Crude closed Monday at 85.44... underneath the two reversals it elected on Thursday. And the timing window opened the same day.

That is the whole situation. Price rallied for three weeks, took out seven daily bullish reversals on the way up, printed the last two at 86.90 and 86.92 on August 21, and then backed off. Now it sits below its own elections while the arrays say a turning point is due right here.

I read this one off screenshots, not the full text report. So the turning point numbers below are bar-height estimates, give or take 6 points, and a few things I normally check were not available to me at all... Fibonacci time targets, the energy model, the indicating ranges. Keep that in mind. That’s how I read it - I don’t know how you chart it.

Full credit where it belongs: the arrays and reversals are from Martin Armstrong’s Socrates platform. The execution framework layered on top is the rule set I’ve written about before.

The big picture

The monthly structure is bearish. June 2026 closed below 76.60, a Major reversal, and nothing since has undone that. The bullish monthly reversals start at 105.20... twenty dollars overhead. Until a month closes above that, every rally in crude is a rally inside a bear structure.

Every rally inside a bear structure is a rally you sell or you skip.

The August run from the 78 area was strong. Seven daily elections from one low is a real cascade. But the rule that adjudicates conflicting elections is blunt about degree: a monthly Major beats a daily cluster, whatever the count, whatever the date. The dailies gave us a tradeable counter-trend leg..... not a new trend.

And Monday’s close at 85.44 put price back below the 86.90 / 86.92 pair it elected three sessions ago. When a market retraces through its own fresh elections right as the timing arrives, I pay attention.

The window

Top down, the way Martin’s arrays are meant to be read:

Monthly. The trading cycle row runs red across August, September and into October. Red at monthly scale means the cycle is counting toward a low. Not this week’s business, but it colors everything below it.

Weekly. The week of August 24 is the tallest bar on the weekly array... this week. A direction-change block covers August 17 through 31. The next weekly panic sits on September 21, which lands in contract-roll territory, so I treat it with suspicion.

Daily. Two tall bars: August 25 and September 7. Direction changes scattered on the 24th, September 1, 4 and 7. One daily panic cycle on September 3. And September 7 carries a yellow trading-cycle bar, the kind where a high and a low are projected on the same day.

So the window is August 24 to September 7. Time is due. The question is only whether price shows up at a level while it’s open.

Right now it hasn’t. The nearest active bullish reversal is well overhead, the nearest active bearish is miles below, and in between there is nothing but the consumed election pair. Price is in the void. The rule about the void is short: no wall, no trade.

Below the line: the exact levels, both scenarios with stops and targets, the honest score, and the Pine script with everything already drawn.

Read more

The daily stochastic printed 86 and crossed down on the same bar.

That combination is the whole post.

The big picture

Some history first, because the levels only make sense with it.

March 2026, Socrates elected ten monthly bullish reversals on Bunge in a single month. Ten. Everything between 119.13 and 126.79 fired at once. I keep a small corpus of these whole-stack election events and it now reads six for six.... when eight or more reversals elect in one bar, it marks the end of the move. Every time. Martin Armstrong’s reversal system is honest about this if you let it be. A market that eats its entire ladder in one sitting has nothing left to eat.

July’s monthly close came in around 106. Below every single one of those ten elected levels.

The week of June 15 the other side confirmed. Five weekly bearish reversals elected, three of them Majors... 121.30, 117.70, 115.00. That shelf is the structure I trade around now. July 29 the dailies capitulated, six bearish in one day. The low printed at 103.37 a week later, and on August 10 a daily bullish Major re-elected at 111.50.

So the map into this week: monthly bullish but spent. Weekly bearish and unrepaired. Daily trying to turn back up. Thursday’s close is the daily leg punching into the bottom of the weekly shelf.

The timing, top down

Full disclosure before the dates. This read is built from array screenshots, not the text report... bar heights estimated by eye, good to maybe plus or minus six TP points, and the convergence score below is capped at 10 of 14 because of it. I’d rather tell you that than pretend a precision I don’t have.

Monthly. The tallest aggregate bar in the next twelve months is January 2027, carrying a direction change and a trading cycle bar that projects a high AND a low in the same month. February ‘27 has the panic cycle. That’s the big window and it’s not tradeable yet. Nearer in, September is the strongest month this quarter.

Weekly. Week of September 14 is the tallest bar on the board, and its trading cycle is green... which in the array language is the ideal timing for a HIGH. The week after carries the volatility spike.

Daily. Monday August 25 is the number one daily bar. Today, August 21, is a direction change day sitting right in front of it... the kind that starts the thrust into a target rather than marking it. September 2 stacks a direction change and a panic cycle on the same session.

Price arrived at the weekly shelf one session in front of the number one daily bar. Time and price met. Everything arrived except the trigger.

Convergence score, disclosed like always: 4 of 14. Low, and I’m publishing anyway, because for once the block has nothing to do with location. Price is ON the levels. The block is the trigger... and below the line I’ll show you exactly which trigger, for both directions.

Everything past this point is levels, stops, targets and the two conditional plans. Paid side of the fence.

Read more

Wednesday’s close was 2.812. The daily bullish Major sits at 2.810.

That is the whole election. Two tenths of a cent, and the level is technically gone.

I am writing this one up because it is a good example of a signal that is real and useless at the same time. The rules say a close strictly above the level elects it. The close was strictly above. So it elected. And I would not put a dollar behind it.

Big picture

August made a new monthly low at 2.616 on the 17th, and that low broke July’s. That matters more than the bounce that followed, because a new monthly low regenerates the whole Monthly Bullish Reversal ladder once the month closes. Same story on the weekly. The report says so directly.... which means every bullish level I am about to list has a shelf life of eight trading days.

Then Tuesday closed 2.774. Wednesday closed 2.812 and printed 2.875 intraday, a four week high.

Here is the part that got my attention. The last key high was 2.830, set back on the 12th. Wednesday went THROUGH it intraday and closed back underneath. High of the day 2.875, settle 2.812. That is a failed extension, and under the offset rule it means the daily cycle high may already be printed, two sessions ahead of where the array says the closing turn is due.

So the market took out the high, and gave it back the same session.

Monthly structure has not changed at all. Cyclical strength is bearish on the report’s own read, long term trend is neutral, short term momentum and trend are neutral. Three neutrals and a bearish is not a bull market. It is a market that stopped falling for a week.

And the monthly array puts the big turn in December, on a maximum turning point bar with a panic cycle and a big red trading cycle sitting under it. Red means the low is due there. Not here.

The convergence window

Top down, the way it has to be run.

Monthly. August turning point is mid tier. September carries the directional change. December carries the maximum bar plus the panic plus the red low window. So the structural event is December, and everything between now and then is a reaction inside it.

Weekly. Week of Aug 24 is the tallest bar in the twelve week window, with a big L-Wave and an elevated empirical read under it. Week of Aug 31 carries a directional change. Week of Sep 28 is the other tall bar, and that one lands on the October contract’s expiry, so part of what the array is drawing there is plumbing, not cycle.

Daily. Aug 21 is the tallest bar in the daily window, and it carries a YELLOW trading cycle. Yellow means a high and a low are both projected on the same unit. Violent whipsaw. Aug 24 carries the big yellow long term bar. Aug 26 is yellow trading cycle again. Aug 27 carries the panic cycle, and Aug 27 is also the September contract’s last trade.

So the window is Aug 21 through Aug 26, apex Aug 24 to 26.

Anything past Aug 26 is a different contract. The September contract expires on the 27th, the roll runs Aug 21 to 26, and volume crosses over somewhere around the 25th. The next array target after that sits on Sep 1 and 2, three converging directional changes, and it is worthless to me on these levels because it executes on October gas. Adding the calendar spread to old levels is not re-anchoring. It is guessing with extra steps.

The indicators, which is where this ends

This is the part people skip.

Stochastic

Daily stochastic is at 73.4 and rising. The rule says do not buy when stochastic is already above 70. So the long is dead on the daily.

Weekly and monthly stochastics are at 14.5 and 9.0. Deeply oversold, which is exactly where you want to be buying.... except %K is still UNDER %D on both. The lines have not crossed. A change of direction is not a cross. No cross, no trigger.

So the timeframe that is allowed to buy has not crossed, and the timeframe that has crossed is not allowed to buy.

For the short side it is the mirror. Above 70 is the high confidence zone for selling, so the daily is in the right neighbourhood. But %K is above %D and climbing. The cross down has not happened.

