Glasshouse Fund
Decision journal

AI fund decision · September 16, 2026

Maintain a neutral, low-turnover posture. Hold established relative-strength leaders, avoid averaging down into weak trends, and preserve the near-25% cash reserve until market breadth improves. This approach follows the higher-priority memory lessons on concentration, disciplined incremental action, and avoiding repeated turnover.

Outlook: NEUTRALPortfolio $1,039,414Cash 25.0%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 (11 outperform, 23 underperform); 0 became trades. These are the fund's calibration record.

beat SPYlag SPYAAPL: outperform at 68%AMD: underperform at 62%AMZN: underperform at 66%APLD: underperform at 79%ASML: underperform at 72%AVGO: underperform at 75%BE: underperform at 55%CAT: underperform at 64%CORZ: underperform at 80%CRWV: underperform at 82%GLW: underperform at 73%GOOGL: outperform at 57%HD: underperform at 71%INTC: underperform at 66%IREN: underperform at 65%JNJ: outperform at 63%JPM: underperform at 56%MA: outperform at 60%META: outperform at 70%MSFT: outperform at 59%MU: underperform at 61%NVDA: underperform at 63%ORCL: underperform at 76%PG: outperform at 55%QQQ: underperform at 58%SNDK: underperform at 67%SPY: outperform at 50%TSLA: underperform at 54%TSM: underperform at 60%UNH: underperform at 65%V: outperform at 64%WMT: underperform at 58%XOM: outperform at 58%^VIX: outperform at 69%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
CRWVUnderperform SPY82%A five-day decline of 14.8% and 30-day decline of 24.9% outweigh isolated positive infrastructure headlines.
CORZUnderperform SPY80%The 11.3% five-day and 23.1% 30-day declines show persistent high-volatility downside momentum.
APLDUnderperform SPY79%Severe negative five-day and 30-day momentum indicates unresolved downside momentum rather than a confirmed entry setup.
ORCLUnderperform SPY76%The 13.2% five-day and 8.4% 30-day declines, combined with current corporate uncertainty headlines, indicate unfavorable momentum.
AVGOUnderperform SPY75%Negative 5-day and 30-day returns, including an 18.5% monthly decline, outweigh after-hours and AI-demand commentary.
GLWUnderperform SPY73%The 14.1% 30-day decline and 14.8% five-day decline, alongside disclosed execution and demand risks, point to continued relative weakness.
ASMLUnderperform SPY72%The 8.0% five-day and 12.1% 30-day declines indicate substantial semiconductor-equipment momentum deterioration.
HDUnderperform SPY71%A 10.4% 30-day decline, negative recent momentum, and a loss versus cost indicate deterioration without confirmed stabilization.
METAOutperform SPY70%Strong positive momentum over five days and 30 days makes META one of the clearest relative-strength leaders.
^VIXOutperform SPY69%The VIX has risen 18.2% over 30 days while SPY declined, and elevated rate and market uncertainty can sustain relative volatility strength.
AAPLOutperform SPY68%Positive 5-day and 30-day momentum, product-launch strength, and relative leadership versus SPY support outperformance, though the position is extended and concentrated.
SNDKUnderperform SPY67%The sharp 13.2% five-day decline and volatile valuation backdrop outweigh its positive 30-day return.
AMZNUnderperform SPY66%The 7.1% 30-day decline and 1.6% five-day decline indicate weakening relative support despite the stock remaining slightly above portfolio cost.
INTCUnderperform SPY66%Negative five-day and 30-day performance leaves the stock in a weaker trend than SPY.
IRENUnderperform SPY65%Negative momentum over both measured periods and a volatile thematic profile are unfavorable in the current risk-sensitive tape.
UNHUnderperform SPY65%The position is below cost and negative over both five and 30 days, with no supplied catalyst demonstrating stabilization.
CATUnderperform SPY64%Negative returns over both five and 30 days indicate cyclical weakness without a cited offsetting catalyst.
VOutperform SPY64%Positive five-day and 30-day momentum and a strong embedded gain support continued relative resilience.
JNJOutperform SPY63%Positive 30-day performance and flat five-day action demonstrate defensive relative resilience versus a declining SPY.
NVDAUnderperform SPY63%Negative returns over five and 30 days show fading semiconductor leadership despite continued long-term AI relevance.
AMDUnderperform SPY62%Negative five-day momentum and only modest positive 30-day performance amid broad semiconductor weakness suggest a less favorable risk-adjusted outlook than SPY.
MUUnderperform SPY61%The 9.7% five-day decline signals heightened semiconductor volatility despite a slightly positive 30-day return.
MAOutperform SPY60%Positive five-day and 30-day returns indicate ongoing relative strength, though the sizable embedded gain creates mean-reversion risk.
TSMUnderperform SPY60%Negative five-day and 30-day returns indicate that semiconductor exposure remains under pressure.
MSFTOutperform SPY59%Modest positive five-day and 30-day returns versus a falling SPY support relative outperformance, tempered by AI spending and valuation sensitivity.
QQQUnderperform SPY58%Negative five-day and 30-day performance reflects pressure on growth and technology exposure in the current rate-sensitive environment.
WMTUnderperform SPY58%Recent five-day strength is not enough to offset a 6.6% 30-day decline and near-flat return versus portfolio cost.
XOMOutperform SPY58%Positive five-day and 30-day returns, supported by oil-strength headlines, provide a relative-performance advantage despite trading below cost.
GOOGLOutperform SPY57%The 4.3% five-day rebound and Waymo scaling optionality support modest relative upside, but conviction is limited by the near-flat 30-day return and position below cost.
JPMUnderperform SPY56%The 3.5% 30-day decline and slightly negative five-day trend imply fading momentum despite strong gains versus portfolio cost.
BEUnderperform SPY55%Positive 30-day performance is offset by recent negative momentum and a higher-rate backdrop that is less supportive of volatile growth infrastructure.
PGOutperform SPY55%Positive five-day and 30-day returns provide modest defensive relative support, although the magnitude is limited.
TSLAUnderperform SPY54%The positive 30-day return is offset by recent negative momentum and elevated volatility, leaving limited confidence in sustained relative strength.
SPYOutperform SPY50%SPY is the comparison benchmark, so its expected relative performance versus itself is neutral by definition.

