Glasshouse Fund
Decision journal

AI fund decision · September 2, 2026

Maintain all current positions and retain the 25.7% cash reserve. The portfolio should preserve exposure to MSFT and NVDA leadership while avoiding additions to concentrated AAPL and AMZN positions or weakening HD, UNH and GOOGL holdings. Sources used include current 5-day and 30-day returns, supplied market and company news, risk_lesson:run_20260729T184341Z_54100baf:report_risk, risk_lesson:run_20260803T191113Z_b5f11a59:report_risk, trade:run_20260812T183945Z_0d1bf062:AAPL:SELL, and report_summary:run_20260826T160411Z_f30a1737.

Outlook: NEUTRALPortfolio $1,042,879Cash 25.7%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 (18 outperform, 16 underperform); 0 became trades. These are the fund's calibration record.

beat SPYlag SPYAAPL: underperform at 62%AMD: outperform at 58%AMZN: outperform at 57%APLD: underperform at 59%ASML: outperform at 63%AVGO: underperform at 53%BE: outperform at 56%CAT: underperform at 60%CORZ: underperform at 72%CRWV: outperform at 61%GLW: outperform at 60%GOOGL: underperform at 58%HD: underperform at 64%INTC: outperform at 57%IREN: outperform at 55%JNJ: underperform at 52%JPM: underperform at 51%MA: underperform at 54%META: underperform at 57%MSFT: outperform at 73%MU: outperform at 66%NVDA: outperform at 69%ORCL: outperform at 55%PG: underperform at 56%QQQ: outperform at 59%SNDK: outperform at 62%SPY: outperform at 50%TSLA: outperform at 58%TSM: outperform at 64%UNH: underperform at 65%V: underperform at 53%WMT: underperform at 63%XOM: outperform at 61%^VIX: underperform at 55%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
MSFTOutperform SPY73%Its 28.5% 30-day gain, positive 5-day trend, and AI datacenter expansion narrative remain the clearest portfolio leadership signals.
CORZUnderperform SPY72%Losses of 7.7% over 5 days and 11.3% over 30 days indicate a deteriorating high-volatility trend.
NVDAOutperform SPY69%Positive returns of 3.7% over 5 days and 14.4% over 30 days materially exceed SPY, supporting continued leadership.
MUOutperform SPY66%A 26.3% 30-day advance reflects strong memory-cycle leadership, though labor and semiconductor-cycle risks remain.
UNHUnderperform SPY65%Negative 5-day and 30-day returns, trading below cost, and unresolved margin-recovery concerns indicate relative weakness.
HDUnderperform SPY64%The 5.5% 30-day decline, 4.5% 5-day decline, and cooling-demand concerns outweigh valuation commentary.
TSMOutperform SPY64%A 10.5% 30-day gain supports semiconductor leadership despite a modestly negative 5-day return.
ASMLOutperform SPY63%Its 7.4% 30-day gain and semiconductor-demand exposure support modest relative outperformance despite a weak recent week.
WMTUnderperform SPY63%A 7.0% 30-day decline outweighs a positive recent week and leaves the trend weaker than SPY.
AAPLUnderperform SPY62%A 3.7% 5-day rebound is constructive, but its 3.8% 30-day decline versus SPY strength and concentration risk temper the outlook.
SNDKOutperform SPY62%Exceptional 51.3% 30-day momentum and positive 5-day performance support outperformance, with elevated reversal risk after the run.
CRWVOutperform SPY61%The 34.6% 30-day return supports leadership, but the 7.0% 5-day decline and financing sensitivity make it speculative.
XOMOutperform SPY61%Oil-price strength amid Middle East escalation supports its positive 5-day and 30-day relative momentum.
CATUnderperform SPY60%Negative returns over both 5 days and 30 days point to weaker cyclical relative momentum.
GLWOutperform SPY60%A 17.6% 30-day gain supports relative strength, although a 4.7% weekly decline signals near-term volatility.
APLDUnderperform SPY59%A 9.6% 5-day decline signals elevated volatility despite a positive 30-day return.
QQQOutperform SPY59%A 6.9% 30-day gain exceeds SPY, though recent performance is slightly negative and concentration in growth leaders remains a risk.
AMDOutperform SPY58%Its 7.0% 30-day gain exceeds SPY, although the 4.4% 5-day decline makes this a moderate-conviction call.
GOOGLUnderperform SPY58%Negative 5-day and 30-day performance, combined with trading below cost, indicate weaker relative momentum despite AI-product optionality.
TSLAOutperform SPY58%Positive 5-day and 19.4% 30-day momentum support relative strength, although volatility remains high.
AMZNOutperform SPY57%The 12.5% 30-day advance supports relative strength, but the recent pullback and FTC-related headlines reduce conviction.
INTCOutperform SPY57%Positive 5-day and 8.7% 30-day returns indicate improving relative momentum, albeit with cyclical semiconductor risk.
METAUnderperform SPY57%The 1.2% 30-day decline versus SPY and only marginal 5-day strength point to fading relative leadership.
BEOutperform SPY56%A 30.5% 30-day advance remains powerful, though volatility and a negative recent 5-day return constrain confidence.
PGUnderperform SPY56%Near-flat 30-day performance and limited recent momentum are unlikely to match SPY absent a stronger defensive rotation.
IRENOutperform SPY55%Strong 30-day performance and contracted-ARR commentary are supportive, but the 7.0% 5-day drop reflects high volatility.
ORCLOutperform SPY55%The 20.0% 30-day gain supports relative strength, but a 5.1% pullback and concerns about spending and layoffs reduce confidence.
^VIXUnderperform SPY55%The VIX is down 20.9% over 30 days despite a 7.4% 5-day increase, indicating a contained but rising near-term volatility risk.
MAUnderperform SPY54%The 3.2% 30-day gain trails SPY and a 2.9% 5-day decline suggests softer near-term relative momentum.
AVGOUnderperform SPY53%Positive 5-day momentum is offset by a nearly flat 30-day return, leaving insufficient relative-strength confirmation.
VUnderperform SPY53%The 1.3% 30-day return trails SPY and the 2.9% 5-day decline signals modest near-term weakness.
JNJUnderperform SPY52%Its positive 30-day trend is modest, and its large existing allocation limits incremental appeal versus SPY.
JPMUnderperform SPY51%A positive 30-day trend is offset by flat near-term momentum and heightened macro/rate uncertainty.
SPYOutperform SPY50%As the comparison benchmark, SPY is assigned neutral baseline confidence; its 4.4% 30-day gain reflects moderate market resilience.