There is a teaching example I keep coming back to, on CBOT wheat. Time and price meet at a bearish reversal, and the question is whether that is a sell. The answer is no, because the indicators are still long. That exact shape is on my screen right now, on nat gas, in the other direction. Time is arriving. Price is arriving. The indicators are pointing the wrong way for both trades.

I have taken that trade before by ignoring this. It is a named error in my own notes for a reason.

The honest score

Convergence: 6 out of 14. Firm. It goes to 8 only if price actually retests the wall above before Aug 26.

That comes from: the tall weekly bar in the target week, a weekly directional change adjacent, the panic cycle landing within one session of the apex, and price sitting inside 3% of the key monthly reversal. Nothing else scores. No Fibonacci dates, no Elliott targets, no ECM date, because this input set does not carry them.

Verdict: Wait for the fake down

I am not going to leave it at that, because a bare WAIT is not information. Here is what is actually failing and when each one expires.

1. The monthly close. Not resolved until Aug 31. And on that date both the weekly and monthly bullish ladders regenerate off the new low. Levels have an expiry.

2. Election count. One bullish election off the Aug 17 low, and it elected by 0.002. One is not a cascade. This expires the moment a second bullish level elects, or the count resets on a new low.

3. The indicator gate. Covered above. Expires on the Aug 21 close, one way or the other.

4. The contract. Aug 27. Everything after Aug 26 is October gas.

Void discipline. After the 2.810 election, the active walls are 2.905 overhead and 2.720 underneath. Price is at 2.795, which is 3.9% below one and 2.7% above the other. It is testing NEITHER. That is the avoid zone by definition. No entry, no stop reference, no trade.

And before anything else, the risk maths already refuses it. Selling here against a stop above the overhead wall gives me 0.38 to 1 on the first objective. The rule is 1.5 to 1 minimum. The trade fails on arithmetic before it ever reaches the gates.

So no long at all

Not today. That is not the same as never, and the difference is the only part of this post I would actually act on.

What blocks it today is one number. Daily stochastic at 73.4. Above 70 is a hard abort for a buy, and it is not a preference, it is the single most common way I have lost money in this system. Buying an already extended oscillator. On top of that price is 6.8% off the low. The buy was Aug 17 at 2.616.... and on Aug 17 the monthly close was unresolved, the election count was zero, and none of the stochastics had crossed. The system said no that day too. Consistent, and annoying.

Now look at the table again, at the two rows nobody reads.

Weekly stochastic 14.5 against 15.7. Monthly 9.0 against 11.2. Both deeply oversold. Both with momentum ALREADY crossed up underneath. Both sitting roughly one bar away from %K crossing %D.

That is a completely different trade from anything on offer this week. Buying an oversold cross at weekly and monthly degree is not the same animal as buying a daily bounce into overhead supply, and it is worth more than every level in this post.

It has not crossed. Friday’s close is when I find out.

And then there is December, which is where the monthly array actually points. Maximum turning point bar, panic cycle, big red trading cycle underneath. Red is the low. If the model is right about that then the long that matters is four months out and everything between here and there is positioning.

So the short answer is that the long side is one weekly stochastic cross away from being interesting, and it is not that cross yet.

That is how I read it. I do not know how you chart it.

Read more

Gold printed $4,509.1 late Wednesday and backed off.

The daily reversal on my sheet is $4,509. The number came out of Martin Armstrong’s Socrates computer before the market ever went there. Price stopped a tenth of a dollar past it and gave back forty bucks into the close.

I didn’t trade it. That probably sounds wrong, so here’s the working.

The big picture

Gold topped in January at $5,627. Five months down into $3,955 on June 30th, then a five week bounce. Through that whole 29% decline the monthly level never elected a single bearish reversal.... on the monthly, the bull market never broke. That’s the structure I have to respect no matter what the daily chart is doing to my nerves.

Then the strange part. Socrates flagged what it calls a rare superposition event : a long-term weekly bullish reversal elected right into the August 3rd high, negated by an intermediate bearish election a week later. Rule 4 - the later election consumes the earlier one. A bullish signal that fires into a high and dies the next week reads as exhaustion to me. That’s how I read it... I don’t know how you chart it.

One more thing that bothers me. This rally is a dollar story. Priced in a basket of currencies, gold peaked back on June 2nd. In dollars it peaked Wednesday night. A market making new highs in one currency only is thinner than the chart looks.

Quick housekeeping before the numbers : none of this is advice. I trade my own book and publish the working, wrong parts included. Full disclaimer at the bottom.

The window

August is the strongest bar on Martin’s monthly array, and the machine printed that bar back when this was still a falling market. Inside August, the week of the 17th. Inside that week, Friday the 21st, with a directional change opening the window on Monday. A panic cycle sits two sessions behind it, on the 25th.

So Wednesday’s tick-perfect touch came on the wrong date. Price arrived at the level. My oscillators crossed down from deep overbought, 91.5 on the daily. Time didn’t show up..... two legs out of three. The rule I trade by needs all three, and the one mistake I keep paying for is entering on two. Last time it was time and price without the cross. This time it’s price and the cross without time. Same tuition, different classroom.

What I’m watching

• Friday’s weekly close vs $4,432.3

• A daily close under $4,415.7 ... opens the path to the wall

• The 200-day at $4,497.7, still untested from below

• $4,578, the first weekly bullish reversal. A weekly close above it ends the bear case entirely

• My other open windows in the same August-September cluster. Correlated setups get one aggregate cap, not five separate ones

• or gold gaps over the whole shelf Monday morning and this post becomes a museum piece ?!

Everything past this line is the executable part : entries, stops, targets, and the second wall nobody is looking at.

Read more

Sunday night I pulled the Socrates arrays for NY Crude off screenshots... no text report this time, just the visual tabs. Ran the digest, wrote down the gate: short below 78.00. By Tuesday’s close the September contract printed 74.30.

The gate fired while I was still reading the report.

Normally that’s the good problem. This time there’s a catch, and I’d rather tell you the catch before the setup: the September contract stops trading around August 20. The strongest timing in this whole read lands AFTER that date. So the big trade... the one the monthly arrays are actually pointing at... isn’t takeable on these levels. What’s left is a smaller pre-roll trade, and an honest score that doesn’t flatter it.

That’s how I read it. Here’s the working.

Big Picture

June closed below 76.60 and elected the Monthly Major 4 bearish. That’s the whole story on the monthly level.

Yes, March and April elected a stack of monthly bullish reversals, some of them Major Extended. Doesn’t matter anymore. When both sides elect, the later Major nullifies the earlier one.... the rule Martin’s system runs on is degree first, then count, then date. The June Major 4 came after the April bullish block, so the entire 83.40 to 104.82 bullish ladder is consumed. Not support. Not targets. Gone.

What would flip this back? A MONTHLY close above 105.20. That’s 41% above Tuesday’s close. Until then the structure is bearish and rallies are for selling.

The monthly timing array says August wants a low, and October is the bar I actually care about... the Long Term row maxes out there with the Trading Cycle calling for a low. The destination below is ugly and specific: after the daily levels run out at 67.03, there is NOTHING until the weekly shelf at 61.70 to 60.00, and below that the monthly Majors at 57.10 to 54.70. Martin’s reversal density rule says wide gaps travel fast once breached. That gap is nine points wide.

The Convergence Window (Monthly → Weekly → Daily)

Cascading down:

Monthly: Aug 2026 (TP ~45, low due) near-term... Oct 2026 structurally (TP ~40, but the highest Long Term confluence in the whole 12-month window).

Weekly: week of Aug 24 carries a TP ~51 with a Direction Change. Week of Oct 12 is the tallest bar at ~52. Both strong.

Daily: Aug 4 was the tallest daily bar (TP ~52) with a Direction Change on Aug 5. The market fell INTO it and closed on the low. When price falls into a high bar, the bar usually marks the low of the reaction, and the bounce starts from there.

Now the catch, in full. All of this was read off images, not the text report... so the TP numbers are bar-height estimates, good to maybe ±6, and the score below is capped at 10 of 14 because six scoring inputs (Fibonacci dates, Elliott dates, ECM, the indicator grid) simply aren’t visible in screenshots. And the week of Aug 24, the strongest near-term weekly bar, sits AFTER the September contract’s last trade date. Different contract. These levels don’t carry over, and no, adding the calendar spread to old levels doesn’t fix it. Only a fresh report on the October or December contract does.

