The Morning Bell No. 9: One Up Thirteen, One Down Twenty

Lululemon lost about a fifth of its market value overnight. Samsara gained thirteen percent. Seven companies reported after yesterday's close into the same macro and the same tape, and the distance between the best and the worst of them was more than thirty points of price.

Nothing here is financial advice. I am not a licensed advisor. The account in this newsletter is a small automated test account, entirely separate from my personal investing. Entertainment and education only.

The tape before the bell

Every price here carries the time I pulled it. These are premarket marks, not opening prints.

SPY 772.04 at 9:27 ET against a 773.17 close, down 0.15 percent. QQQ 719.28 at 9:27 ET against 717.67, up 0.22 percent. IWM 294.16 at 9:27 ET against 295.19, down 0.35 percent.

VIXY 16.77 at 9:26 ET against 16.95, down 1.06 percent. TLT 82.27 at 9:27 ET against 82.07, up 0.24 percent. GLD 404.30 at 9:27 ET against 410.22, down 1.44 percent. Bitcoin 79,366 at 9:27 ET against an 80,849 prior close, down 1.83 percent.

The regime reads green. SPY sits above its 20 day average of 769.21 and its 50 day of 756.14, and the VIX printed 14.12 this morning against a 60 day median of 16.27. Both of my green tests pass and neither red test fires, so sizing runs at full multiplier.

One honest caveat on all of that. My research file was built at 7:46 ET and the August employment report came out at 8:30 ET, so what I have is the market's reaction and not the number itself.

The reaction is worth reading on its own. Gold is down 1.44 percent and bitcoin down 1.83 percent, long bonds are up a quarter percent, and the Nasdaq is green while the S&P and small caps are red. Gold took the biggest hit of anything I track.

What the system flagged overnight

Seven companies reported after yesterday's close. Exactly one of them passed my beat and raise test.

Samsara printed 0.20 against a 0.13 estimate, which is a 54 percent beat. Revenue came in at 508.4 million, up 30 percent, and the full year outlook was raised to a 2.043 to 2.047 billion range. It gapped 13.03 percent.

The other six went the other way or went nowhere. Lululemon fell 20.35 percent, Guidewire 15.08 percent, UiPath 8.34 percent and Zscaler 3.26 percent, all of them on guidance rather than on the quarter itself.

A down gap after a guidance break is not a defined setup in my rules, which class beats only. So the four biggest moves on the board generated nothing at all, and I would rather say that plainly than invent a short thesis to fill the space.

The panel of outside voices I track was close to useless overnight. Two of thirteen sources had a usable stance inside the window, which is 15 percent and the weakest reading this system has produced, so convergence is suppressed and those two get reported as description only.

One of them opened a real position, SK Hynix through the SKHY listing at 158.51, structured long dated. The other is hedging for higher US rates. Neither mentioned the jobs report and neither mentioned any of the seven companies that reported, which tells you where their attention was.

One source claim graded wrong on a live check. A note timed 17:40 ET yesterday said bitcoin had broken 92,000. My broker mark was 81,275 at 07:50 ET this morning, about 13 percent away, so it is logged as wrong rather than stale and that source's price calls carry less weight today.

The setups on my screen

These are research notes. They get graded here tomorrow whether they work or not.

The account I run this in is a small automated test account, entirely separate from my personal investing. It holds 101.43, all cash, and it may only buy single leg options outright, never spreads and never short options.

Samsara long, on post earnings drift. The mistake I think is being made is that analysts were anchored to a 0.13 estimate against a 0.20 print, and index and passive flows reprice a name like this over days rather than inside one session.

Reference price 43.80 from a 07:57 ET premarket quote, so treat it as a stamp rather than a live number. Stop 41.29, target 48.82, earnable R of 2.00.

That stop needs resetting to the actual earnings day low once one exists, because a pre gap ATR of 2.09 understates post gap range every time. The five session math on this setup fails and the ten session math passes, so the horizon is ten sessions and I am saying the failed leg out loud rather than burying it.

It dies if it opens under 39.53 or loses VWAP inside the first thirty minutes. Crowding reads moderate, which after yesterday is the single best thing about it.

Then there is the sizing, which is where this ends. Post earnings implied volatility rank on Samsara sits above my 65 bar for single leg longs, and protocol size floors to zero contracts against a 101.43 balance.

So the answer is nothing, at any conviction tier. That makes fourteen consecutive fires of this account with no eligible entry, and that is a structural finding about a hundred dollar account, not bad luck.

Three names stay on watch. Lululemon and Guidewire both flip to a setup on a failed bounce rejected at the gap edge, which needs at least one full session of price action first. DocuSign at 1.73 percent sat under my 2 percent bar and flips if it opens above it and holds.

Today's calendar

The August employment report landed at 8:30 ET, before this went out. Weekly options expire at 16:00 ET.

Markets are closed Monday for Labor Day. Tuesday brings ABM and UNFI in the morning, then GME, CASY, BRZE, TTAN and AVO after the close.

Wednesday has CHWY, SIG and ASO in the morning and AEO, AVAV and COO after it. Thursday is the heavy one, with M and SKHSY before the bell and ORCL, ADBE, CPRT and RH after. Friday brings KR in the morning.

Worth noting that SKHSY is the SK Hynix American listing, reporting six days after the panel's largest disclosed position was opened in it.

This month's CPI date I still have not verified, second day running, so it stays off the calendar rather than going on as a guess.

The report card

Yesterday's setup was Snowflake long after a beat and raise. It triggered.

It confirmed the gap and hold at 09:33 at 377.60 and then faded all session to close at 356.47. Against a 377.42 anchor and a 361.64 stop, day one came in around minus 1.33R. The stop broke.

The ten session horizon runs to September 17 so the claim is technically still open, but a setup that loses its stop on day one is a loss and I am grading it as one. Nothing was bought, because the account could not afford the shares and no option on it fit.

Here is the part that actually matters. I wrote yesterday that crowding on that name was high, that it was the top pre bell ticker on the retail forums with a televised endorsement behind it, and that crowded gaps give back more than quiet ones.

Then it gave back 5.5 percent from the opening print. Naming the exact risk that later kills the trade and taking the trade anyway is not a win. It is a louder kind of loss.

Second grade, this one on the machinery. I called the regime amber yesterday morning. The S&P closed up 1.05 percent that day, 765.16 to 773.17, and the flip to green only registered after the close.

The rule was cautious on a day that paid to be long. It is mechanical and I am not going to override it on feel, but a miss is a miss and it goes in the book.

The standing failure is still the same one. No weekly grades file exists yet, so every base rate in this system reads unproven at zero samples. I have written that sentence every day for a week and it is the most important open item I have.

None of this is exotic. It is a scheduled task, a research file, and a set of rules strict enough to make the system say nothing when there is nothing to say. I wrote up how the scheduling works, the research agent itself is documented here, and if you want something like it pointed at your own work, that is what I do.

One question I keep chewing on, and I would like an outside opinion. When a system names the exact risk that later kills the trade, and takes the trade anyway, is that a discipline problem or a sizing problem?

Not financial advice. Entertainment and education only.