Daily Desk
Morning edition · Fri 31 Jul 2026 · is there a trade today, and am I too late? · preview snapshot, frozen 21:36 UTC — not auto-updating
Snapshot
0
Actionable
nothing clears the bar
2
Candidates
watch only · no risk levels
2
Moving, no thesis
crowd early
1,077
Top mentions · MU
+37 vs last
World context
SnapshotS&P fut
7,504
+0.42%
Nasdaq fut
28,405
+0.59%
Nikkei
64,362
+4.03%
10Y
4.67%
+1.3%
Gold
4,100
−1.46%
WTI
84.77
+1.41%
Nothing actionable yet — watch only. MU leads chatter on 1,077 mentions (+37), 7 of the top 10 rising, but no candidate carries a stop, target or settled lateness. The desk refuses to call a trade it can't back.
ApeWisdom
live
Alpaca
724
News
live
Crowd
wiring
Lateness meter
UNKNOWN · needs price history
Sector heat
MockSemis
+2.1%
AI infra
+1.4%
Software
+0.3%
Memory
+3.2%
Energy
−0.9%
Fintech
+0.1%
Consumer
−1.1%
Candidates · watch only — no risk levels yet
SnapshotAMZNLong · swingcomposite 0.50
AWS + Bedrock momentum with improving retail profitability into the print.
grounded · cited "strong performance in AWS, Amazon Bedrock…"
Size
$500
Entry
$235.50
Stop
needs history
Target
needs history
AAPLLong · swingcomposite 0.34
Mac Studio clusters running local AI models driving record Mac revenue.
grounded · cited "Mac's record revenue — jumped 29 percent…"
Size
$500
Entry
$333.43
Stop
needs history
Target
needs history
Moving — no thesis yet
Mock · ranking wiring⚡ The crowd is early — surfaced, not buried
RDDT+101 (+13%)biggest mover on the board — no press has caught it yetwatch
SNDK+12 (+3%)memory names heating, thesis not yet gradedwatch
The cut · chatter · alerts
SnapshotWhy they didn't make it
| RDDT | 1.25 | moving — no thesis |
| MU | 0.36 | no thesis — noise |
| NVDA | 0.00 | cooling |
Top chatter
| Sym | Ment. | Δ | News |
|---|---|---|---|
| MU | 1,077 | +37 | 0/3 |
| RDDT | 906 | +101 | — |
| AAPL | 713 | +8 | 2/3 |
| AMZN | 611 | +10 | 2/3 |
Signal feed Mock
15:40RDDT mentions +13% — surfaced to moving lane
15:40AMZN news graded 2/3, added to candidates
12:57MU thesis zeroed — past-tense recap rejected
11:36morning digest delivered · 490 gate
What this can't yet prove
- Deltas are day-over-day, not anomaly detection — 6 of 10 baseline snapshots collected; 10 needed before "unusual" is a claim.
- No ticket carries a stop or target: both need a price series, and only one spot quote per name is stored so far.
- Crowd tone and news thesis are separate signals, never blended. Crowd-prose grading is wiring in now (density solved: 16,819 comments/week).
Watchlist
Your names — heating, flips and alerts surface to the top
Mock — pipeline wiring
2
Sentiment flips
3
Heating
9
Names tracked
+1.4%
Avg 5d move
Alerts
Mock14:20SKHY crowd sentiment flipped bearish — first flip in 12 days
13:05NVDA mentions +142 — moved Quiet → Heating
11:40DRAM memory-cycle chatter accelerating
09:12PLTR new DD thread, sentiment neutral-to-up
Your names
Mock| Symbol | Mentions Δ | Sentiment | Price | 5d | News | Status |
|---|---|---|---|---|---|---|
| NVDA | +142 | bullish ↑ | $182.30 | 1/3 | Heating | |
| PLTR | +58 | neutral | $71.44 | — | Watching | |
| SKHY | −4 | bearish ↓ flip | $14.02 | 1/3 | Alert | |
| DRAM | +31 | bullish ↑ | $28.90 | — | Heating | |
| HBMX | +18 | bullish ↑ | $9.71 | — | Watching | |
| SCHD | +1 | neutral | $29.10 | — | Quiet | |
| VXUS | +2 | neutral | $68.55 | — | Quiet | |
| GOOGL | — | — | $355.22 | — | Held · never traded |
Add a name
MockType a ticker — validated against Alpaca before it's added
TICKER…
Add
⬍ Drop or paste a screenshot
One vision call reads the tickers → confirm before adding. GOOGL can't be added (D-037).
