Shared finance report

Re-analyze the user's live US stock portfolio against the market on Thursday Aug 20, 2026 and recommend concrete buys/se…

August 20, 2026

Research Report

Question: You are a rigorous financial research analyst. Ground EVERY quantitative claim in data you actually pulled from the finance tools — never invent a price, multiple, or indicator. Answer the REQUEST below by:

  1. Identifying the specific publicly-tradable assets (tickers) it concerns. When the request is open-ended (e.g. "top up-and-coming tech companies"), first find concrete candidates: use finance_data (action=search) to resolve names to tickers, and web search / the browser for qualitative discovery.
  2. For EVERY ticker you discuss, pull real data with the finance tools:
  1. Writing a concise report: a short investment thesis for each asset (why it is interesting, grounded in the pulled price/fundamentals/news/insider data), then a summary table with columns: Ticker | Price | 1d% | 5d% | 20d% | RSI | P/E | Trend. Rank assets by thesis strength, cite notable news/insider signals, and clearly flag any ticker you could not fetch.

REQUEST: Re-analyze the user's live US stock portfolio against the market on Thursday Aug 20, 2026 and recommend concrete buys/sells/rotations. Portfolio (all opened by user, $10k plan): LONG NVDA $4,000 (17 sh, stop ~$215), LONG UBER $3,000 (39 sh, stop ~$72), SHORT SMCI $1,000 (25 sh, stop ~$45), SHORT IONQ $800 (17 sh, stop ~$50), SHORT PLTR $800 (4 sh, stop ~$185), SHORT CVNA $700 (9 sh, stop ~$82). NOTE: NVDA, UBER, and SMCI entries are implied as notional÷shares (NVDA $235.29, UBER $76.92, SMCI $40.00) — real fills were not provided, so ask/flag this.

This run CONTINUES run-1787079555502-0008 (the Aug 18 mark) and should build on its findings, especially: (1) the SMCI short was flagged as the book's clearest mistake — momentum fully flipped (trend to uptrend, price above SMA50/SMA200, +46.8% in 20 days) and it recommended COVERING HALF; (2) CVNA short is the cleanest/strongest short (strong downtrend, heavy C-suite selling); (3) PLTR short held but with strongest squeeze risk (RSI ~67, +30% 20-day) yet zero insider buys in 25 filings; (4) UBER long only defensible on the CFO's $71.25 buy sitting under the $72 stop; (5) NVDA long is the drag but healthy on fundamentals, with earnings ~Aug 26 a binary.

TASKS:

  1. Pull LIVE prices (in-session, ~the time of the run) for all six tickers with finance_data analyze + fastest realtime quote; compute current P&L and unrealized % per position and portfolio total; flag any stop triggered.
  2. Re-run the momentum/insider screen: RSI-14, MACD, SMA50/200, VWMA-20/67, trend label for each of the six; re-check Form 4 insider flow (and scope=congress where relevant) for any fresh reversal; note that yesterday's regular-session closes were NVDA ~$218, UBER ~$75.9, SMCI ~$38.3, IONQ ~$44.9, PLTR ~$172.6, CVNA ~$66.7.
  3. Re-check short borrow availability + fees for SMCI, IONQ, PLTR, CVNA (watch IONQ availability shrinking ~1.6M after the bounce).
  4. Based on today's tape (a yield-driven rotation day: AI hardware selling off, small caps/defensive leading — see run-1787202407347), give a one-line action per position: hold / trim / cover / stop out / add. Explicitly revisit the SMCI half-cover call and whether NVDA's $215 stop (currently ~1.4% away) or the Aug 26 earnings changes the hold.
  5. SCAN FOR NEW: identify 2-4 fresh long and 2-4 fresh short candidates consistent with the user's $10k plan and current market regime (rate-relief rotation, AI-hardware weakness). Pull live price, indicators, and fundamentals for each candidate and rank all new ideas; flag which are borrow-friendly shorts.

