GOOG Alphabet Inc - Trade Plan — 2026-05-20

TL;DR

LONG GOOG via spot equity, limit entry $370.00, stop $365.40 (hard), TP1 $398, TP2 $420, horizon 45 trading days (carry through the ~July 1 DOJ remedy ruling), size 67 shares = $24,790 notional = 2.48% of assumed $1.0M NAV. This is a wait-for-pullback long sized at the memory concentration cap (2.5% NAV), not standard conv-3 sizing — the system already carries three correlated mega-cap tech longs (MSFT, AAPL, AMZN). Per-trade risk is just 0.031% NAV ($308) because the stop is narrow ($4.60/share) and the cap binds size well below what fixed-fractional risk would produce. R:R to TP1 is 6.1:1; to TP2 is 10.9:1. No market order at current $384.90 spot under any circumstance — if the pullback to $370 doesn't fill within 30 days, the trade expires unfilled and is re-evaluated.

Source verdict

"LONG — conviction 3/5. Limit entry $370-$375. Stop $365.40 (2×ATR below spot per technical.md). Hard sizing cap 2.0-2.5% NAV per memory.md concentration discipline — not the 3-4% a standalone conv-3 setup would warrant. This is a directional research view; the trader composes the plan and the risk committee stress-tests it. No market order at the $384.90 spot under any circumstance."

— research-verdict.md, decision section

Direction & instrument

LONG via spot equity (common, GOOG Class C shares).

Considered and rejected: structured options overlay. The options-flow.md analyst correctly flagged that IV is cheap (IV 32.5% vs HV20 40.0%, spread −7.5pp) and that a bull call spread or risk reversal would be a more capital-efficient expression of the directional view, taking advantage of cheap IV in the body of the curve (DTE 9-58 is flat at ~32-33%). I considered three structures:

1. Long Jul-17 (DTE 58) $385/$420 bull call spread. Approximate debit: $11.50 per share ($1,150/contract). Max profit at $420: $23.50 ($2,350/contract). Risks the full debit if GOOG closes below $385 on Jul 17. Would carry through both the June CPI/FOMC and ~July 1 DOJ ruling but not Q2 earnings (Jul 29). Capital efficiency: ~2× the max-payoff/cost vs equivalent equity exposure, with bounded downside.

2. Long Jul-17 $370 calls outright. Cleaner expression of the verdict's $370 entry view; pays for elevated call IV from the call-skew read (call IV > put IV at OTM strikes).

3. Jul-17 risk reversal: long $385 call / short $370 put. Cost-efficient (puts are cheap on the skew); exposes to $370 if GOOG falls, which is exactly the entry zone we want anyway. Effectively a synthetic-long with no premium outlay but unbounded downside below $370 unless paired with a deeper put.

Why I am declining the options overlay and going with spot:

• The verdict is explicitly a limit entry at $370, not a directional bet at $384.90 spot. The whole structure of the thesis is "wait for the pullback, buy where the technical level supports a defined stop." An options overlay collapses that price-discipline into a premium-paying instrument that pays off regardless of entry price — it abandons the entry discipline the verdict requires.
• The position is sized to 2.48% NAV at the cap. Spot at this size has trivial capital deployment ($24,790) and is the cleanest instrument for the risk committee to model. Adding Greek risk (theta decay through the 45-day horizon, vega exposure to a VXX spike on June CPI) on top of the existing 3-mega-cap-tech-long correlation problem is layering complexity on an already-correlated basket.
• The journal and backtest infrastructure are spot-equity focused. Logging an options structure introduces P&L measurement complications (early assignment, IV-marking) that the system is not yet calibrated for.
• The IV-cheap signal is a structure recommendation, not a trade recommendation per the options-flow analyst's own caveat. The directional view is the verdict; the instrument decision is the trader's, and conservative wins here.

If the risk committee wants to push back on this and route the trade through a Jul-17 $385/$420 bull call spread, the structure is documented above and switchable. The default is spot.

