Zubifi.
my positions →
v0.1.0 · as of · data refreshed 2026-09-09 · static build
Coverage
23
23 tickers
Buys
7
moderately-aggressive
Sells
0
downside markers
Flagged
3
guarded / high-risk
Top own
GOOGL
Own — accumulate: the dip is here, add a tranche
Top swing
OSCR
BUY · +0.34
Macro deskmixedOil above $100, long yields making fresh highs, and a wider Canada trade fight are tightening the macro backdrop even as mega-cap tech remains relatively resilient ahead of next week's Fed decision.
Oil shock broadens into input-cost and consumer-spending pressureHigher-for-longer rates reassert through the long endCanada trade-war escalation adds supply-chain and sentiment riskMega-cap platform and AI leaders still act as relative defensives inside a toughnext: 2026-09-16 FOMC rate decision
Stock · Technology · brief
AMD
Advanced Micro Devices, Inc.
Thesis Bullish · moderate · 6–12mo
BUYthe owner's call leads — buy now; the short-term read BUY is timing, not a veto · moderate · 7/7
Own — and a good entry now
Quality name at a fair price with a live long entry; buy / accumulate. No pullback to buy yet: add a tranche (~1 sh, $521, 3% of account) on weakness into the 440.50–521.09 zone; re-underwrite below ~431.25.
edge unproven — no resolved BUY trades yet (need 30) · report card Q2 FY2026: thesis tracking · guidance raised · conf 85% · fair value ~535.31 (price -3%) · A add below 347.95 · B trim above 722.67 · C re-underwrite 431.25 · quality 6/6 · valuation cheap · implied return ~33% vs 9% hurdle · thesis bullish
Signal history: BUY since 2026-09-08 (was HOLD) · 42 runs since 2026-09-05
Prev close
521.09
Entry
511.84
Entry vs close
-1.8%
Fill odds
~88%
Target
539.59
Stop
493.34
Reward : Risk
1.5:1
Win odds (baseline)
~40%
Size @ 1% risk · gap-aware
7 sh · $3,583
Category: Technology · XLK 20d +1.0%
Price, moving averages & levels
PriceSMA20SMA50EntryTargetStop
The case

Fair value ~535.31 (price -3% vs it). A: add below 347.95 (-33% from here). B: trim above 722.67 (+39% from here). C: re-underwrite below 431.25, hard exit 388.12 (or any price on a broken thesis).

The trade call is to own AMD with a good entry now, based on an edge that is unproven but shows promise, with no resolved buy trades yet. The strategy is to take advantage of the positive trend, strong momentum, and positive sentiment, while the thesis remains bullish. The fundamentals are growing, the analysts are optimistic, and the news is positive. However, the high beta and wide analyst dispersion present risks that make this call more aggressive. Not financial advice.

⚑ Macro: Linked to higher-for-longer rates and AI/tech resilience: semiconductor demand tied to compute remains supportive, but the stock stays sensitive to rate-driven valuation compression.
⚠ High beta (2.48): moves are amplified ~2.5x versus the market.
⚠ Wide analyst dispersion (targets span ~170% of price): low consensus on where this goes.
Headlines & filings
Trade conviction
−1 sell-0.15+0.15buy +1
Factors
Trade factors · ~1–4wk
Trend · w0.20
+0.17
Momentum · w0.18
+0.99
Mean-reversion · w0.10
-0.60
Volume · w0.10
+0.25
Rel. strength · w0.17
+0.49
Sentiment · w0.15
+0.31
Options · w0.10
+0.34
Thesis factors · 6–12mo
Analyst · w0.35
+0.49
Fundamental · w0.30
+0.63
Insider · w0.20
-0.76
Earnings qual. · w0.15
+0.31
Quality & valuation
Quality 6/6 checks
Profitable
Free cash flow positive
Net debt / EBITDA < 3
Covers near-term bills
Growing
Stable sector
Valuation: cheap P/E 33.6 · PEG 0.49
Quarterly report card
Q2 FY2026 · filed 2026-08-04 · gpt-5.4
Change vs prior card is positive but narrower than the headline implies: Data Center is accelerating hard and guidance is strong, while cash conversion softened sequentially and Gaming remains a drag.
revenue: trackingmargins: trackingcash flow: driftingoverall: tracking

AMD can compound if it keeps converting AI and server product strength into sustained Data Center share gains while preserving gross margin and funding the ramp without balance-sheet strain. This quarter supports that thesis through 107% Data Center growth and better margins, but the business is becoming more concentrated in AI/Data Center and requires heavy commitments and execution to justify the valuation.