Honest score: 3 to 5 out of 14 for this week’s window. The Aug 24 window scores 4 but it’s untradeable on this contract. Neither clears my own entry floor on convergence alone. What carries the remaining trade is structure: an elected level owed a retest, and all three timeframes pointing the same way on the oscillators. I’ll show you exactly what I mean below the line... but the score is the score, and I’m not dressing it up.

For the big September-October move: standing aside until I re-anchor on the next contract. Right direction, wrong contract is still the wrong trade.

Below the line: four setups... the one I’m arming, the one the rules block, the one I rejected and why, and the placeholder for the real trade... with entries, stops, targets, R:R, and the kill conditions that override all of it.

Read more

https://www.tradingview.com/chart/OXY/EAnhGypT-Ready-for-a-bounce/

Start with the bar that decided everything, because it happened five weeks ago and almost nobody logged it.

On June 30, Occidental traded down to $47.77.

Sitting at $48.30 was a Major monthly Bearish Reversal. Price went through it by fifty-three cents intraday, spent a few sessions down there, and then turned. July closed at $57.07.

Eighteen percent above the level it had supposedly broken.

So the June decline hit the wall, failed to take it, and reversed 23.5% off the low.

That was the trade. It's over. What I want to talk about is why the setup sitting in front of us right now — with a genuinely strong structural read behind it — is still a wait.

The Big Picture

June was a clean cascade down. Four consecutive weekly Bearish Reversals elected, one per week:

| Week of | Level elected |

|---|---|

| Jun 08 | 56.60 · Major |

| Jun 15 | 52.20 |

| Jun 22 | 51.80 · Major |

| Jun 29 | 49.70 · Major |

Four weekly elections in four weeks is a real cascade, not a wobble. And a cascade has a destination. This one's destination was the monthly shelf at 48.30 — and it got there, and it couldn't close through.

Here's the part I got wrong on the first pass, and it's worth being blunt about it, because it changes the read.

When I first ran the elections I treated that June weekly bearish sequence as the structure still in charge. It isn't. Once a reversal is executed, its validity as a signal ends. All four of those June levels are spent. They were targets, they were reached, they're consumed. What actually governs now is what happened at the wall they ran into — and the wall held.

Then the market did the other half of the job. On July 21 and 22, ten daily Bullish Reversals elected off the low, a mix of Major and Extended, stacked between $55.35 and $57.40. Ten elections from a single low is a strong signal by any measure.

But look at what has not happened.

Zero weekly Bullish Reversals have elected from that low. The nearest one sits at 59.20, untouched. Zero monthly bullish — nearest is 65.20, and that's another world away. The entire bullish cascade is daily-degree only, and the sequence that's supposed to follow a daily turn is: dailies elect, then weeklies begin to elect in the same direction, then the monthlies come into play.

Step two hasn't happened. That's the whole hesitation.

Worth noting on the other side: the monthly bullish cluster from March 2026 — four Majors between 58.69 and 62.00 — is also consumed. Price elected the lot in March, then round-tripped all the way back through them. Those levels are history, not support.

The Convergence Window

Top down, monthly to weekly to daily.

Monthly. September 2026 is the target month. The Aggregate bar is at 96% of the array's maximum height, it carries a Direction Change, and the Long-Term row prints a tall yellow bar underneath it. August, by contrast, is the quietest month in the whole array — low Aggregate, no Direction Change, no Panic Cycle. August is a drift month. September is where the model expects something to resolve.

One flag for the longer horizon: the single biggest bar in the entire twelve-month monthly array is March 2027, and it carries a yellow Trading Cycle — meaning a bullish count and a bearish count converge on the same month. Yellow is the violent one. Both a high and a low projected. That's eight months out and I'm not trading it today, but it's on the wall.

Weekly. This is where the shape lives.

Week of Aug 17 — the array maximum. Trading Cycle prints green: the ideal time for a HIGH.

Week of Aug 24 — near-max Aggregate. Trading Cycle prints red: the ideal time for a LOW.

Week of Aug 31 — carries the weekly Direction Change, and Aug 31 is also the monthly close.

Week of Sep 7 — Trading Cycle prints red again, the strongest red in the array.

Read in sequence, the weekly array is describing a rally into mid-August, a decline into late August and early September, and a low somewhere in that red zone. Then green again into mid-September and the first week of October, where a Panic Cycle also sits.

Daily. And here's the problem.

The biggest daily bar in the array was August 3 — yesterday. Highest Aggregate in the window, tallest Long-Term bar, and a yellowTrading Cycle, same signature as March 2027 in miniature.

Monday's range was 54.86 to 56.13. The nearest reversal in either direction was 53.30, a dollar and a half below the session low.

Time showed up. Price didn't.

The Honest Score

6 out of 10.Moderate. WAIT

The cap is 10, not 14, because this read was built from screenshot arrays rather than a Premium text report. Bar-height estimation gets you turning-point scores to roughly plus-or-minus six points, and it gets you nothing at all on Fibonacci time targets, Elliott Wave dates, the Economic Confidence Model, the nine indicating ranges, or the pivot ladder. Four of the fourteen scoring lines are structurally unavailable. I'd rather publish the ceiling than pretend the number is out of fourteen.

What earned points: the monthly turning-point score in September, the weekly scores in the Aug 24 and Sep 7 windows, the monthly Direction Change in September, and the weekly Direction Change at Aug 31.

What earned nothing, and why it matters more than what earned something:

The void.Price closed Monday at 55.47. The nearest reversal below is 53.30. The nearest above is 58.30. The midpoint of those two is 55.80. Price is sitting 0.6% off the exact center of the rangewith no wall to lean on in either direction. Step out to the monthly and it's worse — 13% down to the nearest monthly bearish, 17% up to the nearest monthly bullish. On the monthly timeframe this stock is in open water.

The oscillators. Three timeframes, and they don't agree.

| Timeframe | Stochastic 13,3,3 | Momentum vs 3MA | Read |

|---|---|---|---|

| Monthly | %K 52.0 below %D 54.6 | −9.53 vs −0.79, both negative | Down |

| Weekly | %K 65.7 above %D 61.2, rising | 6.56 vs 6.37, positive | Up |

| Daily | %K 51.1 above %D 50.2 | 0.54 vs −0.45 | Up, barely |

Two of three point up. The monthly points down. And that daily spread is nine-tenths of a point with the lines converging, which is a change of direction, not a cross. The distinction is not academic — trading a turn on lines that changed direction without crossing is the single most common way a good structural read turns into a loss.

The rule I keep coming back to on setups like this one is the plainest in the whole framework:

> If price reaches the reversal on the array date, buy or sell. Otherwise do nothing.

Time alone is not a trade. Price alone is a coin flip with a slight edge. Both together is where the win rate lives. Yesterday delivered time without price, and the correct response to that is to write it down and wait for the next date.

🔒 For paid subscribers

Everything above is the read. Everything below is what I'd actually do about it — the levels, the stops, the targets, the position sizing, and the full indicator file.

Read more

Soybeans ran a 117 cent range in July and closed in the middle of it.

Today is the last trading day of the month. It is also a Friday. The daily array has a panic sitting on it. One print this afternoon closes the day, the week and the month at the same time, and Martin’s timing says this is the bar.

I’m doing nothing.

Here is what lined up, because it is worth seeing how close this got.

The array first. July 31 is one of the three tallest turning point bars in the daily window. Panic cycle on the same bar. The weekly closes here and the monthly closes here, so whatever prints at settlement outranks anything the daily does .

Then price. Nearest monthly reversal above the market is 1205.00. Beans are at 1193.75. That is 0.94% away, which is inside the 1% band I use, on the exact day the timing is due.

Price arriving at the reversal on the array date. That is the setup. It is the one you sit through months of nothing waiting for.

Then I opened the oscillators.

Daily stochastic: 10.2 . Buried.

Weekly: 77.7.

Monthly: 73.4, and the lines have already crossed down.

The rule says do not sell into a stochastic under 30. Same rule says do not buy into one over 70. So the daily forbids selling and the weekly forbids buying, and there is no side of this market I am allowed to stand on.

Time and price meeting is the SETUP.

The indicators are the TRIGGER.

They are two different things and I have paid real money for treating them as one.

I did it in wheat. Array date was due. Price was sitting on the reversal. Both of the things I check had checked out, so I took it. The oscillators had not crossed and I did not wait for them. Market went straight through the level and kept going.

That was me deciding the setup looked good enough to skip the last step.

Now the part where I tell you where my own read is soft, because you should not take any of this at face value.