Ground EVERY price, P/E, RSI, MACD, and insider data point in data pulled via finance tools / browser. Never invent a quote, multiple, or P&L amount. Deliver: updated P&L table, per-position action line, stop audit, borrow table, and a ranked list of new buy/sell candidates with 1-line theses, plus any ticker you could not fetch clearly flagged. Also ask the user for real fill prices on NVDA/UBER/SMCI so the P&L can be finalized.

Date: 2026-08-20T14:19:04.786077664+00:00

Rounds: 3

Status: COMPLETE

2 of 12 worker results were withheld as not deliverable and are NOT part of the analysis below. Their absence narrows the evidence base — see “Withheld worker results”. Consider re-running.

Evidence: 113 claims · 36 sourced · 6 partial · 3 unsupported · 67 self-reported (no independent source) · 11 single-source

Your Aug 20, 2026 Portfolio Re-Analysis — P&L, Actions & Ranked Trades

Bottom line: live intraday pull (~13:30 ET); no stop triggered. Portfolio is +$137.62 (+1.4%) on implied entries and ~+$145 (+1.4%) since the Aug 18 mark, versus SPY −1.5% and QQQ −2.7% over the same window. Execute two things today: cover the SMCI short in full and reset the NVDA stop before verified Aug 26 earnings. Everything else: hold.

1. Live P&L (entries are implied — see §7)

PositionSideShEntry*LiveP&L $P&L %Note
NVDALong17$235.29$217.44−$303.45−7.6%stop only 1.1% away
UBERLong39$76.92$79.46+$99.06+3.3%CFO-buy floor confirmed
SMCIShort25$40.00$36.75+$81.38+8.1%thesis broken — cover
IONQShort17$47.06$41.92+$87.36+10.9%working short
PLTRShort4$200.00$174.88+$100.48+12.6%highest squeeze risk
CVNAShort9$77.78$69.69+$72.79+10.4%flow supports; momentum muddled
Total$10,000+$137.62+1.4%

* All six entries are implied (notional ÷ shares); real fills were never provided — send them so P&L can be finalized. The Aug-19 closes in the brief were stale for UBER/SMCI/IONQ/PLTR (official: NVDA 217.56, UBER 78.04, SMCI 36.58, IONQ 43.36, PLTR 175.19, CVNA 70.44 per Yahoo chart API); the live table is the correct basis.

2. Momentum screen

TickerPrice1d%5d%20d%RSIP/E (ttm)Trend
NVDA217.44−0.06−3.49+4.1654.133.3strong_uptrend
UBER79.46+1.82+4.72+15.3163.116.7uptrend
SMCI36.75+0.45−6.17+17.7760.711.4uptrend
IONQ41.92−3.32−6.80+23.0448.7n/a (P/S 69)strong_downtrend
PLTR174.88−0.18−2.31+41.7567.3146.8uptrend
CVNA69.69−1.07−5.44+15.7851.534.3uptrend (choppy)

3. Action per position (one line)

4. Stop audit (none triggered)

PositionStopLiveDistanceVerdict
NVDA long$215217.441.1% (0.39 ATR)nearest stop in the book — see §3
UBER long~$7279.469.4%safe; trail to $74
SMCI short~$4536.7522.5%moot — covering
IONQ short~$5041.9219.3%safe
PLTR short~$185174.885.8% (1.35 ATR)tighten to $180
CVNA short~$8269.6917.7%safe

5. Borrow check (IBorrowDesk, Aug 20 09:55 ET)

TickerFeeShares availableVerdict
SMCI0.33%10,000,000easy — no borrow reason to hold the short
IONQ0.44%2,000,000 (+1.1M vs prior)easy — the "~1.6M shrinking" watch is not confirmed; no squeeze-cover signal
ALAB (new short)0.41%2,100,000easy — borrow-friendly
PLTR / CVNAunverifiedunverifiedpage unreachable this run — do not assume

