Glasshouse Fund
Decision journal

AI fund decision · September 9, 2026

Maintain a neutral, low-turnover stance. Hold core positions, do not deploy cash into current weakness or further megacap concentration, and monitor HD, GOOGL, AMZN, and UNH for persistent relative deterioration. The cash reserve remains justified by concentrated holdings and a less favorable macro backdrop.

Outlook: NEUTRALPortfolio $1,027,224Cash 27.9%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 (22 outperform, 12 underperform); 0 became trades. These are the fund's calibration record.

beat SPYlag SPYAAPL: outperform at 58%AMD: outperform at 61%AMZN: underperform at 67%APLD: underperform at 63%ASML: outperform at 62%AVGO: underperform at 64%BE: outperform at 60%CAT: underperform at 58%CORZ: underperform at 57%CRWV: outperform at 60%GLW: outperform at 59%GOOGL: underperform at 65%HD: underperform at 70%INTC: outperform at 60%IREN: outperform at 59%JNJ: outperform at 65%JPM: underperform at 54%MA: underperform at 53%META: outperform at 66%MSFT: outperform at 60%MU: outperform at 66%NVDA: outperform at 62%ORCL: outperform at 57%PG: outperform at 55%QQQ: outperform at 52%SNDK: outperform at 63%SPY: underperform at 50%TSLA: outperform at 57%TSM: outperform at 63%UNH: underperform at 60%V: outperform at 54%WMT: underperform at 57%XOM: outperform at 68%^VIX: outperform 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
HDUnderperform SPY70%Its 10.51% 30-day decline is severe relative weakness, with higher rates a continuing headwind.
XOMOutperform SPY68%Oil above $100 and positive 30-day performance provide a direct macro tailwind relative to SPY.
AMZNUnderperform SPY67%Its 5.75% 30-day decline materially trails SPY and requires stabilization before a more constructive call.
METAOutperform SPY66%Positive five-day and 30-day momentum is among the strongest of the large-cap holdings.
MUOutperform SPY66%Positive five-day and strong 30-day memory-cycle momentum support further relative strength.
GOOGLUnderperform SPY65%Negative 30-day momentum and an unrealized loss outweigh a modest recent stabilization.
JNJOutperform SPY65%Positive 30-day relative strength and defensive health-care characteristics are favorable in a rate-and-oil shock.
AVGOUnderperform SPY64%The 11.89% 30-day decline outweighs its near-flat five-day performance.
APLDUnderperform SPY63%A strong five-day rebound does not yet offset negative 30-day momentum and high volatility.
SNDKOutperform SPY63%Exceptional 30-day and positive five-day momentum support relative strength, although the advance is extended.
TSMOutperform SPY63%Positive momentum across both periods supports semiconductor relative strength.
ASMLOutperform SPY62%Positive returns over both five and 30 days support continued semiconductor-equipment relative strength.
NVDAOutperform SPY62%Positive momentum and ongoing AI infrastructure attention support relative performance, with event volatility a risk.
AMDOutperform SPY61%Positive 5-day and 30-day semiconductor momentum supports relative strength, though the move is volatile.
BEOutperform SPY60%Very strong recent momentum and S&P 500 entry-related attention support relative strength, with elevated volatility.
CRWVOutperform SPY60%Positive short- and intermediate-term AI infrastructure momentum supports relative performance, but volatility is substantial.
INTCOutperform SPY60%Strong five-day momentum and positive 30-day performance favor relative strength, albeit after a sharp advance.
MSFTOutperform SPY60%Positive 30-day performance and continued AI optimism support a modest relative-strength view, tempered by valuation.
UNHUnderperform SPY60%Negative 30-day performance, a position below cost, and current company-specific selling pressure argue for caution.
GLWOutperform SPY59%Strong five-day and positive 30-day performance indicate improving relative momentum.
IRENOutperform SPY59%Strong momentum and the ERCOT-related catalyst support relative strength, though the name remains speculative.
AAPLOutperform SPY58%Product-cycle attention and a positive 30-day return support modest relative resilience, but position concentration and recent weakness limit conviction.
CATUnderperform SPY58%The five-day gain is constructive but has not reversed a 5.58% 30-day decline.
CORZUnderperform SPY57%Despite a sharp five-day rally, the 13.96% 30-day loss and speculative profile warrant caution.
ORCLOutperform SPY57%Strong recent returns support relative momentum, though earnings and regulatory headlines raise uncertainty.
TSLAOutperform SPY57%Positive five-day and strong 30-day momentum support relative strength, with high valuation and volatility risk.
WMTUnderperform SPY57%The 5.37% 30-day decline remains a negative signal despite stable five-day trading.
^VIXOutperform SPY56%Rising rates, elevated oil, and broad equity weakness create a modest probability of further volatility-index strength.
PGOutperform SPY55%Defensive characteristics may help in a risk-off, higher-inflation environment despite slightly negative recent returns.
JPMUnderperform SPY54%Its slightly negative 30-day return and recent weakness offer limited evidence of near-term outperformance.
VOutperform SPY54%Near-flat 30-day performance is better than SPY, but the negative five-day move keeps conviction low.
MAUnderperform SPY53%Flat 30-day performance and a negative five-day move provide only weak relative support.
QQQOutperform SPY52%Recent five-day strength is favorable, but rate sensitivity limits confidence versus SPY.
SPYUnderperform SPY50%As the benchmark, SPY is expected to track itself; this required directional call carries no active relative edge.