The report I worked off was screenshots. No indicating ranges, no ECM dates. Every turning point number I have is a bar height I ESTIMATED off an image, give or take six either way. My convergence score is capped at 10 out of 14 for that reason before I score anything at all, and it came in at 7.

I also had the price wrong. I ran the whole cascade off 1192.60, because that is what the monthly bar readout said when I built it. Actual is 1193.75. A cent and a bit..... except that cent and a bit is what moved 1205.00 from OUTSIDE my 1% band to INSIDE it, and that band is a thing I actually trade off. So pull your own price before you use any level in here.

And the contract. Socrates runs grains on a nearest series and does not tell you which month it is sitting on. August beans expire on the 14th. The genuinely interesting timing on this report is late September into October.... which is on the other side of the roll. The DATES survive a roll. The PRICES do not. Anyone reading those October levels off an August chart is trading a ghost and will not find out until it costs something.

So: panic cycle, monthly close, three timeframes resolving on one print, price under 1% from the wall, 7 out of 14.

And I am sitting on my hands.

Somebody helped me see the wheat thing for what it was, which is the only reason I caught this one before I clicked. That is how I read it . I don’t know how you chart it.

Ask me Monday if I still feel clever about it.

Timing arrays and reversal levels from Martin Armstrong’s Socrates. Estimates from a visual report, not the text feed, so treat every number in here as approximate. Nothing here is advice and none of it is a recommendation to buy or sell anything.

Two numbers from this month have been sitting on my desk all week.

July’s high in the S&P 500 E-mini was 7,632.00. The unelected bullish reversal it was running at sat at 7,632.30.

It missed by thirty cents.

July’s low was 7,324.00. The bearish reversal underneath it — a Double, the kind Socrates only generates a few times a year on a daily chart — sits at 7,324.60.

It missed by sixty cents.

A 308-point monthly range, and both walls were mapped to under a single point before the month began. Not a zone. Not a “region of interest.” Two numbers, printed in advance, and the market walked up to each one, touched it, and turned around.

I want to be honest with you about what that does and doesn’t mean. Because the temptation — and I’ve paid for this lesson more than once — is to read that kind of accuracy as an invitation.

It isn’t. Precision tells you where. It never tells you when, and it never tells you go.

The big picture

Zoom out and the structure is still a bull market that has simply gotten tired.

The move off the April 2025 low at 4,940.25 ran thirteen months and put on roughly 54%. Every monthly bullish reversal from the March low was elected. On Armstrong’s monthly indicating ranges, all nine indicators are still reading bullish. That is not a broken market.

But July is closing down 2.26%, and it’s the first down month of the run. The monthly stochastic crossed down from 93.9 — about as overbought as this chart has ever printed. Monthly momentum crossed down with it. Both of the readings I care most about have rolled over at the same time, from an extreme.

Here’s the part most people will get wrong this week, so hold onto it:

The nearest monthly bearish reversal is 6,814.40. That is 7.6% below where you’re sitting.

July cannot elect it. Nothing about Friday’s close can elect it. Which means the monthly uptrend is not breaking — the indicators are just unwinding a very stretched condition. What you’re looking at is a reaction inside a trend, not a trend change. Those two things demand completely different trades, and confusing them is how people give back a year of gains defending a short.

The convergence window

I run these top-down, always. Monthly first, then weekly, then daily. Never the other way around.

Monthly. August carries a turning-point score of 40. October carries 40 and a Panic Cycle. The dominant target in the whole array is March 2027 at 60 — far out, but worth writing down now.

Weekly. The strongest weekly target is the week of September 7. I’ll flag why I trust this one: I read it independently in a Socrates Premium text report dated June 29, and then again in a fresh set of arrays pulled this week. A month apart, two different formats, same answer. A weekly Elliott Wave target lands on 09/07 as well.

One catch you need to know about if you trade the front month: September 7 falls inside the ESU roll window. The September contract expires the 18th, roll runs roughly the 10th through the 16th, and quarterly OpEx sits on top of it. The biggest weekly target on the board is anchored to a contract that won’t be front-month much longer. I’ll re-anchor that one on a fresh report in late August rather than carry these levels into it.

Daily. This is where it gets interesting.

Thu Jul 30 — red Trading Cycle. Ideal time for a low. Highest internal volatility bar in the window.

Tue Aug 4 — tallest Aggregate bar in the array, plus a Directional Change, plus a heavy Long-Term reading.

Thu Aug 6 — Directional Change and Panic Cycle together.

Fri Aug 7 and Wed Aug 12 — green Trading Cycle. Ideal time for a high.

Read as a sequence rather than as isolated dates, the array is drawing a low into month-end, a thrust into the 4th, a range-expansion event on the 6th, and a high candidate the following week.

But none of that is the actual event.

Friday is the whole trade

Friday July 31 is a daily close, a weekly close, and a month-end close landing on the same print.

When timeframes collide like that, the higher one governs. A daily close means very little on the last trading day of a month — the monthly close supersedes it. So every conditional I’m carrying resolves at one moment, and nothing before then is actionable.