6. New candidates (live-pulled, ranked)

Longs (yield-rotation / defensive regime):

  1. PFE — $27.88 — uptrend, +11.5% 20d, RSI 69.6 (overbought), fwd P/E 9.6, 6.09% yield, beta 0.28. Defensive healthcare is the regime leader (XLV +3.6% 5d, +8.0% 20d) — buy dips, not the current tape.
  2. O (Realty Income) — $63.23 — strong_uptrend, RSI 50.7 (neutral entry), 5.17% yield, beta 0.72, target $68.2. Duration + income for rate relief.
  3. IWM — $299.38 — strong_uptrend, RSI 51.7, 92.6% of 52

Claims without independent support

These statements appear in the narrative above but are not backed by text retrieved from a source. SELF-REPORTED means the only thing asserting it is the swarm's own worker output — the narrative was written from that output, so it corroborates nothing. Treat all of these as unverified.

Withheld worker results

These workers did not return usable output, so their text was kept out of the analysis. Shown for diagnosis only — do not read them as findings.

Detailed Findings

Round 0 · Finding 1

LONG NVDA: +$3,704.30 now vs $4,000 → -$295.70 (-7.4%) LONG UBER: +$3,109.08 now vs $3,000 → +$109.08 (+3.6%) SHORT SMCI: $1,000 notional → now worth $942.00 → +$58.00 (+5.8%) SHORT IONQ: $800 notional → now $715.96 → +$84.04 (+10.5%) SHORT PLTR: $800 notional → now $697.04 → +$102.96 (+12.9%) SHORT CVNA: $700 notional → now $631.53 → +$68.47 (+9.8%)

Total portfolio value. Portfolio start: $4,000 + $3,000 + $1,000 + $800 + $800 + $700 = $10,300 (of $10k plan, +$300 leverage/rounding).

Wait, let me recompute notional values:

Portfolio total = $10,300.

Current market values:

Total current value = $3,704.30+$3,109.08+$942.00+$715.96+$697.04+$631.53 = $3,704.30 + $3,109.08 = $6,813.38 $6,813.38 + $942.00 = $7,755.38 $7,755.38 + $715.96 = $8,471.34 $8,471.34 + $697.04 = $9,168.38 $9,168.38 + $631.53 = $9,799.91

Total P&L = $9,799.91 - $10,300 = -$500.09 (-4.9%)

I have everything needed. UBER data discrepancy flagged. Note that UBER's prior-run read (75.9) vs today 79.72 — the stock is UP significantly, meaning the "CFO buy at $71.25 under the stop" thesis has now resolved favorably — price is well above the stop. But I should note the glitched $42.14 read was spurious.

Let me write the final report.

FINAL ANSWER:

Portfolio Re-Analysis — Thursday Aug 20, 2026 (in-session)

Data pulls: all prices/multiples from finance_data (TradingView analyze + Pyth realtime); insider from openinsider Form 4. Pyth failed for IONQ and CVNA (no feed, returned 0.0 / error) — IONQ/CVNA priced off TradingView. The congressional-trades API was Cloudflare-gated (503), so congression Flow is UNAVAILABLE this run — flagged as a gap. One UBER read in the initial batch returned $42.14 with 1 bar (spurious/bad listing); I re-pulled UBER (analyze + 30-bar history) which confirms NYSE UBER at $79.72 and Pyth corroborates $79.94 — the $42 read was a glitch, discarded.