Entry

• Method: limit
• Price: $370.00 (mid of $368-$375 technical-analyst zone)
• Validity: GTC, 30 calendar days. If the limit is not hit by 2026-06-19, the order cancels and the verdict is re-evaluated against the then-current chart and macro state.
• No partial-fill scaling. The position is small enough (67 shares) that scaled entries at $370/$372/$374 would produce 22-share fills that are not operationally meaningful. Single-clip limit at $370 is the right granularity.

Why not market at $384.90 spot: The research verdict is explicit: no market order at spot. The technical setup is a pullback in uptrend, not a strong uptrend. The MACD has crossed bearish (technical.md: -1.44 histogram, 10th consecutive decline), RSI has rolled from 83.3 → 62.3, and the 2026-05-19 down day printed on 1.35× the 50-day average volume. Chasing $384.90 here means buying into a fresh momentum deceleration on heavy volume — the opposite of the risk-controlled long the thesis demands.

Why $370 specifically and not $372 or $375:

• $370 is the mid-zone of the technical-analyst's $368-$375 entry range
• $370 sits between the 38.2% Fibonacci retracement of the April-May leg ($375.30) and the 50% retracement ($366.80) — the structural pullback target
• $370 is below the SMA20 ($375.79) and the 2026-05-12 swing low ($379.72), meaning a fill at $370 implies both near-support levels have already given way. This is feature, not bug: the entry only triggers if the pullback is real, and the stop at $365.40 sits just below the next structural level ($366 zone / 50% Fib).
• $370 + the $4.60 stop produces the highest R:R configuration in the entry zone (6.1:1 to TP1 vs ~5.0:1 if entry were $372).

Stop

• Price: $365.40
• Type: hard stop (close-only execution; intraday wicks below $365.40 that recover by close do not trigger)
• Justification:
  • 2×ATR(14) below 2026-05-19 close of $384.90 per technical.md (ATR = $9.75)
  • From the $370 entry, the stop is $4.60 (1.24%) — equivalent to 0.47×ATR from entry, which is admittedly tight in absolute terms but...
  • ...the stop snaps to the structural floor around $366 (the 50% Fibonacci retracement zone and the bottom of the technical.md "$368 gap zone") rather than being a pure volatility multiple. A break below $365.40 means the entire pullback-buy thesis is wrong: support didn't hold, the regime has flipped to topping, and there is no structural reason to be long here.
  • Cross-check against vol-implied minimum: daily 1σ at HV20 = 41.1% is $384.90 × 0.411/√252 = $9.98/day. A stop of $4.60 from entry is 0.46σ daily — well inside one day's expected move. This would be noise-vulnerable if entry were at $384.90 spot, but the entry is at $370, and the stop sits at the structural support level the pullback itself targets. The geometry is "we are buying because $366-$370 should hold; if it doesn't, we are wrong."
  • Stop is tighter than the worst plausible drawdown in risk-assessment terms. A CPI-shock scenario (per macro-factor.md: VXX +20%, TLT −2% combined for -8 to -14% factor drag) would move GOOG -3 to -5% from $370, taking it to $352-$359 — through the stop. That is the risk-committee's job to challenge; my position-sizing response is that the position is so small (2.48% NAV, $25K notional) that even a stop run to $352 (a -4.9% gap) costs the book 0.12% NAV. The stop catches normal invalidation; the small size catches catastrophic gap invalidation.

Alternative stop considered: Technical-analyst flagged $362 as a wider, "structurally cleaner" stop below the April 2026 breakout gap area. I am rejecting this alternative because: (a) the entire purpose of the $370 limit is to put the stop above the 50% Fib at the structural level, not to widen the stop further; (b) at 67 shares the per-trade dollar risk is trivial — the value of a wider stop is to absorb more noise, which we don't need; (c) a $362 stop with $370 entry would lower R:R to TP1 from 6.1:1 to 4.5:1 without material noise-protection benefit given the position is sized to a hard cap, not to a per-trade risk budget.