Guidance
raised
Leadership
4/5
Offering / moat
4/5
Fair value (card)
fair
Confidence
85%
Next report
2026-11-03
Guidance: Q3 FY2026 revenue guided to about $13.0B +/- $0.3B, implying ~41% y/y growth and ~13% q/q growth, with non-GAAP gross margin expected at ~56%; management explicitly expects Data Center sales growth to accelerate in 2H26.
Watch next quarter: Data Center revenue growth and whether it again outgrows company revenue, Non-GAAP and GAAP gross margin progression versus the 56% Q3 guide, Free cash flow margin and operating cash flow conversion as capex and working capital rise, Client processor ASP trends versus unit growth, Gaming revenue decline and any stabilization in semi-custom
  • AI/Data Center ramp slips due to product timing, customer deployment delays, or weaker-than-expected MI/Helios demand
  • Export controls, licensing rules, or geopolitics again disrupt GPU sales
  • Supply chain or manufacturing constraints reduce yields, availability, or delivery timing
  • Large customer concentration and associated warrant structures weaken economics or bargaining power
Analysts & insiders
Consensus
strong buy
Analysts
49
Target lo/mean/hi
365.00 / 615.38 / 1250.00
Implied upside
+18.1%
Actions 90d
2↑ / 0↓
Insider 90d
0 buy / 10 sell
Earnings beat rate
+62.5%
Avg surprise
3.82%
Fundamentals
Market cap
$850.7B
Sector
Technology
Rev growth
+50.1%
Earnings growth
+159.5%
Net margin
+15.6%
Fwd P/E
33.6
PEG
0.49
Beta
2.48
Next earnings: 2026-11-03 (55d)
Sentiment
News (7d)
50 articles
News mix
42+ / 1- / 7o
News score
0.37
Social (StockTwits)
30 msgs
Bull ratio
+52.9%
Social score
0.03
Google Trends
1.00
Wikipedia Δ
0.20
Options
Implied move
+4.3%
Put/Call OI
0.66
ATM IV
+43.0%
Expiry
2026-09-14
Owns record
The owner's calls are graded on the owner's clock: every A/B/C state is recorded once per company per day and scored, as prices arrive, on the return in excess of SPY and of the sector ETF over the next quarter, two quarters and a year, measured from the next session's close. Adds should beat the benchmark; trims and exits should be followed by under-performance. 'Hit' is the share of adds that beat SPY. Live only — fair value depends on point-in-time filings, so there is no replay; a class is 'unproven' until 10 calls have a 1Q grade.
34 calls · 17 companies · since 2026-09-09 · first 1Q grades due ~2026-12-09
classcalls / cosstatus1Q · resolved / vs SPY / vs sector / hit2Q · resolved / vs SPY / vs sector / hit1Y · resolved / vs SPY / vs sector / hit
A · add zone (value)12 / 6pending000
A · dip add4 / 2pending000
wait / hold12 / 6pending000
B · trim zone4 / 2pending000
C · re-underwrite2 / 1pending000
Track record
Every plan is graded against what happened next: did the entry fill within 15 trading days, did the target print before the stop, and what did price do 5 and 20 days later. Win rate is shown next to the driftless baseline (1 ÷ (1 + reward:risk)) — beating it is the only evidence of an edge. No costs or slippage; overlapping windows; current-watchlist survivorship. This record governs the calls: each signal class (signal × conviction) is 'unproven' until 30 live trades have resolved (rolling 120 days); once measured, a class with negative expected value after ~0.2% costs is not presented as a swing. Edge is measured per class × price regime when that sample is large enough, and only trades from the current model (id ec01b77311b8) count — 112 plan(s) from earlier model versions are shown in the totals but never certify the current one. One plan per ticker per day.
Live signalsthe full model's real calls, scored as prices arrive
46 plans · 28 pending · 13 open · fill rate 100% (predicted 88%) · resolved 5: win rate 0% vs 45% driftless baseline · fwd 5d — · fwd 20d —
groupplansfilled / pred.resolvedwin / baselinefwd 5dfwd 20d
By signal
BUY45100% / 88%50% / 45%
SELL1 / —0 / —
By conviction
high6100% / 100%0 / —
low12100% / 81%10% / 43%
moderate28100% / 85%40% / 45%
By price regime
neutral11100% / 84%20% / 51%
ranging22100% / 85%10% / 40%
trending5100% / 100%0 / —
unknown8100% / —20% / 41%
Replay (simulated)price-only sub-model stepped through history — no fundamentals or sentiment
199 plans · 16 pending · 10 open · fill rate 76% (predicted 88%) · resolved 130: win rate 32% vs 39% driftless baseline · fwd 5d +0.4% · fwd 20d -0.0%
groupplansfilled / pred.resolvedwin / baselinefwd 5dfwd 20d
By signal
BUY11576% / 89%7633% / 38%-1.4%-2.6%
SELL8477% / 86%5430% / 39%+2.8%+3.3%
By conviction
high48100% / 100%4613% / 33%-0.0%-1.8%
low5364% / 79%2952% / 44%-0.2%+1.6%
moderate9870% / 86%5536% / 41%+1.1%+0.0%
By price regime
neutral4776% / 86%2841% / 39%-0.7%-2.5%
ranging8064% / 84%4137% / 41%+0.6%-0.6%
trending7290% / 92%6125% / 37%+1.0%+1.9%
Strategy lab — plan rules comparedreplay, simulated: same signals, different entry/stop rules, gap-realistic scoring
entry rulestopplansfilledwin / baseexpectancyavg lossworst 5%losses > 1R
dip_limit×119988%26% / 39%-0.50R-1.20R-1.82R26
dip_reclaim×119976%32% / 39%-0.55R-1.25R-2.13R17
market×1199100%31% / 40%-0.38R-1.12R-1.36R19
dip_limit×1.519988%34% / 50%-0.38R-1.08R-1.43R20
dip_reclaim×1.519976%38% / 50%-0.43R-1.09R-1.36R15
dip_reclaim = price must dip to the level AND close back above it (no resting limit into a crash); dip_limit = a resting limit that fills on the way down; market = next open. Expectancy is the mean realized result per plan in units of planned risk; 'losses > 1R' are gaps through the stop — the losses a 1%-risk sizing quietly assumes never happen.
⚠ Zubifi is an educational research tool. It aggregates public data and applies heuristic scoring. It is NOT financial advice, not a recommendation to buy or sell any security, and makes no guarantee of accuracy. Markets are risky; do your own research and consult a licensed professional.