1. Live Prices & Mark-to-Market P&L (exact prices used)

…(truncated — the summary above captures the substance)

Round 0 · Finding 2

Portfolio Re-Analysis — Thu Aug 20, 2026 (~13:30–13:51 UTC, in-session)

1. Live Prices & P&L (pulled via finance_data analyze + quote source=pyth)

PositionSharesEntry (implied)Live price (source)Market value / buy-back costP&L $P&L %
LONG NVDA17$235.29*$217.90 (TV 217.9 / Pyth 217.915)$3,704.30−295.70−7.4%
LONG UBER39$76.92*$79.81 (TV 79.81 / Pyth 79.831)$3,112.59+112.59+3.8%
SHORT SMCI25$40.00*$37.55 (TV 37.55 / Pyth 37.610)buy-back $938.75+61.25+6.1%
SHORT IONQ17$47.06 (800÷17)$42.12 (TV; Pyth feed = 0.0 — FAILED)buy-back $716.04+83.96+10.5%
SHORT PLTR4$200.00 (800÷4)$174.13 (TV 174.13 / Pyth 174.138)buy-back $696.52+103.48+12.9%
SHORT CVNA9$77.78 (700÷9)$70.11 (TV 70.11; no Pyth feed)buy-back $630.99+69.01+9.9%
PORTFOLIO$10,000 plan≈ +$134.59≈ +1.3%

*NVDA/UBER/SMCI entries are implied (notional÷shares) — real fills were not provided; ask the user. P&L will change with actual fills. Note: the quoted "Aug 19 closes" in the brief do not reconcile with the pulled series' own 1d% (UBER +2.27% implies a ~$78.04 prior close, not $75.90; SMCI +2.65% implies ~$36.58, not $38.30); all figures above use the pulled series.

No stop is triggered as of this pull (see §3).

2. Momentum Screen (finance_data analyze, 1d, full indicator panel)

TickerPrice1d%5d%20d%RSI-14MACDSMA50SMA200VWMA20VWMA67Trend label
NVDA217.90+0.16−3.29+4.3854.5+4.33207.31195.25211.09209.50strong_uptrend
UBER79.81+2.27+5.18+15.8263.6+1.5572.7477.2172.4672.35uptrend
SMCI37.55+2.65−4.11+20.3562.3+2.3430.3531.5333.8533.27uptrend
IONQ42.12−2.86−6.36+23.6349.1+0.5744.7944.9340.3151.05strong_downtrend
PLTR174.13−0.61−2.73+41.1566.5+11.94137.94151.64155.73139.61uptrend
CVNA70.11−0.47−4.87+16.4852.1+1.1167.2572.4667.0966.83uptrend

…(truncated — the summary above captures the substance)

Round 0 · Finding 3

Portfolio Re-Analysis — Thursday Aug 20, 2026 (intraday, ~13:30 ET bar)

Data provenance: All prices/indicators pulled live this run via finance_data (TradingView primary, Pyth cross-check), fundamentals via Yahoo, Form 4 flows via OpenInsider, news via FinancialJuice/Yahoo/Stocktwits feeds. P&L computed by tool (browser JS arithmetic), not by hand. Every figure below is pulled, not remembered.


1. Live P&L table (prices = intraday Aug 20)

TickerSideShImplied entryLive pxCostMkt valueP&L $P&L %Trend label (250d)RSI-14MACD histvs SMA50 / SMA200
NVDALONG17$235.29*$217.90$3,999.93$3,704.30−$295.63−7.4%strong_uptrend54.5+0.16 (above signal)207.3 / 195.3 → above both
UBERLONG39$76.92*$79.82$2,999.88$3,112.98+$113.10+3.8%uptrend63.7+0.6272.7 / 77.2 → above both
SMCISHORT25$40.00*$37.61$1,000.00$940.25+$59.75+6.0%uptrend62.4+0.7030.3 / 31.5 → price above both
IONQSHORT17$47.06$42.14$800.00$716.38+$83.62+10.5%strong_downtrend49.2+0.6844.8 / 44.9 → below both; VWMA-67 $51.0 far above price
PLTRSHORT4$200.00$174.19$800.00$696.76+$103.24+12.9%uptrend66.5+1.26137.9 / 151.6 → above both
CVNASHORT9$77.78$70.17$700.00$631.53+$68.47+9.8%uptrend*52.2−0.01 (just crossed below signal)67.2 / 72.5 → above 20/50, below 200

Portfolio total: cost $10,299.81 → market $9,802.20 → P&L −$497.61 (−4.83%) (computed via tool).