Where does that leave price? Look at how it’s sitting:

```

7,398.00 ← lowest elected bearish — invalidation

↑ 21.25 points

7,376.75 ← Wednesday’s close

↓ 21.75 points

7,355.00 ← top of the unelected support cluster

7,336.10

7,324.60 ← the DOUBLE that held

Twenty-one points up. Twenty-one points down. Price is within half a point of the exact center of its own bracket.

That is the textbook description of the void — sitting between the walls, testing neither. The rule I trade by is unambiguous about it: no entry between walls, no stop reference, no trade. You wait for price to reach a wall before you evaluate anything.

Note on my inputs: this week’s arrays and reversal tables came from platform screenshots rather than a full Premium text report, which caps how much I can verify. More on exactly what that costs below.

Read more

Here is the bar that matters.

VIX opened at 19.05 on Tuesday, ran to 19.52, and closed at 18.21 — in the bottom fifth of its own range. Down 2.46%.

That 19.52 print did something specific. It pushed through an unelected bullish reversal sitting at 19.22 — and then the market closed a full point beneath it.

In this framework, an intraday poke through a level means nothing. Only the close elects. So the wall held, the signal fired in the opposite direction, and by the letter of the model, Tuesday said short volatility.

I’m not shorting it. Neither should you, and the reason is the entire point of this post.

A signal firing and a trade being takeable are two different tests. Tuesday passed the first one and failed the second one three separate ways. Learning to tell those apart is worth more than any single setup I could hand you.

The Big Picture

Start with what the market has actually done, not what it looks like it’s doing.

Since the July 10 low at 14.96, VIX ran nine sessions to 20.31 on July 23 — a 36% spike. It has since bled back to 18.21. On a chart that reads as a volatility scare that’s cooling off.

Underneath, the structure is more decided than that.

The monthly is bearish and it’s been bearish since May. A Major bearish reversal elected at 15.60 in the month of May 2026. The only monthly bullish election on the board fired at 22.00 back in March — and it isn’t a Major.

That ordering matters more than it looks. When a later Major election fires against an earlier non-Major one, the earlier one is consumed. It stops counting as a target, a stop, or a direction signal. The March bull case didn’t weaken. It was cancelled.

The weekly repeated the pattern. A Major bearish reversal elected at 15.10 in the week of July 6. Every weekly bullish election on record — 27.36, 28.20, 29.60, 30.33 — dates to March 2026. Four months stale, and sitting on the wrong side of a later Major.

So the July rally isn’t a regime change. It’s a counter-trend bounce inside a structure that already picked a side.

The momentum picture says the same thing in a different language:

Timeframe Momentum Signal Reading Monthly −1.65 −0.76 Short Weekly −0.20 +1.02 Short Daily +1.16 +0.33 Long

Monthly and weekly agree with the structure. The daily hasn’t crossed. After a −2.46% session, daily momentum is still above its signal line — and at +1.16 on a panel that spans roughly ±2.5, it’s barely off flat.

That’s not a short. That’s a market that fell and hasn’t confirmed it meant it.

The Convergence Window

The Socrates arrays are why July 28 was circled in the first place. Working top-down, which is the only way this works:

Monthly — July 2026. The Aggregate row sits in its tallest tier, tied with November and March. Underneath it: a big blue Empirical bar and a big yellow Long Term bar. Yellow Long Term carries the heaviest weight in the array, and July has it.

August is the interesting neighbour. August carries the Direction Change. The extreme prints in the tall month; the decisive move ignites at the DC. That’s the classic pairing, and it puts the turn in July and the follow-through in August.

Weekly — week of July 27. Tallest tier on the Aggregate, and it carries both a Direction Change and a Panic Cycle in the same week. That combination is uncommon. It’s also the week we’re in right now, closing Friday.

Daily — July 28 and August 3. Tuesday scored 51 on the daily Turning Points row, second-highest in the twelve-session window, and it landed exactly on a Fibonacci time target projected from the July 23 high.

The tallest bar in the entire daily array is Monday August 3 at 57 — and it holds the only Panic Cycle in the window.

One thing to flag about August 3 before anyone circles it. Scored on its own cascade path it’s much weaker, because August’s monthly Aggregate is mid-tier and the week of August 3 is unremarkable. A tall daily bar sitting inside a quiet week inside a non-target month doesn’t outrank a loaded week under a heavy month. August 3 is where the move gets violent. It isn’t where the model points.

There’s one more structural fact worth sitting with: Friday July 31 is both the weekly close and the monthly close. Higher-timeframe closes supersede daily closes for anything structural. Whatever happens Wednesday and Thursday is provisional until that Friday print lands.

The Honest Score

I publish this whether or not it flatters the setup.

Convergence: 8 out of 14. HIGH confidence — at the very bottom of the band.

What earned points: the tallest-tier monthly Aggregate with big yellow Long Term, the tallest-tier weekly Aggregate, the verified daily score of 51, the weekly Direction Change, the adjacent monthly Direction Change, and the Fibonacci time target landing precisely on July 28.

What didn’t: no Direction Change on Tuesday itself, the Panic Cycle sits four sessions away, no Elliott Wave or ECM data in this report, and — the big one — the market is nowhere near a monthly reversal. The nearest monthly wall above is 25.32, roughly 39% away. The nearest below is 14.30, about 21% away.

Four of those eight points rest on bar-height estimates rather than published numbers. If the monthly estimate misses its threshold, this is a 6 and the label drops to MODERATE.

Verdict: STAND ASIDE. Not “wait,” not “watch closely, might scale in.” No position.

Three reasons, and they’re independent of each other:

One — the monthly close hasn’t happened. July is still open until Friday. Any read taken now is conditional by definition, and the framework doesn’t permit entries on conditional bias. Gate A fails on a technicality that isn’t a technicality.

Two — zero elections. From the July 23 high at 20.31, exactly zero bearish reversals have elected. The first one doesn’t come into play until 16.02, which is 12% below spot. A directional signal with no elections behind it is a forecast, not a cascade.

Here’s the part that surprised me. Measured from the July 10 low instead, the bullish side has four elections — a strong reading. So the direction with cascade support is long, and the direction with structural support is short. They point opposite ways. When your own gates disagree with each other about which way to face, that is the market telling you it hasn’t decided.

Three — the void. At 18.21 the market sits 5.5% below the wall at 19.22 and 12% above the wall at 16.02. It isn’t testing anything. The test happened at 19.52 intraday and it’s over.

Entering between walls has no stop reference, which means it has no defined risk, which means it isn’t a trade — it’s a position with a hope attached. Rule 4 exists to prevent exactly this.

And there’s a fourth thing, more subtle and more useful than the other three.

Tuesday spiked to an intraday high through the reversal and then closed near its low. Intraday says one story, the close says another. When those two disagree on an array target date, the framework is explicit: the close is the authority, and the setup is void. You want the intraday extreme and the close arriving together — both high, or both low. Split them and you have a failed signal wearing the costume of a good one.

Mode C Notice

Transparency on inputs, because it changes what the numbers are worth.

The daily section of this analysis came from full Premium report text — those figures are exact. The weekly and monthly sections came from chart images only. That means the weekly and monthly turning-point scores are bar-height estimates, accurate to roughly ±6 points, and several signals simply weren’t available: weekly and monthly indicating ranges, weekly Fibonacci retracements, Elliott Wave targets, ECM dates, and higher-timeframe pivots.

Image-only input caps the maximum achievable score at 10 out of 14. This one scored 8, so the cap didn’t bind — but the two monthly points are the most fragile in the tally.

One known bias worth naming: this reading method systematically under-estimates tall weekly clusters. Both co-tallest weekly bars were leaned to the high end to compensate.

🔒 The Trade Plan — Paid Subscribers

Everything above tells you why there’s no position today. Everything below tells you exactly where the position comes from, at what price, with what stop, and what each one is actually worth.

Three setups are on the board. One of them is the trade. One of them is a trap that looks like a trade. One of them is a long that most people reading this wouldn’t think to consider.

Read more

I spent last night building a pine script for natural gas. Reversal arrays, timing, the lot.

The contract it’s built on stops trading Wednesday.

That’s most of the post right there. But the way this setup falls apart is more useful than most of the ones that work, so here’s the working.

Big picture

Martin’s report came out Friday with gas at 2.888. By Monday’s close it was 2.743.

Five percent. Over a weekend.

Every daily bearish reversal on that report is elected now. 2.845, 2.800, 2.795, 2.755, 2.750 .... all five, gone in a session and a half. The shelf I was watching as support is above us.

What’s underneath is one number. 2.720. Monthly, major.

Price is sitting about 1% over it.

The convergence window

Top down.

Monthly. The tallest bar in the array is December, with a panic on it and a red trading cycle. That’s the structural low and it’s five months out, so it isn’t this week’s problem. What July has is a direction change and a YELLOW trading cycle ... the both-directions-at-once bar, whipsaw by construction. Which is what we got.

Weekly. Week of August 3rd is top tier, tied with the two September weeks. Mode C underreads clusters in that tier, it always has, so I’m reading the whole band hot rather than cold.

Daily. July 31st carries the biggest bar in the array. It’s also a Friday. It’s also the last trading day of the month.

Three closes land on one date, and the monthly one is the only one that counts.

So price arrives at a reversal on the array date, and the array date is the date the monthly resolves. Time and price meeting. That’s the whole thing.

Except.

Where it breaks

The daily stochastic is at 6.2.

