GYP Weekend Market Intelligence & TLT Technical Breakdown
The S&P finished the week up about a percent. Gold was roughly flat. If that's all you saw, you missed the most instructive week of the year.
This Week’s Market Intelligence
The S&P finished the week up about a percent. Gold was roughly flat. If that’s all you saw, you missed the most instructive week of the year.
Underneath those numbers: Amazon rose 15.32% on Friday alone. Microsoft finished the week up 21.75% — and is still down 3.91% on the year. Sandisk fell 15% on the week after falling 47% on the month, then ripped 20.83% in a single session. Apple dropped 7.35% after beating. And momentum as a factor had its worst month in twenty years.
That is not a one-percent week. That’s a violent redistribution wearing a calm suit.
The story almost nobody led with
Midweek, CNBC reported that Leopold Aschenbrenner’s Situational Awareness was unwinding after steep losses. By Thursday the scale was clear:
Assets reportedly fell from $45 billion to around $10 billion
The fund was said to be running roughly 4x leverage
Down about 67% in July — and still up 80% year to date
Citadel bought the bulk of the public-equities portfolio
Amit put it more precisely than any wire did: “Semis get destroyed all month. Turns out it was a structural deleveraging event that really caused the worst parts of it. Best day for semis in a year once the biggest AI-levered hedge fund gets liquidated.”
That single fact reorganizes the entire month. And nothing shows it better than putting the month and the year side by side:
Read the Sandisk row twice. Up 412% on the year, down 47% on the month, and still 48% below its high. Those three numbers describe the same stock in the same year.
And look at the last row. NVIDIA — the name everyone calls “the AI trade” — is up 7.64% on the year while Sandisk is up 412%. Whatever happened in July, it did not happen to “AI.” It happened to the specific book somebody was levered into.
A complex falling 30–50% on no change in fundamentals, then ripping 10–20% the instant the selling stopped, isn’t a market forming a view. It’s a forced seller finishing.
Forced selling and informed selling look identical on a chart and mean opposite things. Almost every retail investor who sold semis in July sold into someone else’s margin call and called it risk management.
And the lesson underneath is older than any of this. That fund was right about AI. Right about the stocks. Still up 80% on the year. It got taken apart anyway, because 4x leverage on $45 billion doesn’t let you be early. As Tony B said on Office Hours this week: “Size kills. He just wanted to show he had a big stick.”
Trader’s lens. Two things to carry into Monday. Dispersion is the opportunity — our own scanner has SPY’s 14-day implied vol at 16.7 against QQQ’s 28.4, a twelve-point spread. You’re paid roughly double on the Nasdaq side for more cushion. And the vol collapse is over: SPY’s expected daily move fell from 0.78% Wednesday to 0.487% Friday, but VIX rose 7.8% Friday to 18.42. The front day is quiet; the next month got dearer. “Vol is dead” was right Thursday and wrong by Friday’s close.
Investor’s lens. When a levered fund liquidates, the prices it prints are not information about the businesses. If you were watching Micron fall 30% and concluding something about memory demand, you were reading a margin clerk’s handwriting. The useful takeaway isn’t “buy the dip” — it’s that you cannot distinguish forced selling from informed selling in real time, which is exactly why position sizing has to survive being wrong about which one you’re looking at.
The other thing that genuinely changed: bonds
The Fed held, but two regional presidents — Dallas’s Lorie Logan and Cleveland’s Beth Hammack — publicly backed a 25 basis point hike. The 10-year pushed to 4.7388%, its highest since January 2025.
By Friday the market had stopped asking when cuts arrive and started asking whether another hike does. You could see it in what broke: Russell −1.4% (three and a half times the S&P’s loss), gold −1.7%, bitcoin −3.5%. Every long-duration, no-cashflow asset at once.
That’s a discount-rate move, not a growth scare. University of Michigan sentiment came in at 55.2 against 54.0 expected, up sharply from 49.5. The consumer is getting better, not worse.
Sentiment: a month of fear, two percent from the highs
CNN’s Fear & Greed Index closed the month at 42 — Fear. Here’s the run:
The market has been in Fear for a solid month — while the S&P sits 1.76% below its all-time high and the Russell is up 18% on the year.
That gap is the single most useful sentiment fact available right now. People are not behaving like owners of a market near record highs. They’re behaving like people waiting to be hit.
Note also that the gauge rose from 38 to 42 on a day the market fell. That’s the midweek volatility collapse feeding through the components — sentiment improving on the same session that yields spiked.
Trader’s lens. A month of sub-45 readings with the index near its highs is a premium seller’s environment: people are paying up for protection they mostly won’t need. That’s precisely what we were paid for on Tuesday and Wednesday. But respect the second derivative — 29 to 42 means the fear is draining, and drained fear is thinner premium. This is late in that particular trade, not early.