Targets

Target 1 (TP1): $398.00

• Rationale: The $398-$400 zone is the most heavily-defined near-term resistance on the chart: it has been tested twice (2026-05-08 swing high $398.37; 2026-05-13 swing high $399.93) and rejected both times. The 52-week high of $404.47 (2026-05-18) sits just above. A move from $370 to $398 is +7.6% and represents the full mean-reversion to the prior swing-high cluster.
• R:R from $370 entry / $365.40 stop: ($398 − $370) / ($370 − $365.40) = $28 / $4.60 = 6.09:1
• Action at TP1: Sell 50% of the position (round to 34 shares of 67) at $398 limit. Move stop on the remaining 33 shares to $370 (entry / break-even).

Target 2 (TP2): $420.00

• Rationale: The current 2σ Bollinger Upper Band is $420.82 (technical.md). The technical-analyst flagged a $420-$425 "BB upper band projection / post-$400 breakout extension." This is a measured-move target: if the $398-$400 wall breaks on volume, the next resistance is the BB upper rail plus prior 2025 highs. It is not the bull-flag pattern target of $451 — that pattern is only ~50% formed (technical-analyst's own assessment) and does not meet the

  60% threshold for a defensible projection.

• R:R from $370 entry / $365.40 stop: ($420 − $370) / ($370 − $365.40) = $50 / $4.60 = 10.87:1
• Action at TP2: Close the remaining 33 shares at $420 limit. Trade is complete.

Trail rule

Why not a target beyond $420: The technical-analyst's $451 bull-flag projection is real but pattern-confidence is only ~50%. Pricing a 12.2:1 R:R target on a half-formed pattern is the "target = prayer" failure mode the role guide warns against. $420 is the BB-projected and pattern-supported level; beyond that, we manage with a trailing stop, not a fixed target.

Horizon

45 trading days (entry fill date through ~2026-07-22).

Catalyst calendar overlay:

Why 45 days and not 21: A 21-day horizon expires 2026-06-18, which lands exactly between the June CPI/FOMC and the DOJ ruling — the worst possible exit window (we'd close before the binary catalyst the thesis is designed to capture). 45 days carries us through the DOJ ruling, which is the load-bearing catalyst per the verdict.

Why not 63 days through Q2 earnings: Q2 print on ~Jul 29 introduces gap risk (consensus 25-30% Cloud growth is the bar; a 25-27% print is neutral-bad per the verdict's invalidation criteria, and an earnings gap can blow through any stop). The DOJ ruling is the catalyst-driven exit opportunity; Q2 earnings would be the gap-risk extension. Closing the trade on the DOJ outcome (or by the 45-day deadline) avoids the earnings-gap exposure entirely and matches the conviction-3 risk posture.

Size

sizing:
  method_used: hard_cap_concentration
  cap_binding: "memory.md 2.5% NAV concentration cap"
  size_pct_nav: 2.48
  shares: 67
  notional_usd: 24790
  entry_price: 370.00
  risk_per_share: 4.60
  risk_dollars: 308.20
  risk_pct_nav: 0.031
  conviction: 3
  rationale: >
    Methods computed: fixed-fractional at conv-3 (0.50% NAV risk
    / $4.60 risk-per-share) would produce 1,086 shares = 40.2%
    NAV — far exceeds both the 10% single-position cap and the
    2.5% concentration cap. Vol-targeted (0.20% daily 1σ NAV /
    HV20 41.1%) suggests 7.7% NAV. Quarter-Kelly (p_win 0.55,
    r_win 6.09) suggests 11.9% NAV. The minimum of these
    methods is 7.7% (vol-targeted), but the memory.md 2.5%
    concentration cap (4th correlated mega-cap tech long;
    MSFT/AAPL/AMZN already open) binds well below that. Final
    size at the cap: 67 shares = $24,790 = 2.48% NAV, risking
    just $308 ($4.60 × 67) = 0.031% NAV at the stop. This is a
    very small per-trade risk by design — the cap reflects
    portfolio concentration, not idea conviction. NAV is
    assumed at $1.0M for journal calibration.
  alternatives:
    fixed_fractional_pct: 40.2   # over caps; informational
    vol_targeted_pct: 7.7
    kelly_quarter_pct: 11.9
    concentration_cap_pct: 2.5   # BINDING
  caps_checked:
    single_position_pct: 10.0    # 2.48 ≤ 10.0 OK
    single_loss_pct: 1.0         # 0.031 ≤ 1.0 OK (massively)
    sector_pct: 35.0             # GOOG + MSFT + AAPL + AMZN in tech; assume each at ~2-3% NAV, basket ~10% NAV ≤ 35% OK
    correlation_pct: 25.0        # Σ|ρ × size| with MSFT/AAPL/AMZN longs estimated ~6-8% NAV ≤ 25% OK
    liquidity: 1_day             # 67 shares at $370 = $24,790; GOOG ADV ~$8-10B; trivial, clears immediately
  nav_assumed_usd: 1000000
  basket_correlation_note: >
    GOOG joins MSFT (OW conv 2), AAPL (OW conv 3), AMZN (OW
    conv 3) as the fourth open US mega-cap tech long. Pairwise
    correlations among these names are typically 0.55-0.75. The
    risk committee should specifically stress: a CPI-driven
    XLK -5% session hits all four positions simultaneously.
    Sized at 2.48% NAV here means the GOOG slice of any
    basket-level shock is bounded.