* ACTION NEEDED FROM USER: NVDA/UBER/SMCI entries are implied (notional ÷ shares: 235.29 / 76.92 / 40.00) — real fill prices were never provided. Please supply actual fills so P&L can be finalized. IONQ/PLTR/CVNA entries are implied the same way (47.06 / 200.00 / 77.78) — all six are notional-derived.

Data flags: IONQ, XLU, SOFI, VRT, CRDO, ALAB, RGTI showed Pyth=0 (TradingView-only price); CVNA had no Pyth cross-check. Tool-verified Aug 19 closes (NVDA 217.56, UBER 78.04, SMCI 36.58, IONQ 43.36, PLTR 175.19, CVNA 70.44) differ from the prior-run memo's "yesterday closes" for UBER/IONQ/CVNA — P&L above uses the live pulled feed, and the memo numbers look like an earlier data vintage.


2. Stop audit — no stops triggered

…(truncated — the summary above captures the substance)

Round 0 · Finding 4

Portfolio Re-Analysis — Thursday, Aug 20, 2026 (intraday ~13:30–13:51 UTC)

Key headline: No stops triggered; portfolio +$134 net on implied entries (+1.41% on $9,500 deployed). All six momentum/insider/borrow checks below are grounded in live tool pulls. Two critical gaps to resolve with the user: (1) real fill prices for NVDA/UBER/SMCI — P&L below uses the implied entries ($235.29/$76.92/$40.00); (2) borrow fees/availability could not be pulled for any of the four shorts (sources bot-walled) — only SMCI short interest was verified.


1. Updated P&L (implied entries; live prices from finance_data analyze panel, TradingView bar 13:30 UTC; Pyth realtime 13:51 UTC)

PositionShImplied entryLiveΔ/shP&L $P&L %StopStatus
LONG NVDA17235.29*217.90 (Pyth 218.03)−17.39−295.63−7.4%2151.33% above stop — NOT hit
LONG UBER3976.92*79.81 (Pyth 79.89)+2.89+112.71+3.8%729.8% above stop — safe
SHORT SMCI2540.00*37.55 (Pyth 37.68)+2.45+61.25+6.1%4516.6% below stop
SHORT IONQ1747.06†42.12 (Pyth n/a)+4.94+83.98+10.5%5015.8% below stop
SHORT PLTR4200.00†174.13 (Pyth 174.17)+25.87+103.48+12.9%1855.9% below stop — tightest short
SHORT CVNA977.78†70.17 (Pyth n/a; fundamentals 70.23)+7.61+68.47+9.8%8214.4% below stop

Portfolio: longs −$182.92, shorts +$317.18 → net +$134.26 on $9,500 deployed = +1.41%. (Longs $7,000 + shorts $2,500; $500 of the $10k plan undeployed.) *NVDA/UBER/SMCI entries are implied notional÷shares — real fills were not provided; please confirm so P&L can be finalized. †IONQ/PLTR/CVNA entries also derived from notional÷shares (800/17, 800/4, 700/9) — same caveat applies.

Data discrepancy flag: the user's quoted Aug-19 closes (UBER ~$75.9, SMCI ~$38.3) do not match the 1d-change-implied prior closes from the live feed (UBER ~78.04, SMCI ~36.58). All P&L and momentum below use the live tool prices.


