Glasshouse Fund
Decision journal

AI fund decision · August 20, 2026

Maintain a neutral, low-turnover posture. Hold existing leaders, avoid adding to laggards or weak watchlist names, and retain the 25.86% cash reserve pending either clearer breadth improvement or a more attractive entry opportunity. Sources used in forming the decision include risk_lesson:run_20260803T191113Z_b5f11a59:report_risk, risk_lesson:run_20260729T184341Z_54100baf:report_risk, trade:run_20260812T183945Z_0d1bf062:AAPL:SELL, trade:run_20260807T183247Z_07bc8571:AAPL:BUY, and report_summary:run_20260810T195022Z_25113609.

Outlook: NEUTRALPortfolio $1,034,987Cash 25.9%Grounding: ok

Trades

No trades this day. The fund held.

Market calls

A directional call (beat or lag the S&P 500 over the horizon) on every researched name, whether or not the fund traded it. 34 calls (12 outperform, 22 underperform); 0 became trades. These are the fund's calibration record.

beat SPYlag SPYAAPL: underperform at 60%AMD: underperform at 68%AMZN: outperform at 59%APLD: underperform at 62%ASML: underperform at 61%AVGO: underperform at 72%BE: underperform at 67%CAT: underperform at 66%CORZ: underperform at 70%CRWV: underperform at 63%GLW: underperform at 65%GOOGL: underperform at 59%HD: underperform at 56%INTC: underperform at 71%IREN: outperform at 55%JNJ: outperform at 66%JPM: outperform at 55%MA: outperform at 54%META: underperform at 76%MSFT: outperform at 62%MU: outperform at 58%NVDA: outperform at 53%ORCL: outperform at 57%PG: underperform at 58%QQQ: underperform at 54%SNDK: outperform at 61%SPY: outperform at 50%TSLA: underperform at 66%TSM: underperform at 57%UNH: underperform at 69%V: underperform at 53%WMT: underperform at 57%XOM: outperform at 63%^VIX: underperform at 56%50%60%70%80%90%100%
Each tick is one call, placed by stated confidence. Hover for the name.
All 34 calls, with reasoning
SymbolCallConf.Why
METAUnderperform SPY76%The -17.83% 30-day decline and -8.23% five-day decline indicate severe momentum deterioration versus SPY.
AVGOUnderperform SPY72%The -13.24% five-day decline and negative 30-day return indicate sharp near-term relative weakness.
INTCUnderperform SPY71%The -11.25% five-day decline and negative 30-day return show pronounced relative weakness.
CORZUnderperform SPY70%The -10.94% 30-day decline and reported capital-spending scrutiny outweigh the recent data-center expansion narrative.
UNHUnderperform SPY69%Below-cost status, an -8.21% 30-day return and governance-related uncertainty indicate unfavorable relative momentum.
AMDUnderperform SPY68%Negative five-day and 30-day momentum indicate weaker semiconductor relative strength versus SPY.
BEUnderperform SPY67%The stock's -12.53% five-day move and flat 30-day result trail the benchmark's positive monthly return.
CATUnderperform SPY66%Negative returns across five and 30 days indicate cyclical momentum is weaker than SPY.
JNJOutperform SPY66%Strong five-day and 30-day returns, plus defensive characteristics, support continued relative strength versus SPY.
TSLAUnderperform SPY66%The five-day rebound does not offset the -10.21% 30-day decline and continued high volatility.
GLWUnderperform SPY65%Negative five-day and 30-day momentum, alongside documented operational and customer-financing risks, argue for caution.
CRWVUnderperform SPY63%Strong 30-day appreciation is offset by a -14.51% five-day reversal, making the risk-adjusted setup less favorable than SPY.
XOMOutperform SPY63%Strong five-day and 30-day returns, supported by oil-price and Hormuz-risk headlines, favor relative strength.
APLDUnderperform SPY62%Despite positive 30-day performance, the -7.81% five-day decline signals elevated volatility and weakening near-term relative strength.
MSFTOutperform SPY62%Azure and AI growth narratives accompany a 20.75% 30-day gain, although recent weakness and extension reduce conviction.
ASMLUnderperform SPY61%Negative five-day and 30-day returns provide insufficient evidence of relative leadership.
SNDKOutperform SPY61%Positive returns over both five and 30 days demonstrate clear relative momentum versus SPY.
AAPLUnderperform SPY60%The 3.79% five-day rebound is constructive, but the -4.85% 30-day return materially trails SPY and concentration limits appetite for adding.
AMZNOutperform SPY59%Positive 30-day relative performance, positive five-day momentum and AWS-focused headlines support a modest edge over SPY.
GOOGLUnderperform SPY59%Waymo and AI commercialization are long-term supports, but current negative 30-day momentum and below-cost position status lag SPY.
MUOutperform SPY58%A 9.83% 30-day return materially exceeds SPY, with only modest recent weakness.
PGUnderperform SPY58%Flat five-day action and a -3.99% 30-day return lack the momentum needed to match SPY.
ORCLOutperform SPY57%The 15.78% 30-day gain and AI-backlog narrative support relative strength despite a sharp recent pullback.
TSMUnderperform SPY57%Near-flat 30-day performance and negative five-day momentum trail SPY's stronger monthly result.
WMTUnderperform SPY57%Negative five-day and 30-day returns, together with the reported sharp market reaction, indicate weaker momentum.
HDUnderperform SPY56%The stock is stable and slightly above cost, but frozen housing-market conditions and negative 30-day momentum imply limited relative upside.
^VIXUnderperform SPY56%The VIX is down 11.00% over 30 days and remains low, though current macro headlines prevent high conviction.
IRENOutperform SPY55%Strong 30-day performance and AI-infrastructure interest support upside, though the recent five-day decline makes conviction low.
JPMOutperform SPY55%Its 30-day gain modestly exceeds SPY and the position is well above cost, though recent five-day momentum is negative.
MAOutperform SPY54%Positive five-day and 30-day returns with strong gains versus cost support a narrow relative-performance advantage.
QQQUnderperform SPY54%The index has only modest 30-day gains and negative five-day momentum amid pressure on major technology constituents.
NVDAOutperform SPY53%AI demand and a 4.90% 30-day gain support a slight edge, but earnings risk and negative five-day momentum make this low conviction.
VUnderperform SPY53%Returns are positive but near flat over both five and 30 days, offering limited evidence of an edge over SPY.
SPYOutperform SPY50%SPY is the benchmark reference; this forced directional call reflects no claimed relative edge.