R:R summary

The asymmetry is favourable: max plausible loss in a structural DOJ adverse ruling (-0.37% NAV) is roughly equal to the gain at TP2 (+0.34% NAV). The expected value calculation rests on the verdict's implicit P(behavioural ruling) > P(structural ruling) read.

Add-on plan

Conditional add: if Q2 2026 Cloud growth prints > 25% YoY and operating margin holds > 30%, then on the next pullback to or below $400, add 33 shares (matching the remaining position size after TP1) for a maximum combined exposure of 100 shares = ~3.7% NAV (still under standard 10% single-position cap; would push concentration cap to 3.7%, requiring a re-validation in memory).

Add criteria — all must be true:

1. Q2 Cloud YoY ≥ 25% (the verdict's invalidation threshold's upper bound; the bar is "upper half" of consensus 25-30%)
2. Q2 operating margin ≥ 30% (the FY2026 EPS-bar floor per verdict assumption #4)
3. DOJ ruling has been behavioural-only (no TAC/default-search structural constraint)
4. Stock has not made a new 52-week high in the interim (the add-on is a confirmation add, not a chase)
5. Re-validate the concentration cap at point of add — if the three other tech longs are still open, the add raises the basket to 4 positions × ~3% = 12% NAV in correlated names, which is at the upper edge of what the system should carry. If MSFT or AAPL has matured by then, reallocate that slot to GOOG.

This is the MSFT-playbook expression: enter at conv-2-equivalent size, add on confirmation rather than front-running the ramp (per memory.md).

Critical assumptions

These inherit from the research verdict's invalidation criteria and are the inputs the risk committee should stress:

1. The pullback to $370 actually materialises. Current spot is $384.90; the trade requires a -3.9% move from spot to fill. If GOOG breaks above $400 on volume before the $370 limit fills, the entry is missed and the breakout alternative (per technical.md: long above $401 with stop $393) is not automatically authorised — that requires a re-run of the verdict.

2. Google Cloud Q2 2026 growth holds at ≥ 27% YoY. Below 25% triggers cascading revisions per estimate-revisions.md and the verdict flips to NO-TRADE. The Q2 print sits just after the 45-day horizon by design.

3. DOJ remedy is behavioural-only. A structural ruling (Chrome divestiture or TAC ban) is an event-driven thesis-invalidator, not a price-driven one — see next section.

4. June 11 CPI core ≤ 0.4% MoM. Hot print triggers VXX +15-25% and TLT -2 to -3% per macro-factor.md, for -8 to -14% factor drag on GOOG. The stop catches this if it reaches $365.40.

5. Operating margin holds > 30% in Q2 print. Already a verdict-level assumption; only relevant for the add-on decision since Q2 sits outside the primary 45-day horizon.

6. Daily close above $375.79 (SMA20) holds on volume after the entry fills. If the pullback to $370 occurs and then GOOG fails to rebound above SMA20 within 10 sessions of fill, the pullback-buy thesis is mechanically failing and the trade should be reviewed regardless of stop status.