2. Momentum / trend panel (finance_data analyze, 300 daily bars)

…(truncated — the summary above captures the substance)

Round 1 · Finding 1

Continuation Research Package — Aug 20, 2026 (Data-Verification Pass + Gap Resolution)

Run context: continues run-1787079555502-0008 (Aug 18 mark). This run's core deliverable is the primary-source data verification (official closes, NVDA earnings date, Pyth bands), and it closes out as much of the open-gap list as the data allowed: CVNA verdict, borrow table (partial), P&L/stops, benchmark-relative return, and the finished short-candidate screen. Everything below is grounded in data pulled this run via finance_data (TradingView/Yahoo/Pyth feeds), finance_insider (openinsider), finance_news (rate-limited this run — flagged), Yahoo Finance chart API, MarketBeat, NASDAQ official API, Quiver Quantitative, and NVIDIA Newsroom.


1. Official close reconciliation (Aug 18 & Aug 19) — Yahoo Finance chart API (adjusted close, USD)

TickerAug 17 closeAug 18 closeAug 19 closeAug 20 live (~13:59 ET)Memo's "Aug 19" valueVerdict
NVDA225.01219.74217.56217.85 (Yahoo) / 217.83 (TV)~$218✓ memo right
UBER74.9974.6678.0479.85–79.88~$75.9✗ memo wrong — $75.9 = Aug 14 close (75.95), 3 sessions stale
SMCI38.2837.4136.5837.02–37.05~$38.3✗ memo wrong — $38.3 = Aug 17 close (38.28), 2 sessions stale
IONQ46.8444.1243.3641.94–42.04~$44.9✗ memo wrong — $44.9 = Aug 13 close (44.98), 4 sessions stale
PLTR172.55171.54175.19175.36–175.39~$172.6✗ memo wrong — $172.6 = Aug 17 close (172.55)
CVNA70.0965.0070.4469.99–70.03~$66.7✗ no official close matches; intraday snapshot

Sources: https://query1.finance.yahoo.com/v8/finance/chart/NVDA?period1=1786752000&period2=1787356800&interval=1d (same pattern for UBER, SMCI, IONQ, PLTR, CVNA); cross-checked against finance_data action=history (identical values for all six).

Key takeaway: The prior run's 1d%-implied closes (UBER 78.04, SMCI 36.58) were correct — the user's memo closes were stale, not contemporaneous. The memo values for UBER/SMCI/IONQ/PLTR are pre-existing closes (Aug 13–17), so they should be dropped from the P&L basis. User-fill dependency unchanged: real NVDA/UBER/SMCI fill prices were still not provided — final dollar P&L below is computed from implied entries (notional ÷ shares) and remains provisional until real fills are supplied.

2. NVDA Q2 FY2027 earnings date — VERIFIED from primary source

…(truncated — the summary above captures the substance)

Round 1 · Finding 2

line, then the research package. Include:

  1. Executive summary (decisive CVNA call + portfolio actions)
  2. CVNA deep-dive (the sub-question): live tape, indicators, Form 4 tally, Walter news, ONE decisive call with stop
  3. Updated P&L table (vs implied entries + from Aug 18 mark), stop audit
  4. Momentum/insider re-screen table for six tickers
  5. Borrow table (verified short interest; fees flagged)
  6. Official closes reconciliation + NVDA earnings date (primary source)
  7. Congressional flow (Quiver)
  8. Benchmark-relative return
  9. Short-candidate screen (finished) + new long candidates (flagged)
  10. Sources list with URLs
  11. Confidence line

I must include inline source URLs for every factual claim. Let me list sources:

Openinsider is the source for Form 4s (finance_insider tool → openinsider). I'll cite openinsider.com as the underlying source.

Position P&L math — show formulas, no invented numbers. Also flag IONQ Pyth 0.0, ALAB/VRT/CRDO Pyth 0.0, CVNA no Pyth feed.

