straightforward measurement of the decline that started on August 13 projects 7,581. The projected floor for this week is also 7,581. When two unrelated calculations land on the same number, that number tends to matter.

Above the market the picture has quietly deteriorated. The zone where this advance has been projected to end has been 7,900 to 8,000 for months. Two weeks ago the weekly ceiling sat inside it. Last week it was just below. This week the ceiling is 7,733, which is 167 points below the bottom of that zone. The gap between where the market is expected to trade and where this advance is supposed to finish has tripled in a week.

The critical reversal line for the larger trend is unchanged at 7,303. It did not rise this week because the market made no new high. But the distance to it shrank from 6.2 percent to 4.8 percent, because price fell toward it instead.

Gold is the strongest chart on the board and also the most stretched. It ran from roughly 4,366 to 4,690 in three sessions, a gain of more than seven percent, and it is up about 17 percent from its late July low. Spot gold cleared 4,600 for a third straight weekly gain. The measured target for this leg sits just above the market at 4,700 to 4,710, which means gold has perhaps 30 points of room left before this move has done what it set out to do. The first real support underneath is 4,506. This is not a level to chase.

The euro finished at 1.1678 after reaching 1.1714. It has cleared every capping level from this year except 1.1797. Below, 1.1571 is the floor that decides whether the breakout holds or gets given back.

The Narrative Catching Up To the Price

The setup was already pointing lower before any headline arrived. Every timing measure now reads topping or negative, the projected floor broke last week, and the ceiling keeps coming down. What geopolitics supplies is not the cause. It is the permission.

Washington is expected to unveil new sanctions on Iran on Monday, after threatening what it called crushing economic measures and dashing hopes of a deal to reopen shipping through the Strait of Hormuz. Oil rose every single session last week to $94.39, up about 13 percent in two weeks, and coverage is now framing $100 as reachable. Rising oil is the one development that hurts both sides of this market at once, because it pressures the consumer that is already weakening while making it harder for the Fed to ease. It is also the clearest reason gold and the euro are both bid against a dollar that fell even on a hawkish Fed readout.

Three Events, Four Days

This is the densest week of the summer, and it front loads on Wednesday. At 8:30 in the morning the Federal Reserve's preferred inflation measure is released at the same moment as the revised reading on second quarter growth and durable goods orders. Three significant numbers, one timestamp. Then after the closing bell on the same day, Nvidia reports earnings. That company is the single stock most tied to the spending that is holding up the entire manufacturing side of this economy.

Friday brings the other half. At 10 in the morning the Federal Reserve chairman delivers his first keynote address in the role at the central bank's annual conference. Markets will be listening for how firmly he commits to the two percent inflation target, with rate odds stuck near 30 percent and no obvious catalyst to move them until he speaks. There is also a technical turn date on the calendar that same day, which is an unusual overlap worth noting. If Wednesday's inflation number runs hot, the rate increase that has been priced out comes back onto the table and 7,669 will not survive it. If it comes in soft, the market has already shown that it no longer rallies on that, which leaves Friday's speech as the only thing left that can change the tone.

Two Paths, One Level

Scenario A. The market holds 7,669 on a closing basis through Wednesday's data. The decline that started on August 13 gets treated as a normal pause inside a strong year, and price works back toward 7,733 and then the gap left behind at 7,744. This is the weaker of the two paths right now, because it requires the market to absorb the inflation print, the growth revision and a major earnings report without breaking a level it is already sitting on. It becomes credible on a close back above 7,697.

Scenario B. The market loses 7,669 on a weekly close. The measured targets are 7,616, then 7,581, and if the selling extends, 7,541. That would be roughly a three percent drawdown from Friday and about four percent from the August high, which is still an ordinary correction rather than anything structural. The larger trend does not come into question until 7,303, which is 4.8 percent below. The evidence favors this path. The projected floor already broke once last week, every timing measure has turned, and Friday's bounce recovered less than a third of Thursday's decline.

The deeper case, if the autumn plays out as the longer term picture suggests, points toward the 7,275 to 7,300 area by mid October. That would still be a correction inside an uptrend, not the start of something worse, and history says the recovery from it runs hard into year end.

What to Do With This

The market has already made its high for this stretch. The question left is whether the decline from it is shallow or real, and one level answers it.

1. 7,669 is the whole week. Watch the weekly close, not the intraday wick. It was pierced Thursday and recovered Friday, so a single dip through it proves nothing.
2. Wednesday at 8:30 is the pivot. Inflation, growth and durable goods land together. A hot inflation number puts the rate increase back on the table and takes 7,669 with it.
3. Do not chase gold up here. It has run more than seven percent in three sessions into a target zone that is 30 points away. The place to be interested is 4,506, not 4,690.
4. The euro breakout is real until 1.1571 fails. It cleared every level that capped it this year except one. Give it room.
5. Watch the 30 year yield at 5.28 percent. The government tried to push it down last week and failed inside two days. If that happens again, it matters more than the Fed meeting.
6. Being patient is a position. Three market moving events land in four days and there is no clean entry anywhere on the board right now. Waiting through Wednesday costs nothing.