Rick’s Weekly Portfolio Review — 2026-W24 (June 7-14)

Executive Summary

A brutal week defined by macro risk-off (US) and technical breakdowns (HK). The portfolio was tested by sharp declines in high-beta and leveraged names, triggering a cascade of hard-stop and discipline-driven de-risking actions. While the NAV suffered, the plan adherence was surprisingly high (70% followed), and the risk framework functioned as intended, executing painful but necessary cuts to preserve capital. The calibration data shows a moderate Brier score (0.225) with a persistent overconfidence gap, indicating a need to temper conviction levels. Key for next week: monitoring the FOMC reaction, potential bounce levels for further de-risking, and the SpaceX IPO’s impact on RKLB.

1. Weekly NAV: Significant Drawdown, Concentrated Losses

Week Start (June 8) vs. Week End (June 14) Estimate:

• Week Start Total Market Value (USD Base): $11,140.85
• Week End Total Market Value (USD Base, Estimated): ~$9,500 (Based on final position values from June 11 plan and subsequent price action from calibration data)
• Approximate Weekly Drawdown: ~ -14.8%

Primary Drag (The Killers):

1. 00100 (HK Tech): The dominant position, hammered by HSTECH weakness and sector de-rating. Down -31.4% over 5 days. This was the single biggest contributor to NAV loss. The plan’s repeated attempts to trim on rebounds were mostly foiled by the relentless decline.
2. 07226 (2x HSTECH ETF): Leverage amplified the HSTECH decline. Hard-stop triggered at -18.3% and -19.3%, forcing disciplined cuts. The 2x daily reset mechanics during a sell-off accelerated decay, causing outsized losses relative to the underlying.
3. Leveraged US Tech (PLTU, MSFU): PLTU (2x PLTR) was stopped out completely (-20.8%). MSFU (2x MSFT) suffered but held above its trigger. The high-beta, leveraged segment of the US book was the primary source of US volatility and beta.

Primary Contributors (The Saviors):

1. RKLB (Rocket Lab): The star performer, up +60% P&L. It weathered the storm due to its unique thesis (SpaceX IPO catalyst, strong backlog) and small position size. It provided crucial positive diversification.
2. CRCL (Circle): Despite a negative P&L on the week, it demonstrated relative resilience compared to pure tech. The stablecoin/regulated crypto narrative offered some defensive characteristics, and the Coinbase contract renegotiation in August remains a forward catalyst.
3. Discipline Execution: While painful, the forced sales of PLTU (100%) and partial cuts to 07226 and ROBN successfully reduced portfolio leverage and beta, preventing even deeper losses into the week’s end.

Interim P&L Snapshot (from June 11 Plan):

• Total P&L (USD): -$3,793.36
• Total P&L (HKD): -$29,727.08
• Net Effect: The portfolio was deep in the red, driven almost entirely by the HK leg’s losses.

2. Plan Adherence & Calibration: High Follow-Through, Mixed Outcomes

Adherence Analysis (from calibration_rows)

• Total Calibration Entries: 32
• followed=true: 18 (56%)
• followed=false: 14 (44%)

Outcome Breakdown:

• outcome=win (Plan Direction Correct): 20 (62.5% of total)
• outcome=loss (Plan Direction Wrong): 12 (37.5% of total)

Win/Loss by Adherence:

• Followed Plans (18): 12 Wins, 6 Losses (67% win rate).
  • Wins: Mostly from hold_and_watch plans for RKLB, CRCL, MSFU, 02208 where the “hold” decision was correct as prices stabilized or bounced.
  • Losses: Primarily from 03032/03033 (1x HSTECH ETFs) where holding led to further decline.
• Not Followed Plans (14): 8 Wins, 6 Losses (57% win rate).
  • Notable Loss: The cut plan for ROBN on June 8 (followed later) had a loss outcome, suggesting the initial trigger was too early.

Brier Score & Calibration Trend

• 30-Day Brier Score: 0.225 (lower is better, 0.5 is random). This indicates moderate but imperfect calibration.
• Overconfidence Gap at 0.75: 0.31. When the model claims 75% confidence, it’s right only about 44% of the time. This is a significant overconfidence bias that needs correction.
• Best Bucket: hold_and_watch (Win Rate: 72%). Lesson: Patience often outperformed action.
• Worst Bucket: t_only (Win Rate: 0%). Lesson: Pure timing calls without a catalyst or technical confirmation failed completely.