The debate

Bull, bear and risk cases: bull 0.78 · bear 0.86 · risk 0.98
bull case · conviction 0.78

Buy/hold with a growth-led bias: the strongest portfolio exposures are demonstrating clear relative strength versus a nearly flat S&P 500, while company-specific catalysts support continued leadership. MSFT, NVDA, and AMZN are the core momentum engines, and JNJ, JPM, MA, and V provide profitable winners with positive 30-day trends. The case is not uniform. HD, UNH, GOOGL, and AAPL have weaker recent relative performance, and geopolitical escalation, oil above $90, inflation concerns, and a rising VIX argue for disciplined sizing, but the evidence favors maintaining exposure rather than retreating to cash.

  • MSFT is the standout holding: its 30-day momentum is +28.5% versus the S&P 500's roughly +0.4% benchmark return and QQQ's +6.9%; its 5-day return is also positive at +0.9%. The reported golden cross and AI datacenter expansion narrative provide technical and fundamental catalysts.
  • NVDA combines strong momentum with renewed near-term leadership: +14.4% over 30 days and +3.7% over five days, materially ahead of the benchmark. The bullish forecast highlighted in the news is an additional catalyst, although it should be treated as scenario-based rather than certain.
  • AMZN has delivered the strongest 30-day momentum among the major diversified holdings at +12.5%, despite a mild -2.1% five-day pullback. That combination suggests the recent weakness may be consolidation within a still-strong intermediate trend, and the position remains above its average cost.
  • The portfolio already owns several proven relative-strength winners: MA is +3.2% over 30 days and trades well above its average cost; JPM is +3.0% over 30 days; JNJ is +2.6% over 30 days and has news citing an 82% three-year run; and V is still positive over 30 days. These winners help balance weaker positions.
  • Current prices are above average cost for MSFT, NVDA, AMZN, JNJ, JPM, V, MA, and META, giving the portfolio a meaningful cushion and supporting a hold posture rather than forced selling after short-term volatility.
  • AAPL's 5-day rebound of +3.7% is constructive, with news pointing to fresh AI Mac launches and a TV price increase as possible catalysts. However, its 30-day return is -3.8% and one valuation analysis calls it 28% overvalued, so AAPL is better viewed as a hold or selective add, not the primary momentum buy.
  • The broader tape still rewards growth: QQQ is up +6.9% over 30 days, while several semiconductor and infrastructure names show substantial strength, including AMD +7.0%, ASML +7.4%, TSM +10.5%, and MU +26.3%. That supports continued exposure to the portfolio's AI and technology leaders.
  • Risk is visible but not yet decisive: the VIX is down 20.9% over 30 days despite rising 7.4% over five days, while the S&P 500 is only modestly higher. With about $268,000 in cash, the portfolio can hold its strongest leaders and add selectively rather than chase every extended mover.
bear case · conviction 0.86

The strongest bearish case is that portfolio breadth and near-term momentum are deteriorating despite the S&P 500 gaining 0.4% over the last 30 days. Three holdings are already below cost. HD, UNH, and GOOGL, and each is also showing negative 30-day momentum. AAPL remains above cost but has declined 3.8% over 30 days, while META is also fading. The candidate set is no safer: CORZ and WMT have particularly severe 30-day declines, and AVGO and CAT are effectively flat-to-negative. This combination raises the risk that current unrealized gains and recent leadership are vulnerable to further reversal.