Investor’s lens. One year ago this gauge read 63 — Greed — with the S&P lower than it is today. Sentiment told you nothing about the next twelve months except how it felt at the time. Use it to understand what you’re being paid for, not to decide what to own.
Where The Year Actually Stands
Small caps are leading the year — and were the worst performer on Friday’s rate scare. Both things are true and they’re the same fact from two angles: the Russell carries floating-rate debt, so it benefits most from expected cuts and suffers most when a hike returns to the table.
Worth remembering the shape of this year: everything was down 5–10% at the April lows. The Russell’s 18% and the Nasdaq’s 12% are round trips, not straight lines.
The Five Stories That Moved Markets
1 · Mega-cap earnings separated the winners from the losers — but not the way most people are saying
Microsoft and Amazon delivered; Apple and Meta were punished. The lazy version is “execution matters now.” The data says something sharper.
Charlie Bilello published Q2 revenue growth for the group:
Meta grew revenue fastest in the entire group and finished the week down 6.5%. Microsoft grew slowest of the four big spenders and finished up 21.75%. So it isn’t revenue growth. It isn’t even capex — every one of these companies is spending enormously.
The distinction is whether the spending is visibly tied to demand. Amazon raised 2026 capex guidance from $200 billion to $220 billion — and the stock rose 15.32% on Friday and 17% on the week, because it paired the number with “demand still far more than supply,” 2028 demand already booking, and a claim that AWS becomes a $1 trillion-a-year business. Meta raised capex too, and reported free cash flow down 91% to $784 million, the lowest since Q3 2022, with no comparable demand signal attached.
Same spending. Opposite verdicts. The market is not rating AI investment — it’s rating whether management can point at the customer.
The scoreboard backs it up. Amazon ended the week 2.51% below its 52-week high — closer to its highs than any other mega-cap on our list, SPY included. Microsoft is still 3.91% down on the year even after a 21.75% week. Being forgiven is not the same as being whole.
One more from Bilello worth sitting with: Apple has bought back $747 billion of its own stock over ten years — more than the market cap of 486 of the 500 companies in the S&P.
Trader Take: Earnings reactions matter more than earnings. Apple beat and fell 9%; the beat was not the trade. Watch where institutional money goes after the print, and remember that post-earnings implied volatility stays elevated for days — which is when premium sellers get paid.
Investor Take: The test being applied is consistent and worth adopting: is the spending tied to demand you can see? Alphabet posted record revenue with its first negative free cash flow since the 2004 IPO and fell 7%. Meta’s cash flow collapsed and it fell 9%. Amazon guided capex up $20 billion and rose 9% because it explained who’s buying. That’s a usable framework, not a narrative.
2 · The Fed held, but the market turned hawkish
Rates unchanged, tone notably more cautious, dissents in favor of hiking, and by Friday a September hike had gone from unlikely to live.
The data cut both ways. Q2 GDP came in at 1.5% against 1.8% expected. But core PCE rose only 0.1% on the month against 0.2% expected, and claims at 197,000 beat the 200,000 consensus. Liz Ann Sonders’ last data before her break was similarly mixed: July Services PMI 53.6 against 51.5 expected, Composite 53.6 against 52.2, with Manufacturing softening to 53.8.
Then read the level rather than the change: core PCE is still 3.3% year over year against a 2% target, with the funds rate at 3.50–3.75%. The real policy rate is barely positive. That arithmetic is exactly what Logan is arguing from.
Trader Take: Rate expectations are driving intraday volatility again. Expect outsized reactions to employment and inflation data — and note the market’s reading of the Fed reversed twice in 48 hours this week without the Fed doing anything at all.
Investor Take: Higher-for-longer favors strong cash flow, pricing power and clean balance sheets. If you hold long-duration assets, you now hold an interest-rate view whether you chose one or not.
3 · Treasury yields reached new highs
The 10-year hit 4.7388%, the highest since January 2025. Ed Yardeni called it a day early — “Warsh fails first test as Bond Vigilantes drive yields higher” — and the print confirmed him within 24 hours.
The fiscal backdrop isn’t helping. Bilello notes the national debt rose more than $400 billion since July 1st alone. Peter Mallouk adds the longer arc: the CBO projects federal debt held by the public reaching 175% of GDP by 2056. Treasury Secretary Bessent says the administration is “laser-focused on paying down the debt.” The bond market is voting on that claim in real time.
Trader Take: Watch bonds as closely as stocks. Friday’s Russell −1.4% against the S&P’s −0.4% was a rates trade wearing equity clothing, and that rotation happens inside a session.