Six point two. The rule I trade says don’t short into an oversold stochastic and 30 is the line. This is 6.

The monthly is stranger. Stochastic at 12.3, momentum sitting ABOVE its signal since April, price printing new lows the whole time. Price down, momentum up. That’s a divergence on the monthly and I’ve not seen one of those resolve quietly.

So the array says short and the panel says don’t.

Everyone reading this report is going to want to sell 2.74 on Friday. I’m not going to.

Three bullish crosses on the daily stochastic in two weeks, the 14th, the 16th and the 21st, and all three failed. That’s a push. You don’t fade a push, you wait for it to run out of road.

The trap

This is the part I’d want somebody to tell me.

August last trades Wednesday the 29th. The decision date is Friday the 31st.

Friday’s monthly close prints on September.

September gas trades under August. I’ve seen 1% and I’ve seen 2% depending on the day. Call it one and a half.

The gap between price and 2.720 right now is 1%.

So the roll on its own can carry a continuous chart through 2.720 on Friday. No seller involved. Your chart will show a monthly major bearish election that never happened, on the exact date the setup resolves, and it will point you into a short with the stochastic at 6.

Read Friday’s close on the August contract, or re-anchor 2.720 to September before you look at it. Not off a continuous chart. This is the splice problem and the calendar has arranged for it to land on the one day that matters.

The plan

Two branches. Both of them wait for Friday.

July closes above 2.720. That’s a non-election of a monthly major on a month-end array date with the stochastic under 20, and under 20 is where the confidence lives for a long.

Entry on confirmation rather than on the first signal, because of the push ... I want a daily close above the high of the low bar before I touch it. Stop 2.719 on a close. First target 2.950, the daily major overhead. Then 3.140, then the cluster at 3.265 / 3.270 / 3.305.

Roughly six to one on the first target. Half size, counter-trend against a monthly bear regime, and counter-trend gets half.

Worth knowing: 6.4% of nothing between 2.950 and 3.140. No levels in there at all. Gaps like that travel fast once the first one goes.

July closes below 2.720. The monthly major elects, and the ladder underneath is 2.570, then 2.510, then a void down to 2.245.

I still can’t take it on Friday. Stochastic at 6 is stochastic at 6. That short only becomes available on a bounce that lifts the daily back over 30 and then crosses down again at the retest of 2.720. And if the close is more than 1% through, under 2.693, it’s a retest trade or it’s nothing.

Honest score

6 out of 14. Moderate.

Mode C caps me at 10, so read that as 6 out of a possible 10. I ran this one off screenshots. No fibonacci dates, no elliott targets, no ECM, no indicating ranges, no pivots, no breakout or crash numbers. The TP figures are bar heights, give or take 6.

What earned points: weekly TP over 45, daily TP over 45, monthly direction change, weekly direction change on the adjacent week, panic cycle sitting on the target date.

What didn’t: price isn’t inside 1% of the gate on the target date. If it drifts to 2.873 or lower by Friday that flips and the score goes to 8. It’s the only line that can move between now and then.

Not in the rubric but sitting there anyway ... two DOUBLE reversals elected at 2.975 on July 10th. Doubles turn up a few times a year on a daily. When they fire it’s an abrupt change, and it was.

Void discipline: price is 0.8% off the 2.720 wall, so it’s testing it, not floating between walls. Void rule doesn’t apply here. If it were sitting mid-range I’d have nothing to say at all.

That’s how I read it. I don’t know how you chart it.

Pine (full pinescript)

The header carries the warning, because I’ll forget otherwise:

// ⚠⚠ CONTRACT BASIS - READ BEFORE TRADING THESE LEVELS ⚠⚠
//   Report anchored to NGQ2026 (August delivery). Last trade: Wed Jul 29 2026.
//   TIMING is calendar-based and survives the roll. PRICE LEVELS DO NOT.

// Daily - active only. Jul 09/10/16 elections excluded (incl. two Doubles at 2.975)
var array<float> d_bull = array.from(3.305, 3.270, 3.265, 3.140, 2.950)
var array<float> d_bear = array.from(2.845, 2.800, 2.795, 2.755, 2.750)

// Weekly
var array<float> w_bull = array.from(4.203, 4.155, 3.735, 3.710, 3.495)
var array<float> w_bear = array.from(2.865, 2.675, 2.560, 2.245, 2.220)

// Monthly - 2.720 is the floor of the shelf. Close below flips it to resistance.
var array<float> m_bull = array.from(5.263, 5.146, 5.136, 5.053, 4.365)
var array<float> m_bear = array.from(2.720, 2.570, 2.510, 2.230, 2.135)

Every daily bear level in there elected on Monday. I’m leaving them in the file rather than editing them out, because the ladder you just came through is the ladder you retest.

What I’m watching

• Friday July 31, monthly close against 2.720. On the August contract. Not on a continuous chart.

• Daily stochastic back above 30. Nothing shorts until it gets there.

• The monthly momentum divergence. It’s been building since April and it hasn’t paid yet.

• My chart marks a daily panic around July 28. My read of the array put it on August 5. Those disagree and I don’t know yet which one is right. Panic means about a 70% shot at taking out both sides of the prior bar, so it isn’t a rounding error.

• 2.950 on the way back up. First level that would change the direction question at all.

Timing from Martin’s Socrates arrays. Execution rules are an overlay I run on top, and the mistakes in applying them are mine. Educational only, not advice, and I’m not managing your money. Trading futures can lose you more than you put in.

Then I checked the number that actually matters, and it stopped me.

Monthly bearish reversals elected: zero.

A quarter of the company’s value gone in four weeks, and by the strict accounting of Martin Armstrong’s reversal system, the long-term trend has not changed. Not yet. That’s not a technicality — it’s the entire framework. Reversals elect on closes, at their own timeframe. A crash, however violent, is just noise against the monthly ladder until a month-end close takes out a monthly bearish level.

The first one sits at $297.82, and it’s a Major. This Friday, July 31, is the month-end close.

That’s the setup. Everything else in this post is about what happens on the way there.

Big Picture

The bull case wrote itself into the reversal ladder between December 2024 and October 2025: ten monthly bullish reversals elected, spanning $371.74 up to $456.23. That was the machine confirming the uptrend, level by level, for almost a year.

Price is now below every single one of them.

Here’s what most people get wrong about that: retracing through elected bullish levels doesn’t un-ring the bell. Those elections are consumed — they’re history, not support. The trend only flips when the opposite side starts electing. And on the monthly level, it hasn’t. $297.82 is the first domino, $214.24 and $212.35 wait far below it, and between them sits a gap wide enough to matter if this ever gets truly disorderly.

One degree down, though, the flip already happened — and it happened fast. The week of July 20 elected the entire weekly bearish ladder in five sessions: $380.15, $364.46 (twice), $337.23, $331.39, $325.60 — four separate times at that one price — and $314.73. Ten elections, three of them Majors, in one week. I’ve been logging these cascades for a while and I haven’t recorded a faster one.

When a market elects an entire cluster in one pass, the cluster changes jobs. $325.60 isn’t support anymore. It’s the ceiling.

And the timing side? July was the month on the monthly array — the tallest bar in the twelve-month window, carrying a direction-change flag, with the trading-cycle model sitting in its low window. Socrates had July circled as the turning month before the month began. The market delivered the turn as a 25% air pocket.

Convergence Window

Top down, here’s how the calendar narrows:

Monthly: July 2026 is the live turning month. The next clusters sit in November (long-term model confluence) and December (the next low window), with a volatility spike flagged for January 2027.

Weekly: this week — the week of July 27 — carries a direction-change flag. The biggest forward bar on the whole weekly board is the week of September 14, timed as a high. The week of August 24 carries a panic-cycle flag.

Daily: Wednesday, July 29, is the tallest daily bar in the window, timed as a reaction high. Monday, August 10, carries the daily panic-cycle flag.

And the part I keep coming back to: Friday, July 31 is a daily, weekly, and monthly close in the same session. Three degrees of the system resolve at one 4pm print — the daily against $306.93, the weekly against $300.41, the monthly against $297.82. I can’t remember the last time a single close carried this much structural weight on this chart.

One disclosure before the paid section, because I publish it either way: this week’s read is built from the visual arrays, not the full text report. That caps my convergence score at 10 out of 14 and puts a ±6 band on the timing estimates. Same disclosure below the wall.

The full plan — both setups, entries, stops, targets, sizing, and the exact conditions that kill each one — is below.

Read more

A follow-up timing study of Martin Armstrong’s War Index (DEFIND). The June window delivered. Price is back at the ceiling for the July 28 date... and the Panic Cycle now sits one week later than advertised.

Five weeks ago I put a calendar on the War Index: late June first, October most.

The June half already happened.

What June did

The 361 - 372 box didn’t survive the window. It cracked lower right around the June dates, and the June 30 monthly close... the slow-timeframe tell from that post... settled under the 363 shelf. First monthly bearish flip on the board since this run began. By early July the basket was printing the mid-340s.

I had called the near window a watch. It paid anyway.

Three weeks of climbing, straight back to the ceiling

Then the part I didn’t expect to be this tidy. From the mid-340s the index spent three weeks climbing, and Thursday it closed half a percent under the ceiling near 368... the top of the old box, and the first level the model actually cares about up here. The month’s high has already kissed it. On Thursday alone the daily reversal board flipped six bullish signals. One session.

Six flips into a ceiling would normally read as strength.

Except for the date. The daily array puts its tallest turning-point bar of the whole window on Tuesday, July 28, with a directional-change marker on the same day... and the week itself carries one too. Price walked back to the level exactly as the clock came due. A level on a random Tuesday is noise. The same level on the array date is when Martin’s system starts paying attention... and this Tuesday is not random.