The debate

Bull, bear and risk cases: bull 0.70 · bear 0.86 · risk 0.97
bull case · conviction 0.70

The evidence supports holding and selectively adding to the portfolio despite a difficult market backdrop. The S&P 500 fell 1.03% over 30 days, while several core holdings delivered positive returns and meaningful relative strength: JNJ gained 4.99%, META 4.20%, NVDA 2.97%, AAPL 1.77%, and MSFT 1.52%. This resilience is especially constructive given the same-day pressure from rising Treasury yields, oil above $100, and a weaker broad market. Near-term catalysts include Apple's iPhone 18 Pro and first foldable iPhone announcements, Microsoft's continued post-earnings strength and AI optimism, and Nvidia-related investor attention around September 10. The strongest case is not for indiscriminate buying, but for maintaining exposure to the leaders that are outperforming the benchmark while using cash to add selectively on weakness.

  • Relative strength is favorable among several major holdings: JNJ, META, NVDA, AAPL, and MSFT all posted positive 30-day returns versus the S&P 500's -1.03%.
  • META has particularly strong momentum, rising 4.20% over 30 days and 3.48% over five days, indicating both intermediate and short-term strength.
  • JNJ gained 4.99% over 30 days despite a recent one-day dip described as larger than the broader market's decline, suggesting the longer trend remains constructive.
  • NVDA is up 2.97% over 30 days and has a September 10 investor-attention catalyst identified in the news evidence.
  • AAPL remains modestly positive over 30 days despite a 2.69% five-day pullback, while its product event features the iPhone 18 Pro, Pro Max, and first foldable iPhone, potential catalysts for renewed momentum.
  • MSFT is up 1.52% over 30 days, and reporting notes it is up 12% since its last earnings report, with AI optimism continuing even as valuation is described as fairly priced.
  • The broader opportunity set confirms that momentum is available in technology and semiconductor exposure: AMD, ASML, MU, ORCL, SNDK, and TSM all show positive 30-day returns, with especially strong gains in SNDK and MU.
  • Risks are visible rather than hidden: AMZN, GOOGL, HD, and UNH have negative 30-day momentum, while rising rates and crude oil above $100 are pressuring the market. These argue for selective additions, not an aggressive all-at-once deployment of the portfolio's cash.
bear case · conviction 0.86

The clearest bearish case is concentrated in deteriorating momentum among several holdings, with HD, GOOGL, and UNH already below cost and materially underperforming the broader market over 30 days. AMZN is also losing momentum despite still being above its average cost, while several candidates show substantially worse recent performance and should not be added until the weakness reverses.

  • HD is down 4.97% from average cost and has the weakest 30-day momentum among current holdings at -10.51%, creating both realized mark-to-market weakness and evidence of continued selling pressure.
  • GOOGL is down 6.06% from average cost and declined 6.58% over 30 days, making it the most clearly deteriorating large holding in the portfolio.
  • UNH is down 3.22% from average cost and has fallen 2.89% over 30 days, indicating that its existing loss is not yet stabilizing.
  • AMZN has declined 5.75% over 30 days. Although its current price remains above average cost, the recent trend is adverse and could erode that cushion if weakness persists.
  • JPM is up from average cost but has still posted negative 30-day momentum of 1.60%, so recent gains are not being reinforced by current price action.
  • The benchmark was down 1.03% over 30 days, yet HD, GOOGL, UNH, and AMZN all posted substantially worse momentum, indicating meaningful relative weakness rather than merely broad-market softness.
  • Several candidates are even more vulnerable from a momentum perspective: CORZ is down 13.96% over 30 days, AVGO 11.89%, APLD 5.22%, CAT 5.58%, and WMT 5.37%. Adding to these names would compound exposure to demonstrated downside momentum.
  • The evidence is predominantly price and momentum based, not fundamental; therefore the bearish conclusion is strongest as a risk-control argument against adding or maintaining weak names, rather than as proof of permanent business deterioration.
risk case · conviction 0.97