7. The three correlated open longs (MSFT, AAPL, AMZN) do not simultaneously stop out. If any two of them hit stops before GOOG fills, the basket has been signalling regime breakdown and the GOOG entry should be cancelled pre-fill.

Thesis-invalidating events

These are events that, if they occur, mean we exit immediately regardless of price level. The stop catches price invalidation; these catch event invalidation.

1. DOJ ruling includes structural remedy. Specifically: any of (a) Chrome divestiture order, (b) Android divestiture order, (c) prohibition or material restriction on the Apple TAC default-search arrangement, (d) forced default-search choice screens beyond cosmetic UI. Read the ruling text directly — do not trade on headline summaries. Exit market-on-close the session of the ruling regardless of price.

2. Hyperscaler capex guidance cut by AMZN (AWS) or MSFT (Azure) in their next earnings. The verdict's bull thesis rests on contracted AI infrastructure demand. If AWS or Azure guides capex down meaningfully (>10% below prior guidance) and cites demand reasons rather than supply-side constraints, the GOOG Cloud demand narrative is undercut and the trade should be closed.

3. Anthropic / Google Cloud commitment publicly revised down or disputed. The $200B / 5-year figure is a load-bearing bull-evidence item. A confirmed downward revision, public contract dispute, or Anthropic re-platforming announcement triggers exit.

4. Berkshire Hathaway 13F filing shows position reduction. Berkshire tripled the stake under Abel (news.md 2026-05-15). A reduction would reverse the institutional-legitimacy signal that was a real component of the May rally. Exit on confirmation, not on rumour.

5. VIX closes above 35 for two consecutive sessions before the entry fills. Per macro-factor.md, this signals a regime flip from RISK_ON to RISK_OFF. The pullback-buy thesis is regime-conditional; in a high-vol regime, the correct entry is at much lower levels (deeper pullback) or not at all. Cancel the limit and re-run.

6. Operating margin Q2 print < 30% (if the position is still open through Q2 earnings, which the default horizon prevents but the add-on plan extends). Triggers FY2026/27 estimate cuts per estimate-revisions.md; exit position.

What I'm explicitly not doing

• Not chasing $384.90 spot. The verdict explicitly prohibits this and the chart explicitly does not support it.
• Not sizing standard conv-3 (3-4% NAV). The memory concentration cap is the binding constraint; this is the fourth correlated mega-cap tech long, not a standalone idea.
• Not using a Jul-17 bull call spread or risk reversal. IV is cheap per the options analyst, but the verdict's $370 limit-entry discipline is incompatible with a premium-paying structure that pays off regardless of entry price. Documented above; switchable if the risk committee requests it.
• Not setting TP3 at $451 (bull flag projection). The pattern is only ~50% formed per technical.md's own assessment. A trailing stop after TP2 ($420) lets us participate in continuation without committing to it as a target.
• Not carrying through Q2 earnings (~July 29). Horizon ends 2026-07-22, deliberately before the print. The Q2 bar is the add-on trigger, not a hold-through-print bet.
• Not adding a hedge against the basket-correlation risk pre-emptively (e.g. SPY puts or XLK puts). The position is sized at 2.48% NAV precisely to make basket-level hedging unnecessary at the per-position level. If the portfolio-manager wants a basket-level hedge, that is a separate book decision.

Handoff note to risk committee

Three things to specifically stress:

1. Basket correlation. GOOG + MSFT + AAPL + AMZN is a single AI-monetisation / XLK-rotation factor bet. Compute book-level VaR with realistic pairwise correlations (0.55-0.75 historical), not the position-level VaR. A CPI-driven -5% XLK session hits all four simultaneously.

2. Catalyst stacking inside horizon. June CPI (Jun 11), FOMC (Jun 17/18), and DOJ ruling (~Jul 1) all sit within the 45-day window. The probability of at least one regime-flipping event is materially higher than for any single event. Model the joint.

3. Limit-not-filled scenario. If GOOG never reaches $370 in the 30-day GTC window, the trade is unfilled and the journal logs a missed-entry, not a trade. The opportunity cost of not chasing $384.90 is a real consideration but is the explicit point of the verdict's entry-discipline stance. Do not retroactively widen the limit unless the verdict is re-run.