Which way? The array doesn’t say. It never says. After a three-week climb into a ceiling you can guess which side I’m watching first... but that’s how I read it, and you may chart it differently. The date is the signal. Direction is what the close is for.

Friday, three clocks close at once

July 31 is a Friday and a month-end. Daily, weekly and monthly all settle on the same print.

That makes it the referee for whatever Tuesday starts. A month closing back above the 363 shelf softens the June breakdown. A month closing under it locks the correction in for another leg. Same tell as June 30, same shelf.... twice the weight, because this time the weekly closes with it.

Whatever the headlines say next weekend, the first number I’ll read is that close.

The panic week moved. I like it more now.

In the June post the weekly Panic Cycle sat on the week of July 27. It doesn’t anymore. The arrays redraw as data comes in... they warn you they will... and the flag now sits one week later, on the week of August 3.

Quick recap for anyone new here: a Panic Cycle is a range-expansion flag. The week tends to trade through the prior high AND the prior low before it’s done. It doesn’t pick a winner. The room just gets loud.

And the new placement reads cleaner than the old one: turn date Tuesday, verdict Friday, expansion the week after. The panic week now gets to react to the month-end verdict instead of sharing a room with it.

One detail I can’t ignore. The daily volatility rows for late July are blank. Empty. Then the weekly volatility bars spike exactly there, on the week of August 3. The same coiled-spring look the box had in June, one level down.

Same caveat as last time... this is pulled off the visual arrays, the dashboards, and the estimate already moved once this month. It’s allowed to move again. Ask me on the 31st.

The rest of the clock

Nothing after that changed much.

The weekly array’s loudest cluster of the next twelve weeks sits on September 21, with a second panic flag on the 28th... the bridge into the main event. October 2026 still carries the monthly Panic Cycle with the long-term cycle stacked beside it. Still the date the whole sequence points at.

And January 2027, which I called the echo, keeps getting louder. It’s now the tallest monthly bar of the next twelve, and the trading-cycle row turns yellow there... the setting where the model projects a high and a low onto the same month. The violent kind. A problem for January.

What I’m keeping an eye on

Tuesday, July 28 ... does the ceiling near 368 hold on the date, or give on it. Either answer is information.

Friday, July 31 ... the triple close. The month against the 363 shelf is the tell.

Week of August 3 ... the panic week. Expect both edges of the range to trade. Judge the closes... wicks will lie all week.

September 21 - 28 ... the bridge. If the panic week starts a move, this is where it refuels or dies before October.

The clock said late June first, October most. It has kept every appointment so far... and the loud part starts Monday.

I post each window as it resolves... subscribe if you want the July 31 read when it lands. Free, same as this one.

This is a timing study and market commentary for educational purposes only, built on Martin Armstrong’s Socrates models. It is not investment advice, not a recommendation to buy or sell any security, and not a solicitation of any kind. The instruments referenced are volatile and carry real risk of loss. Do your own work and talk to a licensed professional before making any financial decision.

Gold is sitting at 4132 and there is nothing to do here. That’s the whole post if you’re in a hurry.... the rest is why, and what changes on Friday the 31st.

Quick recap of how we got here. Gold topped in the 4520s in early June. Then the weekly reversals started falling - two weekly bearish Majors elected the week of June 8, a string of minors after that, and the week of July 13 closed under 4020 and took out three more. Low of that move printed 3963. And then instead of following through, price turned around and climbed straight back up through the levels it had just broken.... which is how we ended up at 4132, hanging in the air between the walls.

Big Picture

The monthly chart hasn’t blinked once. Ten monthly bullish reversals elected on the way up since July 2024, and zero bearish elected... none. The first monthly bearish Major sits down at 3841 and price has never closed a month below anything that matters. So whatever June and July were, on the monthly timeframe it’s still a correction inside a bull market.

The weekly and daily charts disagree with that. Both elected bearish Majors in June. Same market, opposite verdicts depending on where you stand.

That argument gets settled on July 31.

The Convergence Window

Martin’s arrays, top down:

Monthly - the heaviest bar in the next twelve months is August 2026. No direction change flagged. The Panic Cycle on this timeframe sits out at February 2027... a story for another post.

Weekly - the strongest week is July 27, with direction-change flags on this week AND next week, back to back. A weekly Panic Cycle lands the week of August 3, right behind it.

Daily - the tallest bar of the whole array is Friday, July 31. Direction changes cluster on the 22nd, 24th, 27th and 30th. Daily Panic Cycle on the 28th.

So Friday July 31 is the daily target.... and it’s also a weekly close.... and it’s also the monthly close. Three closes stacked on one print. When timeframe boundaries pile up like that, the biggest close wins - the monthly close is the one that decides the structure, and everything below it is noise until that number prints.

The Honest Score

6 out of 14. WAIT.

Two disclosures before you weigh that number. First, I only had the visual arrays for this one, no Premium text... which caps my scoring at 10 and means the turning-point values are my estimates off bar heights. The bar-height scale is a written rule file, not vibes - one I recently had to rebuild from my own old transcripts after losing the originals... documented method, eyeball inputs. It also tends to UNDERESTIMATE strong weekly clusters, so the real score might be higher. Might. I’d rather show you a true 6 than an invented 9.

Second, why WAIT and not a trade. The rule needs a level under active attack before anything qualifies, and right now the nearest active walls are roughly 3% away in both directions. 4132 is open air. Trading open air is how you donate money.

One more thing, and it’s the most interesting number on the board: the weekly stochastic just crossed bullish from DEEP oversold... 14.8 over 14.7. A cross by a tenth of a point. One bad week un-crosses it, so I’m treating it as armed, not proven, until Friday’s weekly close holds it. But it does two things regardless. It leans the resolution long. And it kills every short idea outright - I don’t short into a weekly stochastic under 20. Ever. That lesson is already paid for in my journal.

Below the line: the exact entry zones, both stop tiers, three targets with R:R, the alternate plan if price never comes back down, and the TradingView levels script.

Read more

Two weeks ago, Kraft Heinz did something it hadn't done in over a year: it elected an entire ladder of daily bullish reversals in a single session. Ten levels, all in one day, July 16. Price spiked past $27. If you watch this stock, you saw the candle and you probably thought the same thing everyone thinks — the turn is in, the move is on.

Here's the gap between that feeling and what the system actually says: nothing has been decided. Price has retraced right back onto the levels it just elected. The monthly trend is still ten-for-ten bearish. And the entire question — bounce or trend change — resolves on a single monthly close against a single price, ten days from now.

You don't need a forecast here. You need one date and one number. That's what this issue gives you.

Disclosure up front: this analysis was run from Socrates array and reversal screenshots, not the full text report. Timing scores are visual estimates; every price level and Direction Change / Panic Cycle flag is exact. The convergence score below reflects that honestly.

The Big Picture

Martin Armstrong's Socrates platform tracks KHC on three timeframes, and right now they disagree in the most instructive way possible.

Monthly: bearish, but empty-handed.

Every one of the ten monthly bearish reversals — from the 32.70s elected back in mid-2024 down through 23.35 elected in March 2026 — has already fired. There is not a single active monthly bearish reversal left below this market. Read that again: the monthly downtrend has nothing left to elect. The first active monthly level is a bullish reversal at $35.45, a full 37% above the market. The bear structure is intact but exhausted.

Weekly: a clean four-stage bullish march.

Since the March low, the weekly has elected bullish reversals in four separate waves — 23.90 in early May, 24.12 in June, a cluster at 24.50–25.20 at the end of June, and the 25.66–25.75 Majors in the week of July 13. When one side elects nine levels and the other elects one, the system's hierarchy rules give the weekly trend to the bulls without ambiguity.

Daily: elected, then retraced.

July 16 elected everything from 25.60 up to 26.19. Price overshot to roughly $27.10, then came all the way back. Friday closed at 25.86 — sitting inside the cluster it just elected. In this methodology, elected bullish reversals become the support shelf. The market is now standing on it.

So: a bear market that has run out of levels, a weekly uptrend building stage by stage, and a daily that just pulled back onto fresh support. That configuration has a name in this framework — a transition. And transitions resolve on closes, not on feelings.

The Convergence Window

Monthly. July 2026 carries a Direction Change on the monthly array — and by the close-basis rule, a monthly signal resolves on the monthly close. That's Friday, July 31. The array's biggest months are still ahead: August and September score high, and October carries both the highest turning-point score on the board and a monthly Panic Cycle. Whatever July 31 decides, October is where it accelerates.

Weekly. The week of August 31 is the strongest weekly bar in the next twelve — by a wide margin — and it carries its own Direction Change. A weekly Panic Cycle follows in the week of September 14. The weekly array is pointing at late summer.

Daily. This week matters. July 21–22 carry back-to-back daily Direction Changes, and July 23 is the strongest daily bar in the array with a tall Trading Cycle signal. Then the array goes quiet until July 31 — which lands a daily Panic Cycle on the exact same session as the month-end close and the monthly Direction Change resolution.

Three timeframes, one bar: July 31 is the decision date. Everything before it is positioning.

What the Indicators Say (And Why It Blocks the Obvious Trade)