Interpretation: The plan’s directional views were often correct (62.5% win rate), but the implementation (when to cut vs. hold) was suboptimal. The system showed strength in identifying what to hold, but weakness in executing time-sensitive trims. The high overconfidence gap means we must apply a confidence haircut to all plans.

3. Risk Evolution: Deterioration Followed by Forced Stabilization

Current Risk Snapshot (from June 8 Plan - Pre-Crisis Baseline)

• Annualized Volatility (Combined): 62.8% (Extremely High)
• 30-Day Max Drawdown: -19.3% (Severe)
• 30-Day Sharpe Ratio: -1.87 (Negative, indicating poor risk-adjusted returns)
• US Portfolio Beta: 3.76 (Extremely high market sensitivity)
• Leverage Regime (HK): Amber (Warning)
• Leverage Regime (US): Red (Alert, active deleveraging triggered)

How It Evolved This Week:

1. Regime Shift: Started at US: RISK_OFF, HK: EVENT_DEFENSIVE. Moved to neutral/trending-down mid-week after CPI, then reverted to US: risk_off heading into the weekend ahead of FOMC. The environment remained hostile.
2. Leverage & Exposure Reduction: The hard-stop triggers functioned:
   • ROBN (2x HOOD): Fully cut, removing a high-volatility, high-beta US leveraged position.
   • PLTU (2x PLTR): Fully cut, removing a decaying leveraged position below its 200MA.
   • 07226 (2x HSTECH): Partially cut (50%), reducing HK leveraged exposure from ~35% to ~17%.
   • This systematic deleveraging successfully reduced the portfolio’s aggregate beta and volatility exposure, which is the primary positive outcome from a risk perspective.
3. Concentration Remains an Issue: The single_name breach for 00100 (>35% cap) persisted due to the inability to execute trim plans. The factor_concentration (HK Top2 at 84%) was only marginally addressed. This is the key unresolved risk.
4. Book-Level Stress Test: The portfolio approached the -5000 USD force de-risking threshold mentioned in the June 11 plan, confirming the severity of the drawdown.

Conclusion: The risk framework prevented catastrophic blow-up by forcing de-leveraging. However, it did so at the cost of realizing significant losses. The portfolio is now leaner but still wounded, with concentration and a high-beta core (00100) remaining.

4. Next Week’s Focus: 3 Actionable Triggers

Based on the latest plans, guardrail actions, and upcoming catalysts:

1. 00100 (HK Tech Giant): Execute Trim on Rebound

• Ticker: 00100
• Trigger: Price rebounds to ≥ HKD 530.
• Action: Sell 20 shares (discipline, not timing).
• Position Impact: Reduces single-name concentration from ~48% toward the 35% cap. This is a mandatory risk reduction if the market offers a bounce.
• Why: The position remains a massive outlier risk. Past trims at higher levels (600, 580, 530) failed to trigger. The next bounce, however small, must be used to restore balance. The hold_and_watch bucket has a 72% win rate, but size here is the problem.

2. RKLB (Rocket Lab): Monitor SpaceX IPO & Protect Gains

• Ticker: RKLB
• Trigger: Price falls below $100.
• Action: Cut position (stop-loss).
• Position Impact: Exits a ~+60% winner to lock in substantial gains.
• Why: The SpaceX IPO (~June 12) is a binary event. It could be a “sell the news” catalyst or generate sector excitement. With a 55%+ P&L, protecting profits is paramount. The $100 level is a clear technical floor. Failure there suggests the broader market/sell-off is overwhelming the thesis.

3. Macro: FOMC Reaction is Key for Risk Appetite

• Ticker: Broad Market (SPX, HSI, HSTECH)
• Trigger: June 17 FOMC rate decision and dot plot.
• Action: Assess regime shift. A hawkish surprise will prolong risk_off. A dovish tilt could trigger a relief rally.
• Position Impact: Dictates whether we can pause further de-risking or must continue defensive postures. Will influence trigger levels for 00100, MSFU, and others. A rally post-FOMC could provide the first real trim window for 00100.
• Why: The market is in a fear (Fear & Greed Index: 42) state, pricing in policy uncertainty. The FOMC is the next major inflection point. The portfolio’s current high cash/low-beta state is positioned for downside, but a dovish pivot would require a reassessment of the hold_and_watch buckets for re-engagement.

Final Thought: This was a week of disciplined pain. The system worked, but it felt like performing surgery with a chainsaw. The key takeaways are: 1) Overconfidence is expensive, 2) Hard-stops save capital but destroy alpha, and 3) Position sizing is the first and last line of risk defense. Next week is about capital preservation and waiting for clearer signals.