Investor Take: Higher yields don’t end bull markets, but they reliably reward quality over speculation. And they raise a question worth answering before you need to: at what yield do long bonds become genuinely attractive to you? Write the number down now, while it’s abstract.
4 · AI spending is being validated — but investors want receipts
Cloud growth confirmed corporate AI budgets are intact and growing. Amazon’s capex raise to $220 billion is not a company hedging. What changed is the standard of proof.
Trader Take: AI has shifted from a momentum trade to a stock-selection story. Expect larger single-name moves around earnings — and expect the index to hide them, exactly as it did this week.
Investor Take: The buildout remains one of the strongest secular trends available. But “the trend is real” and “this price is sensible” are separate questions, and the market started asking the second one. Bilello’s line is the one to keep: AI may be revolutionary, but the price you pay still matters.
5 · August begins with genuinely elevated event risk
The employment report, more earnings, and a Middle East situation that got materially worse over the weekend.
This is the part we’d flag hardest. Reporting from CBS News suggests the US and Israel could begin bombardment of Iranian energy sites as soon as this weekend. Alongside that: CENTCOM has redirected 30 commercial ships, disabled two and boarded two as of July 31. Kevin Gordon flagged Strait of Hormuz vessel crossings at zero. Exxon and Chevron have both cautioned that fuel prices could stay high because of the war, and UK petrol prices have reached their highest since it began.
Crude closed Friday at $85.25 (+2.0%) and Brent at $88.23 (+1.6%) — the first session in a week where price stopped disagreeing with the news.
There’s also a currency situation building. The BOJ appeared to intervene Thursday, and by Friday the US Treasury had reportedly told banks via the New York Fed to stand by for “future actions” in the yen market.
And Keith McCullough flags two signals worth having on your screen: the dollar’s downside is becoming exhausted, and Bitcoin broke its $64K momentum level while retaining its strongest inverse relationship with the dollar. His note on Friday: “No month-end markup by Wall Street in Bitcoin, eh.”
Trader Take: Expect volatility to rise, and expect the gap risk to be in energy. If the weekend headline lands, oil gaps and you don’t get to trade the open. That’s an argument for defined-risk structures over naked exposure into Monday, and for not being maximally deployed right now.
Investor Take: Don’t let a heavy calendar drive decisions. Volatility routinely creates the chance to add quality at better prices — and it always arrives disguised as a reason not to.
What We Actually Did — And What Discipline Cost Us
We publish every trade, so here’s the week in numbers rather than commentary.
We sold into the panic Tuesday and Wednesday. Four starter puts into the memory selloff, then Corning, Nokia and a Circle ZEBRA — every one a name that had been taken apart. Corning was down 51% in a month. Circle was 67.5% below its high.
We harvested Friday. Portfolio 1 closed eleven positions for +$6,105, every one a winner — $9,536 of premium sold, bought back for $4,541, keeping 52%. Portfolio 2 rotated: closed the NVIDIA and Tesla jade lizards, closed and re-sold the same QQQ put twice in one session, and opened an Apple jade lizard into the earnings drop for a $1,060 credit.
But the number that matters isn’t the P&L. Portfolio 1’s buying power usage went from 22.22% to 10.80%. The book ended July at $1,102,870 with 89.2% of its buying power unused — better than 10% in seven months while never using more than about a quarter of available capital.
Trader’s lens. Two lessons the week paid for. We were paid a fear premium created by someone else’s margin call — which, if the deleveraging read is right, is close to the best reason to sell premium there is. And closing early cost us money and we did it anyway: if all eleven had expired worthless we’d have kept another $4,541. We chose certainty over the last 48% because the remaining premium no longer paid for carrying risk through August and a live September Fed.
Investor’s lens. Put our week beside the cautionary tale and the whole lesson is visible in one frame. A fund that was right about AI, right about the stocks, and up 80% on the year lost two thirds of its value in a month because it ran 4x leverage. We were right about far less, and finished the month up, with 89% of our capital untouched. Being right is optional. Surviving is not.
Featured Technical Breakdown — TLT (20+ Year Treasuries)
[CHART — TLT daily, 1 year, with Bollinger Bands (20,2), 200-day SMA and RSI]
We’re featuring bonds because bonds were the story, and because TLT closed Friday sitting on a line that matters.
Where it closed
TLT finished at $82.25, down 0.66%. Four numbers off the chart tell you everything:
What the chart is actually saying — and it’s two things at once
Short term, this is stretched. Closing below the lower Bollinger Band is a statistically uncommon event — it means the day’s move exceeded two standard deviations of the recent range. Pair that with RSI at 31.6 and you have a market that has moved a long way, fast, in one direction.