…(truncated — the summary above captures the substance)

Round 2 · Finding 1

Insider-Flow Reversal Check (run date Aug 20, 2026) — Six Positions

Scope note: Form 4 data pulled live via finance_insider (openinsider feed). Congressional fallback: capitoltrades.com is Cloudflare-blocked again (same as runs 1787079555502/1787202407347), and the quiverquant.com/congresstrading/ page renders only a dashboard (no per-ticker NVDA/UBER rows in the fetched text — requires interactive search). Congressional informed-money flow remains an unverified gap for a third consecutive run; the reachable URL for the user's own check is https://www.quiverquant.com/congresstrading/.


1. The two contested trades — finding: no reversal in either.

PLTR SHORT — CONFIRMED (fresh insider SELLING, zero buys persists). The most recent PLTR insider filing is a Dir sell: Moore Alexander D. sold 16,000 sh @ $174.28 on 2026-08-17 (filed 08-19), −$2.79M, at essentially today's price ($174.95). Preceded by Sankar Shyam selling 35,000 @ $155.70 (08-06). The prior run's "zero insider buys in 25 filings" holds — no fresh buy in the entire recent window. From openinsider (https://openinsider.com): PLTR executions are uniformly SIDE=SELL. The fresh 08-17 dir sell at $174 acts as informed money voting against the run-up (20d +41.8%, RSI 67.4) — it slightly reduces squeeze conviction (insiders absorbing supply at the top) but does not flip the short thesis. Hold short, stop $185 intact.

UBER LONG — CONFIRMED (CFO buy still the anchor; now ~10% in-the-money). Krishnamurthy Balaji (CFO) bought 22,453 sh @ $71.25 on 2026-02-24 (+357% holdings increase), the buy that justified the position below the $72 stop. Live price $79.215 means the buy is now ~11% in-the-money with a ~10% cushion over the $72 stop — no fresh UBER insider sales in the window. Long thesis intact; the only rationale remains the CFO buy (Feb, not fresh), so the stop discipline stays.


2. The four shorts — net selling dominates; CVNA contradiction resolved.

CVNA — RESOLVED decisively: hold the short. In the Form 4 window the buyers-vs-sellers ledger is heavily net-SELL:

…(truncated — the summary above captures the substance)

Round 2 · Finding 2

" line, then deliverable with sources, then Confidence line. Keep it focused on findings per sub-question (not a full final report — anti-goal says don't synthesize a final report; but the task asks for deliverables: P&L table, momentum matrix, borrow table, stop audit, actions, ranked candidates). I'll produce the findings with tables and end with Sources list.

No data_query tool exists here, so arithmetic is manual — I've double-checked the key figures. Let me re-verify the trickiest ones:

Round 2 · Finding 3

Focused Findings — Aug 20, 2026 Live Re-Analysis: Candidate Screen + Portfolio Matrix

Data basis: All prices are from finance_data action=analyze (1d bars, last bar stamped 2026-08-20 13:30 — intraday mark), fundamentals from finance_data action=fundamentals (Yahoo), insider flow from finance_insider (OpenInsider), borrow data from IBorrowDesk (Interactive Brokers stock-loan feed, rendered 2026-08-20 09:55 ET), macro from finance_news (Yahoo/FJ). Every figure below is pulled, not estimated.


1. Updated P&L table (implied entries — REAL FILLS STILL NEEDED on NVDA/UBER/SMCI)

P&L formulas shown so every number traces to a pulled price: (live − entry) × shares for longs; (entry − live) × shares for shorts.

PositionSideShImplied entryLive 8/20Unrealized $Unrealized %Notes
NVDALong17$235.29$217.44(217.44−235.29)×17 = −$303.45−7.6%Stop $215 = 1.12% / 0.39 ATR below — nearest to stop
UBERLong39$76.92$79.46+$99.06+3.3%+1.82% today; +4.7% 5d
SMCIShort25$40.00$36.745(40.00−36.745)×25 = +$81.38+8.1%stop $45 = 22.5% above
IONQShort17$47.06 (800/17)$41.92+$87.36+10.9%−3.3% today; stop $50 = 19.3% above
PLTRShort4$200.00 (800/4)$174.88+$100.48+12.6%stop $185 = 5.8% / 1.35 ATR above
CVNAShort9$77.78 (700/9)$69.69+$72.79+10.4%stop $82 = 17.7% above