The debate

Bull, bear and risk cases: bull 0.68 · bear 0.92 · risk 0.98
bull case · conviction 0.68

The strongest bullish case is to hold the portfolio’s proven leaders and selectively add to the names showing clear relative strength rather than broadly de-risking. The S&P 500 rose 4.05% over 30 days, while MSFT gained 20.75%, JNJ 9.38%, AMZN 6.38%, NVDA 4.90%, and JPM 4.11%. Several of these holdings are also materially above cost, providing evidence that the portfolio has exposure to durable winners. Near-term catalysts include Microsoft’s Azure/AI growth narrative, Nvidia’s earnings next week amid reported 'rock-solid' AI demand, Apple’s potential 2027 foldable iPhone and camera-equipped AirPods launches, and Alphabet’s expanding Waymo commercialization. The case is not uniform. META, UNH, GOOGL, and AAPL have negative 30-day momentum, but the leadership group’s relative strength and identifiable technology and healthcare catalysts support holding and selectively adding rather than selling indiscriminately.

  • MSFT is the clearest momentum leader: up 20.75% over 30 days versus 4.05% for the S&P 500, with news focused on Azure and AI growth and evidence of increased institutional positioning.
  • JNJ has strong defensive and relative-strength characteristics, rising 9.38% over 30 days and 4.32% over five days; its current price is also well above the reported average cost.
  • AMZN is outperforming the benchmark over 30 days by more than two percentage points, with positive five-day momentum and a current price above average cost.
  • NVDA is up 4.90% over 30 days, slightly ahead of the S&P 500, and has a near-term earnings catalyst tied to reported strong AI demand despite recent five-day weakness.
  • JPM is also outperforming the benchmark over 30 days, while MA is nearly matching it; both positions are above average cost, supporting continued exposure to financial and payments momentum.
  • AAPL’s five-day rebound of 3.79% and reported potential 2027 product launches provide a recovery catalyst, although its negative 30-day return argues for holding or staged buying rather than an aggressive increase.
  • Market risk is real: recent reporting cites pressure from Treasury yields and concern about an equity-market bubble. However, the VIX is down 11.0% over 30 days, indicating that broader volatility conditions have improved despite the latest session’s weakness.
bear case · conviction 0.92

The portfolio and candidate set show broadening relative weakness, with several current holdings already below cost and many names losing momentum while the S&P 500 gained 4.05% over the last 30 days. The clearest downside risk is concentrated in META, UNH, GOOGL, and AAPL, while the candidate list is dominated by negative 30-day performance rather than improving leadership.