  • HD is below its average cost by 2.3% and has the weakest negative momentum among the directly held large-cap names at -5.5% over 30 days.
  • UNH is below cost by 1.5% and has declined 5.8% over 30 days, indicating both realized weakness and deteriorating trend.
  • GOOGL is below cost by 4.1% and has negative 30-day momentum of 0.5%, making it the weakest current holding on a cost-basis comparison.
  • AAPL is still above average cost, but its 30-day decline of 3.8% threatens to erode that cushion if the fading trend continues.
  • META is modestly above cost but has declined 1.2% over 30 days, suggesting recent gains are losing momentum.
  • The benchmark rose 0.4% over the same 30-day period, so the negative momentum in AAPL, HD, UNH, GOOGL, and META represents relative deterioration rather than merely a broad-market decline.
  • Candidate risk is substantial: CORZ fell 11.3% and WMT fell 7.0% over 30 days, while CAT and AVGO also posted negative returns. Adding candidates with already-deteriorating momentum could compound downside exposure.
  • The evidence does not show broad-based momentum support across the names supplied; several key holdings and candidates are weakening simultaneously, increasing the risk of further portfolio drawdown if the reversals persist.
risk case · conviction 0.98

The primary capital risk is concentration: 52.3% of invested assets is held in AAPL and AMZN, and the top three positions. AAPL, AMZN, and JNJ, represent 71.5% of invested assets. This is compounded by a 71.0% combined allocation to Information Technology and Consumer Discretionary. Cash at 25.7% provides a substantial liquidity buffer and limits total portfolio exposure, but does not diversify the invested capital.

  • AAPL is the largest position at 28.3% of invested assets; AMZN is second at 24.0%. Together they account for more than half of invested assets.
  • JNJ adds a further 19.2% position, making the top three holdings 71.5% of invested assets.
  • Information Technology is 41.5% and Consumer Discretionary is 29.5%, for a combined sector concentration of 71.0%.
  • Additional exposure to the largest technology-oriented holdings comes through MSFT at 8.8%, GOOGL at 6.7%, and NVDA at 4.3% of invested assets.
  • Cash is 25.7% of the portfolio, which reduces the amount exposed to position and sector losses and provides liquidity; however, the remaining invested capital is still highly concentrated.
  • The evidence does not establish a broad, evenly diversified equity exposure: five positions. AAPL, AMZN, JNJ, MSFT, and GOOGL, comprise 87.0% of invested assets.

Cash thesis

Cash is 25.7% of portfolio value and is intentionally retained. It offsets concentration in AAPL, AMZN, JNJ and technology/consumer-discretionary exposure, provides protection against geopolitical and inflation-driven volatility, and preserves flexibility for a broader market pullback or clearer relative-strength confirmation. Holding cash is preferable to chasing extended leaders or adding to weakening positions.

Full commentary

Bear case response

Accept: HD, UNH and GOOGL are below cost and have negative 30-day momentum; AAPL and META also lag SPY over 30 days. This is genuine relative deterioration and supports no new purchases in those names. Accept: benchmark gains alongside these declines show the weakness is portfolio-specific rather than entirely market-driven. Accept: candidate risk is elevated in CORZ and WMT, with CAT and AVGO lacking compelling momentum, so they are not appropriate diversification purchases now. Partially reject: the weak names do not justify wholesale selling because their positions are modest except AAPL, the portfolio has substantial cash, and there is no evidence of a broad market breakdown. Accept the risk rebuttal that MSFT, NVDA and AMZN are legitimate leaders but not portfolio stabilizers; they are held without adding because leadership is narrow and AMZN has a negative 5-day move plus regulatory headlines. The bear case therefore supports a defensive hold-and-cash posture, not forced liquidation.

Market summary

The market is mixed: SPY is up 4.4% over 30 days and QQQ is up 6.9%, but recent breadth is uneven and several growth and cycliclical names are pulling back. Escalation in the Middle East, oil above $90, and renewed inflation concerns raise rate and margin risks. The VIX remains low relative to 30 days ago despite a 5-day increase, supporting caution rather than a wholesale risk exit.

Portfolio assessment

The portfolio has strong unrealized gains in MSFT, NVDA, JNJ, AAPL, AMZN and payment networks, but invested capital remains highly concentrated in AAPL, AMZN and JNJ. AAPL and AMZN together are over half of invested assets, while AAPL, AMZN and JNJ dominate portfolio outcomes. Recent fund memory favors low turnover, selective risk control, and avoiding renewed additions to already concentrated core positions. No trade is warranted today because the strongest holdings should not be sold solely on short-term volatility, while weak holdings lack sufficient confirmation for either averaging down or forced liquidation.

Risk assessment

Primary risk is concentration rather than liquidity: AAPL, AMZN and JNJ represent 71.5% of invested assets, while Information Technology and Consumer Discretionary together account for approximately 71.0%. AAPL, HD, UNH, GOOGL and META have weaker recent relative trends, and AMZN faces regulatory headline risk despite strong 30-day momentum. MSFT and NVDA are strong but extended growth leaders and could be sensitive to AI-spending, valuation, rates, or geopolitical risk reversals. Cash moderates total portfolio drawdown but does not diversify invested capital.

← September 1, 2026