Voices
The few who've actually been right — track record, not tips
Mock — pending density + backtest
6
Proven voices
of 1,240 evaluated
+5.9%
Cohort avg α
vs population +0.4%
61%
Cohort hit-rate
14
Provisional
building record
Proven voices · cleared n≥10 · α>0 @ T+5, walk-forward
Mock@deepvalue_danProven
41
settled
+7.9%
avg α
63%
hit
tenure 6.1y90d: 7
@semi_cycleProven
28
settled
+5.4%
avg α
57%
hit
tenure 4.4y90d: 5
@quietlongProven
19
settled
+6.8%
avg α
68%
hit
tenure 8.2y90d: 3
@newflow_22Provisional
2
settled
—
avg α
n=2
thin
tenure 3.3ybuilding
Cohort proof · do proven voices beat the room, out of sample?
SimulatedProven cohort+5.9% α
Provisional cohort+1.2% α
Population (all callers)−0.4% α
Illustrative until the walk-forward backtest posts. If proven ≈ population out-of-sample, the feature dies — that's the honest test.
Settled record · @deepvalue_dan
Simulated · walk-forward| Date | Ticker | Dir | Entry | SPY | α @ T+5 | Called |
|---|---|---|---|---|---|---|
| 07-24 | MU | Long | $118.40 | +0.9% | +6.2% | "HBM sold out, pricing reset" |
| 07-11 | AVGO | Long | $284.10 | +1.4% | +9.1% | "custom-silicon ramp underpriced" |
| 06-28 | INTC | Short | $31.05 | +2.1% | +4.4% | "foundry losses worse than guided" |
| 06-15 | SMCI | Long | $61.80 | −0.6% | −3.8% | "rack demand into Q3" — missed |
- Visibility ≠ size: substance earns the board, a settled record earns dollars.
- Walk-forward, no look-ahead; karma unused (tenure from the archive). Cohort-level while per-author n is thin.
Performance
Does the signal actually predict? — the merit proof
Pending backtest
—
Hit rate vs SPY
—
Avg α @ T+5
—
Proven vs population
—
Sharpe (sim)
Equity curve vs SPY
PendingAwaiting the walk-forward backtest
Density is solved (16,819 comments/week) and the run is pre-authorized. Real SIMULATED numbers land here after the overnight replay — no placeholder returns shown as fact.
Analytics that will render
PendingHit-rate by conviction tier
Pending
Calibration curve
Pending
Signal decay
Pending
- Everything here ships labeled SIMULATED with its confounders stated — thin baseline, survivorship, correlated calls, selection bias.
- The verdict that matters: do pre-T proven voices beat population after T? A negative result kills the author feature — and that's a valid answer.
Compare
Everyone else follows the money. We read the room.
Draft · public
Options-flow tools
Unusual Whales · Benzinga · the rest
- Follow institutional options flow — money that's already moved
- Count crowd mentions with no per-voice track record
- Blend sentiment into one number
- $50–$200+ / month
Stonkspex
Read the room — and know whose voice to trust
- Grades the crowd's actual theses, with verbatim citations
- Scores individual authors on whether they've been right before
- Two labeled signals, never blended: what's moving · is it real
- Free-first, honest about what it can't yet prove
Feature matrix
Illustrative| Capability | Flow tools | Aggregators | Stonkspex |
|---|---|---|---|
| Crowd mention velocity | — | ✓ | ✓ |
| Graded thesis w/ citation | — | — | ✓ |
| Per-author track record | — | — | ✓ |
| Signals kept un-blended | — | — | ✓ |
| Options flow | ✓ | — | calc |
| Honest degradation | — | — | ✓ |
| Price | $$$ | $$ | free-first |
Informational only · not investment advice · not an RIA · paper only, no live orders · not affiliated with any named product. Figures illustrative Mock until the forward track record accrues.