  • META is already 4.89% below cost and has the weakest momentum among the named current holdings at -17.83% over 30 days, indicating severe deterioration rather than a temporary lag.
  • UNH is 4.98% below cost and down 8.21% over 30 days, combining realized portfolio losses with accelerating weakness.
  • GOOGL is 3.49% below cost and down 2.75% over 30 days, showing both negative position performance and underperformance versus the S&P 500's +4.05% return.
  • AAPL remains above its average cost, but its 30-day return of -4.85% is materially negative and sharply trails the benchmark, making recent strength vulnerable to reversal.
  • HD is only modestly down over 30 days at -1.07%, but that still represents underperformance against the benchmark and provides no evidence of positive momentum.
  • The candidate universe is broadly weak: CORZ fell 10.94%, TSLA 10.21%, AMD 6.89%, CAT 6.78%, AVGO 3.20%, GLW 3.74%, and INTC 4.31% over 30 days. Adding candidates with this profile risks buying into deteriorating trends.
  • ASML, BE, and WMT are also not demonstrating meaningful upside momentum, with 30-day returns of -1.73%, -0.05%, and -0.57%, respectively.
  • GLW combines negative 30-day momentum with documented risk exposure to customer losses, customer financing and receivables, product or component performance, litigation, regulatory developments, technology change, and the need to pace capital spending to demand. These risks could amplify downside if momentum continues to deteriorate.
risk case · conviction 0.98

The primary capital risk is concentration: despite a substantial 25.9% cash reserve, the invested portfolio is dominated by three positions and a small number of sectors. AAPL, AMZN, and JNJ represent 71.6% of invested assets, while Information Technology and Consumer Discretionary together account for 71.1% of reported sector exposure. This creates substantial dependence on a few holdings and sector groups, independent of their expected direction.

  • AAPL is the largest position at 27.7% of invested assets, making it the single biggest source of position-specific capital risk.
  • AMZN is 24.8% and JNJ is 19.1%; combined with AAPL, the top three positions comprise 71.6% of invested assets.
  • Information Technology is 40.5% and Consumer Discretionary is 30.6%, producing a combined 71.1% sector concentration.
  • Health Care adds another 19.3%, so the reported exposure across Information Technology, Consumer Discretionary, and Health Care is 90.4%.
  • The remaining positions are individually small, with each of V, META, MA, JPM, and UNH below 1.2% of invested assets; they provide limited diversification against the dominant holdings.
  • Cash at 25.9% reduces the amount of capital exposed to the concentrated holdings, but it does not change the concentration of the invested portion.

Cash thesis

Cash is 25.86% of portfolio value, above the 25% threshold. This reserve is intentional: it offsets substantial single-name and sector concentration, preserves flexibility for a clearer market pullback or confirmed relative-strength setup, and avoids deploying capital into a watchlist with broadly weak recent momentum. The opportunity cost is acceptable because the portfolio remains materially invested in the strongest current holdings.

Full commentary

Bear case response

Accept the bear case on concentration: AAPL, AMZN and JNJ dominate invested capital, and cash reduces total exposure but does not fix concentration within equities. Accept that META, UNH, GOOGL and AAPL have lagged SPY over 30 days, with META and UNH showing the most concerning deterioration; this supports no new purchases in those names. Reject an implication that all laggards should be sold immediately: GOOGL and AAPL remain core-scale positions where a forced response to short-term momentum would risk repeating recent trading churn, and their position sizes require evidence beyond one weak month before a major exit. Accept that most watchlist candidates have negative or unconfirmed momentum, so no new candidate purchases are justified. Accept that lower VIX is not an all-clear given rate-sensitive headlines and narrow leadership. Hold proven leaders rather than chase them: MSFT, JNJ and AMZN support portfolio retention, but their gains do not justify aggressive additions at current concentration levels.

Market summary

The S&P 500 gained 4.05% over 30 days, but leadership is narrow and the latest session reflects renewed rate and macro sensitivity. MSFT, JNJ, AMZN and selected financial holdings have shown relative strength, while AAPL, GOOGL, META and UNH have lagged. The VIX remains relatively contained at 15.8 despite the weaker session, supporting a measured rather than defensive wholesale stance.

Portfolio assessment

The portfolio is up modestly versus cost in most major holdings, but its construction remains highly concentrated: AAPL, AMZN and JNJ are the dominant exposures, with further correlated AI/cloud exposure through MSFT, NVDA and GOOGL. No trade is warranted today. Recent AAPL trimming followed by re-adding argues against further short-term churn, while the current cash reserve already provides meaningful risk control.

Risk assessment

The central risk is concentration rather than an absence of market exposure. AAPL, AMZN and JNJ account for approximately 71.6% of invested assets, while technology and consumer-discretionary exposure are dominant. A common AI, rates, regulatory, consumer-demand or valuation shock could affect several large positions simultaneously. META, UNH and GOOGL remain below cost and have weak 30-day momentum; NVDA has a near-term earnings event that can create binary volatility. The cash reserve moderates total-portfolio drawdown risk but does not diversify the invested sleeve.

← August 19, 2026 August 21, 2026 →