The rule that governs entries here is unforgiving: time and price meeting is not a trigger. The indicators — a slow stochastic and a momentum cross — have to fire in the trade direction, or there is no trade. Full stop.

Right now the daily indicators are crossed down, rolling over from overbought. The weekly is split. The monthly has crossed up from oversold for the first time in this entire decline.

So the obvious trade — buy the retest of the elected shelf, right now, because the levels are pretty — is exactly the trade the system prohibits. Price and timing are converging; the trigger hasn't fired. Buying anyway is the single most common way to donate money to this pattern. The system has a specific journal entry format for that mistake, and I'd rather not use it.

That discipline is the whole edge. The setup below is fully armed. It just isn't lit.

The Honest Score: 5 out of 14 — Moderate. Verdict: WAIT.

I score every setup on a 0–14 convergence rubric and publish the number whether it flatters the trade or not. This one scores 5: monthly Direction Change in the window, daily Panic Cycle on the decision date, unanimous monthly structure, and price within half a percent of the key weekly level. What's missing: the timing magnitudes are middling, and the image-only input means Fibonacci, Elliott Wave and ECM confluence couldn't be scored at all.

A 5 is not a table-pounder. It's a watch this specific thing score. The specific thing: whether the daily indicators re-cross up while price holds the shelf — before July 31.

Below the line: the exact trigger condition, entry zone, both stop tiers, all three targets with the risk-reward math, why the short side is structurally blocked (this part surprised me), and the TradingView Pine Script indicator with every level and array date pre-loaded.

Read more

I want to go long anyway.

I don't have an opinion on printers. The daily elected two bullish Majors off the June low... 23.40 on July 8, 24.40 on July 13... and the monthly already elected 25.80 bullish back at the May close. The monthly says up. The daily agrees. The weekly is the one still saying no.

Somebody is wrong.

Some history, because the weekly's no is not random. The monthly spent last winter electing bearish.... November, December, twice in January.... a staircase down from 27.50 into the 19s. Then the May close took out a bullish Major at 25.80 and the bigger board flipped. June promptly gave half of it back. That is the whole argument you're watching... a monthly trying to turn, a weekly that still remembers the winter.

The timing says we find out this week. The daily array puts its tallest TP bar with a direction change on July 24. The weekly array peaks into the week of August 3. And there is panic on the 29th and again the week of August 10.

Panic in Martin's model does not mean down. It means a bar that can take out both sides... the prior high and the prior low, same session. Two of those sit right behind the turning point, so whichever side wins this week, expect to be shaken at least once before the move is allowed to run. The volatility bars on the array agree... they build straight through the end of July and peak into early August.

The indicators are split the same way the timeframes are. The weekly momentum washed out in June and crossed back up, which is usually the tell that a low is in. The monthly is pointing up and has been since spring. The daily stochastic is the holdout... 66, under its signal, still rolling off its July 10 sell. Momentum crossed, the stochastic didn't. Half a trigger. Erwin's rule says half a trigger is zero triggers... and that's the rule I break most often, which is why this plan exists in writing.

Price closed Friday at 24.84. Above the level it elected, below the level it needs to elect next.

The void.

Nothing to do in the void... which is exactly when I get the urge to do something. So I wrote the plan down before Monday could talk me into a boredom trade. One more date while you wait: July 31 is a Friday and a month-end at the same time, so whatever gets printed along the way, that close has the final word.

The gate number, both entries, the short I hope I don't take and where the whole thing is wrong are below the line... that's the part worth paying for, so it's paid.

Read more

Uranium Energy tagged 9.21 in the pre-market this morning. The daily array had today marked as the strongest bar of the month, with a direction change sitting on it. Time and price met at a Major while I was still on my first coffee.

Before anything else... this read came off screenshots, not the full Premium text. Image-only mode. Bar heights are estimated and the convergence score is capped at 10 out of 14. Strong weekly clusters usually come out underestimated when I work this way. That’s how I read it..... I don’t know how you chart it.

The big picture

The monthly ladder on Martin Armstrong’s Socrates is still bullish. Ten elections from August 2025 through January of this year, Majors included. Meanwhile the weekly and the daily have been electing bearish reversals since the second week of June without a single skip in the ladder.... twenty of them between the two timeframes.

Both things are true at the same time. Until one degree gives.

The tiebreaker is an unelected monthly Major bearish sitting at 10.00. If July closes below it, the election fires and the entire bullish stack from last autumn is void. If July closes back above it, that’s a failed election at a Major.... and failed elections at Majors have a habit of snapping hard the other way.

One number. Month-end. 10.00.

The close decides. The intraday spike through a level means nothing here.... Martin’s crude example from 2007 settled that argument a long time ago.

The convergence window

Monthly: July carries the tall bar and a direction change. November is the next node. March 2027 is the tallest bar on the whole array, for whoever trades that far out.

Weekly: the week of July 27 is the strongest bar this month. There are no weekly direction changes anywhere in the twelve columns, which I don’t see often. The weekly panic cycle sits in the first week of September.

Daily: today was #1, with the direction change. July 23 is #2. July 30 is #3, and July 29 next to it carries a panic cycle. So the window is July 28 to 31.... and July 31 happens to be the last trading day of the month, which means the whole thing resolves on the monthly close against 10.00. The timing and the level land on the same bar.

Honest score: 9 out of 14, capped at 10 for image mode. What’s in it: the tall monthly bar, the tall weekly bar, the tall daily bar, the monthly direction change, the panic cycle inside the window, and price sitting within 1% of a Major on the array date this morning. What’s not in it: no Fibonacci or Elliott dates available from screenshots, no ECM check, no indicating ranges.

Setup is not trigger

Here’s the part that keeps this a WAIT instead of a trade.

The daily stochastic is at 28 and falling, no cross. The weekly is at 12.... deep oversold. Shorting into that is the exact trade my journal already has a name for, and the name is not flattering. Time and price met this morning. The indicators did not. The rule is the rule: the meet is the setup, the cross is the trigger, and without the cross you post the WAIT and go do something else.

Price is also seven percent below the last elected shelf at 10.00 to 10.10. Fresh shorts down here chase. The system says wait for the retest.

So: WAIT, published before the close, again. The scores and the dates are above. What follows is what I’d actually do about it.

Not financial advice. I trade my own account and publish my homework. The methodology is Martin Armstrong’s Socrates platform for timing and reversals, with an execution overlay for entries and stops. Positions can change without notice.

The chart looks like free money on the short side. Monthly elected two Major bearish reversals in June - 0.006272 and 0.006286 - and the weeklies stacked ten bearish elections on top of that, including a 4-count. When Martin’s system prints a 4-count on a weekly close, that’s continuation. Structure is bearish on every timeframe I track.

Price is sitting right on the last daily Major bearish at 0.006176. Pierced it intraday today. Not elected yet... elections happen on the close, and today is a Friday, so the daily close and the weekly close are the same print. That matters.

The timing

Working from screenshots this time, not the full Premium text.... so the convergence score is capped at 10/14 and the bar heights are estimated. Mode C, disclosed. It scored the cap anyway: 10/14.

Monthly array points at August. Weekly array’s tallest bar is the week of Aug 17, with a weekly panic the week after. But the near thing is the daily array: July 27 is the tallest bar in the whole run, it carries a panic cycle, and it’s bracketed by direction changes on the 24th and the 27th. Paired direction changes around a bar where time and price meet..... that’s the signature of a turn.

So time says July 24-27 and price says 0.006176. They’re meeting. And I’m still not shorting it.

Why not

The oscillators. Daily stochastic is at 22. Weekly at 12, and it already crossed UP. Monthly at 3.

Three. On a scale of a hundred.

The rule I trade by prohibits shorting into oversold, and I have a losing wheat trade in my journal from the one time I decided price plus timing was enough without the cross. It wasn’t. Shorting yen here means selling the low of a three-month move into a panic bar, with the weekly stochastic already turning up underneath you. That’s how I read it - you chart it your way.

What the system actually wants

The panic bar on July 27 takes out both sides more often than not. If it flushes down and the daily stochastic and momentum cross UP off that low... that’s a reaction long. Against the monthly, so half size, per the rule. Stop below whatever low prints, closing basis. First target 0.006255, where five active bullish reversals sit inside 21 ticks, 0.006255 to 0.006276. Roughly 3:1 to the first one.

And that stack is where the real trade lives. A bounce into 0.006255-0.006276 that works the stochastics off the floor, then a daily cross back down.... that’s the short. With the monthly this time, full size, into the week of Aug 17 where the weekly array put its tallest bar. The weekly panic of Aug 24 sits right behind it.

Sell the bounce, not the hole. Same trade everyone wants, just higher and three weeks later.

Verdict

WAIT. Score 10/14, Mode C cap, image-only input, disclosed. No entry until the daily crosses fire. If they never fire, the window expires and waiting was the trade.

What I’m watching:

• tonight’s weekly close against 0.006176

• daily stochastic and momentum crosses, July 24-27

• the 0.006255-0.006276 stack on any bounce

• week of Aug 17

Not advice. I trade my own book and publish the score before the close so you can watch me be wrong in real time.