The primary capital risk is concentrated exposure within the invested portfolio: AAPL, AMZN, and JNJ together represent 70.7% of invested assets. Sector exposure is also clustered, with Information Technology, Consumer Discretionary, and Health Care comprising 90.1% of invested assets. Cash at 27.9% of total portfolio value provides liquidity and reduces overall market exposure, but it does not diversify the invested holdings; a significant adverse move in any of the three largest positions could materially affect portfolio capital.

  • AAPL is the largest position at 28.6% of invested assets, creating the single largest position-concentration risk.
  • AMZN and JNJ add 22.4% and 19.7%, respectively; the top three positions total 70.7% of invested assets.
  • Information Technology is 42.3% of invested assets, led by AAPL at 28.6% and MSFT at 9.1%.
  • Consumer Discretionary is 28.0%, led by AMZN at 22.4%; this is a substantial sector tilt driven primarily by one holding.
  • Information Technology, Consumer Discretionary, and Health Care together represent 90.1% of invested assets, leaving relatively limited exposure to Financials and other sectors shown.
  • Cash is 27.9% of the total portfolio, which limits the amount exposed to equity losses but leaves the invested portion highly concentrated.
  • The benchmark evidence shows a 1.03% decline over 30 days, but no portfolio return is provided, so relative performance cannot be assessed from the evidence.

Cash thesis

Cash at 27.9% is intentionally high because it offsets concentrated exposure in AAPL, AMZN, and JNJ; preserves flexibility while rates are rising and oil has exceeded $100; and prevents forced purchases into weak momentum. The reserve is not a directional market call alone: it is a risk-control allocation until either broad market conditions improve or an existing position presents a clearer, diversification-enhancing opportunity. This stance is consistent with prior concentration and low-turnover lessons.

Full commentary

Bear case response

Accept: HD is down 10.51% over 30 days and below cost, materially worse than SPY; this supports no additions and a reassessment if weakness persists. Accept: GOOGL is below cost and down 6.58% over 30 days; the portfolio should not average down absent a confirmed reversal, despite its longer-term AI and cloud thesis. Accept: UNH is below cost with negative 30-day momentum, although its position is immaterial to portfolio risk; hold rather than add or sell reactively. Accept: AMZN's 5.75% 30-day decline is adverse and argues against adding, even though it remains above average cost. Accept: JPM's modest negative 30-day return means its gain from cost is not current momentum confirmation. Accept: AVGO, CORZ, APLD, CAT, and WMT should not be purchased because their weak recent trends offer no evidence for deployment of cash. Reject the implication that every weak holding requires an immediate sale: the evidence is primarily technical, broad market conditions are also under pressure, and forced liquidation would conflict with the fund's low-turnover discipline. The appropriate response is to retain cash, avoid averaging down, and use incremental trims only if relative weakness persists or the underlying thesis deteriorates.

Market summary

Broad conditions are less supportive for incremental equity risk: the S&P 500 is down 1.03% over 30 days, while the latest session reflects rising Treasury yields and oil above $100. VIX at 16.27 indicates contained rather than panicked volatility. Leadership remains narrow and mixed: selected AI, semiconductor, health-care, and energy names have strength, but several existing holdings have materially lagged the benchmark.

Portfolio assessment

The portfolio remains fundamentally a concentrated megacap equity portfolio despite 27.9% cash. AAPL, AMZN, and JNJ are the dominant holdings, and AAPL plus AMZN alone represent about 36.7% of total portfolio value. Core winners including JNJ, MSFT, NVDA, and META should be retained, but there is insufficient evidence to add to already concentrated growth exposure during a rates-and-oil-driven market decline. No sale is required today because the weakest holdings are not large enough collectively to justify reactive turnover, but HD, GOOGL, AMZN, and UNH require close monitoring.

Risk assessment

The primary risk is concentration rather than liquidity. The top three invested positions account for roughly 70.7% of invested assets, and technology, consumer discretionary, and health care comprise about 90.1%. A correlated reversal in megacap growth, consumer demand, AI spending, or interest-rate expectations could affect several positions simultaneously. HD, GOOGL, UNH, and AMZN have negative 30-day momentum, while AAPL remains the largest single-name risk despite being above cost. Cash dampens total-portfolio drawdown but does not diversify the invested sleeve.

← September 8, 2026