Structurally, this is a downtrend. Price is 4.4% under the 200-day, and look at the shape of that blue line on the right edge: it was flat-to-rising all year and has just curled over. A declining 200-day with price beneath it is not a bottoming pattern. It’s a trend.
Those two readings are not contradictory — they’re different timeframes. And confusing them is one of the most expensive mistakes in technical analysis. Oversold in a downtrend is not the same as a bottom. It usually just means the next bounce starts from here rather than lower — and that the bounce is a bounce.
The line that matters: $81.89
TLT is thirty-six cents above its 52-week low. That’s the whole technical story in one sentence.
If the 10-year keeps pushing toward the 5% Tony B has been calling for, that low breaks, and there is no chart support underneath it — you’d be looking at levels not seen in this data window at all. If yields stall here, the setup for a bounce is already in place: below the band, RSI in the low 30s, and the 20-day mid at $83.88 as the first target, roughly 2% up.
“They were all telling me to buy the ten-year at four twenty-five — it’s a gift. Then four forty-five — it’s a gift. And now it’s a gift that keeps on giving. I can see five percent on the ten year.“ — Tony B, Office Hours, July 31
Trader Take
Do not confuse “stretched” with “reversing.” A close below the lower band with RSI at 31 is a setup for a bounce, not a trend change — and the trend change doesn’t happen until price reclaims $85.87 and the 200-day turns back up. Two levels to trade against:
$81.89 — lose it and there’s nothing underneath in this window. That’s your invalidation.
$83.88 — the 20-day mid band. First target on any bounce, about 2% up, and where a short-term long should be thinking about taking something off.
For premium sellers, the more interesting expression is that bond volatility has been rising while equity volatility collapsed midweek. That relationship rarely stays stretched. If you want long exposure here, defined-risk structures make far more sense than outright longs into a possible weekend energy headline.
Investor Take
You’ve had a bad year in long bonds and the reason hasn’t changed — you’re accepting a fixed coupon while the government issues $400 billion of new paper a month and core inflation sits at 3.3% against a 2% target. TLT is down 5.63% year to date and 10.8% below its high. None of that is a mystery.
But here is the question actually worth answering: what yield makes long bonds a buy for you?
Not “are bonds bad” — they’re priced as though the answer is yes, and that’s precisely the point. Somewhere above here, a government-guaranteed coupon becomes genuinely attractive relative to equities trading at these multiples. Tony B thinks 5% on the ten-year is visible. Decide now, while it’s abstract, what number gets you interested — because when it arrives it will arrive alongside headlines telling you not to buy.
That is the entire discipline. Write the level down today. Act on it later.
Bottom Line
This week marked a transition in leadership and in psychology.
A $45 billion fund at 4x leverage was liquidated, and the resulting violence in both directions was widely mistaken for information
Momentum had its worst month in twenty years — a factor event, not a fundamental one
Mega-cap earnings proved AI winners still exist, but the test is now “can you point at the customer?” rather than “are you growing?”
Treasury yields at an eighteen-month high became a genuine headwind, with a September hike back on the table
August opens with live energy risk that could gap markets before Monday’s bell
For traders: risk management, earnings reaction and macro awareness now matter as much as any chart. The dispersion between index and single-name volatility is the clearest opportunity on our screens — and the reason we’re closing index premium while still selling Nasdaq premium.
For investors: the long-term outlook stays constructive, but the easy gains from simply owning AI have probably passed. The next phase rewards quality, cash flow, valuation discipline and portfolio construction over momentum.
The bull market is still alive. Wall Street has just become far more selective about where it places its bets — and the one fund that placed the biggest bet of all is now 67% smaller.
One More Thing
If you missed Friday’s Office Hours, we spent the hour opening the members portal and walking through every section — the three live portfolios, the full trade history, the Put Selector, the new Max Yield grid, the Black Swan hedging tools, and the Portfolio Optimizer, which we ran live on two members’ real books.
The replay is in the portal. If you’ve been logging in, glancing at open positions and logging out again, that hour will show you what else you’re paying for.
Grow Your Pile is educational. Nothing here is a recommendation to buy or sell any security. We show you our trades and our reasoning so you can learn the framework. Your account, your size, your risk. Options involve risk and are not suitable for all investors. Selling puts carries undefined downside risk — assignment obligates you to purchase shares at the strike price regardless of how far the stock has fallen. Past performance is not indicative of future results, and the results shown are those of our own accounts and are not representative of any subscriber’s results. Read the Characteristics and Risks of Standardized Options before trading.
— Tony Battista & Tony Rihan