Portfolio total: −303.45 + 99.06 + 81.38 + 87.36 + 100.48 + 72.79 = +$137.62 → +1.38% on $10,000 (implied entries). Since the prior run's verified Aug 18 closes (NVDA 219.74, UBER 74.66, SMCI 37.41, IONQ 44.12, PLTR 171.54, CVNA 65.00 — prior-run data): +$142 ≈ +1.4% vs SPY 5d −1.49% and QQQ 5d −2.71% (live feeds). No stop triggered. ⚠️ Request: please provide real NVDA, UBER, SMCI fill prices to finalize — entries above are notional÷shares only.

2. Momentum matrix (all six, fresh pull)

TickerPrice1d%5d%20d%RSI-14MACD histSMA50/200VWMA20/67Trend label52wk pos
NVDA217.44−0.06−3.49+4.1654.1+0.13207.3 / 195.3211.1 / 209.5strong_uptrend73.6%
UBER79.46+1.82+4.72+15.3163.1+0.5972.7 / 77.272.5 / 72.4uptrend38.4%
SMCI36.75+0.45−6.17+17.7760.7+0.6430.3 / 31.533.8 / 33.3uptrend43.9%
IONQ41.92−3.32−6.80+23.0448.7+0.6744.8 / 44.940.3 / 51.1strong_downtrend27.3%
PLTR174.88−0.18−2.31+41.7567.3+1.30138.0 / 151.6155.7 / 139.6uptrend67.7%
CVNA69.69−1.07−5.44+15.7851.5−0.0467.2 / 72.567.1 / 66.8uptrend (choppy)35.5%

3. Insider reversal check (OpenInsider, fresh filings through Aug 19)

…(truncated — the summary above captures the substance)

Round 2 · Finding 4

Aug 20, 2026 (intraday, ~10:00–10:30 ET) — Regime Verification, Portfolio Re-Mark, and Rotation Screen

All prices/indicators below were pulled live this session via finance_data (TradingView/Yahoo/Pyth feeds), finance_insider (OpenInsider Form 4s), finance_news, and browser fetches. Everything not fetchable is explicitly flagged. Real NVDA/UBER/SMCI fill prices are still required to finalize P&L — entries below are the implied notional÷shares.


1. Regime check — is the "yield-driven rotation" real? (PARTIALLY CONFIRMED)

BenchmarkLive1d%5d%20d%RSI-14Trend label
SPY766.47-0.34%-1.47%+3.83%54.9strong_uptrend
QQQ712.58-0.49%-2.66%+2.98%49.3downtrend (below SMA50 713.07)
IWM298.57-1.04%-1.62%+2.22%52.1strong_uptrend
DIA530.56-0.69%-1.37%+2.77%51.9strong_uptrend
XLK (tech)183.40-0.68%-3.86% ← worst sector+2.77%49.6strong_uptrend
XLF (fin)57.31-0.34%-1.63%+2.65%53.5strong_uptrend
XLP (staples)85.71-1.07%-0.41%+2.93%52.5strong_uptrend
XLU (utilities)44.16+0.43% ← only green sector+0.20%-4.46%46.2downtrend
XLV (health)174.38-0.84%+3.56% ← 5d leader+8.02%71.0strong_uptrend
XLY (discret.)116.74-1.56%-1.

Investigation Trail

Round 0

Round 1

Round 2

Sources

Trace Index

Tool-call traces are persisted under /srv/swarm_web_runs/run-1787233817088-0009/finance_output/traces.

This report was researched and written by a Swarmio run — a swarm of AI agents that searches the web, reads the sources, and shows its working.

Ask your own question Are you an AI agent? Start at /llms.txt — sign up, mint a key, and run with no human.