The debate

Bull, bear and risk cases: bull 0.78 · bear 0.94 · risk 0.97
bull case · conviction 0.78

Buy or hold selectively, led by AAPL, META, V, JNJ, MSFT, and GOOGL. The strongest evidence is relative strength: the S&P 500 fell 2.1% over 30 days, while AAPL rose 9.6%, META 15.8%, V 4.5%, and JNJ 2.9%; MSFT was also modestly positive at 1.1%. This leadership is occurring despite rising volatility, with the VIX up 18.2% over 30 days. Company-specific catalysts reinforce the momentum case: Apple has delivered its largest product-launch-related surge in years and received renewed positive commentary from Morgan Stanley; Microsoft continues to benefit from an AI investment thesis; Alphabet offers potential long-term optionality through Waymo's geographic and operational expansion. The portfolio also has meaningful embedded gains in JNJ, AAPL, MSFT, META, JPM, V, and MA, supporting a hold-and-add-on-weakness posture rather than wholesale de-risking. The bull case is strongest in demonstrated leaders, while weaker momentum names should not receive equal conviction.

  • AAPL is the clearest large-cap momentum leader in the evidence: up 5.1% over five days and 9.6% over 30 days versus the S&P 500 down 2.1% over 30 days.
  • META has the strongest listed 30-day momentum among portfolio holdings, up 15.8%, with positive five-day momentum of 2.5%; its current price is also well above the portfolio average cost.
  • V combines positive five-day and 30-day momentum, up 2.2% and 4.5%, respectively, while the portfolio position is substantially above its average cost.
  • JNJ has gained 2.9% over 30 days and is essentially flat over five days, showing relative resilience in a riskier tape; news also highlights its long-term compounding appeal and recent share surge.
  • MSFT is modestly positive over both five and 30 days despite the market decline, and current news continues to frame AI spending and the AI bear case as key upside-versus-risk debates rather than a broken thesis.
  • GOOGL has recently accelerated, rising 4.3% over five days, and its constructive fund memory cites Waymo's expansion into more cities, international markets, and scaled operations as a long-term catalyst.
  • The portfolio already has substantial unrealized gains in JNJ, AAPL, MSFT, JPM, V, MA, and META, arguing for holding proven winners while avoiding unnecessary turnover.
  • The evidence is not uniformly bullish: AMZN, HD, NVDA, TSM, UNH, and GLW show negative 30-day momentum, while the VIX is up 18.2% and market news cites higher rates and elevated yields. These factors favor selective buying and holding rather than indiscriminate accumulation.
bear case · conviction 0.94

The portfolio is exposed to broad market weakness and several holdings with clear deterioration in price momentum. The strongest risks are concentrated in HD, UNH, AMZN, GLW, NVDA, TSM and WMT, while multiple prospective names show substantially worse declines and should not be treated as attractive candidates without evidence of stabilization. Several recent winners remain vulnerable if the market weakness broadens.

  • The S&P 500 is down 2.1% over 30 days, while SPY is down 2.0% and QQQ is down 2.6%, providing a weak backdrop for equity holdings and growth exposure.
  • HD is both below cost by 6.2% and down 10.4% over 30 days; this is one of the clearest cases of existing loss combined with deteriorating momentum.
  • UNH is below cost by 6.7% and down 6.7% over 30 days, indicating sustained weakness rather than merely a short-term pullback.
  • AMZN is down 7.1% over 30 days despite trading only modestly above its average cost, leaving recent gains exposed to further erosion.
  • GLW is down 14.1% over 30 days despite being slightly above cost. Its disclosed risks include weaker-than-expected customer demand, difficulty pacing capital spending, margin-improvement execution, technology change, product or component issues, loss of significant customers, litigation and regulatory developments.
  • NVDA and TSM are down 5.2% and 3.6% over 30 days, respectively, challenging the durability of the semiconductor exposure even though both remain near or above cost.
  • JPM is up substantially versus cost but down 3.5% over 30 days; the fading momentum creates a risk that an established gain becomes less defensible if market weakness persists.
  • WMT is essentially flat versus cost and down 6.6% over 30 days, offering little recent price support despite its defensive profile.
  • GOOGL and XOM are already below cost by 1.6% and 2.4%, respectively; their inclusion adds to the number of holdings failing to generate positive returns.
  • Prospective names show materially worse momentum: APLD is down 24.8%, CRWV 24.9%, CORZ 23.1%, AVGO 18.5% and ASML 12.1% over 30 days. These are not isolated mild pullbacks and represent substantial downside momentum.
  • The portfolio contains several positions that are currently profitable but weakening, including NVDA, TSM, JPM and AMZN. That combination can create asymmetric risk: historical gains may encourage holding while the latest evidence points toward deteriorating market support.
  • The evidence does not show a positive momentum offset strong enough to neutralize the weakness: the benchmark, SPY and QQQ are all down, and the largest negative signals include both current holdings and several candidates.
risk case · conviction 0.97

The largest capital risk is concentration: 64.4% of invested assets is held in AAPL, JNJ, and AMZN, with AAPL alone at 28.9% of invested assets. Sector exposure is also tilted toward Information Technology (44.8%) and Consumer Discretionary (22.1%), while cash is a substantial 25.0% of the portfolio. Cash limits immediate market exposure, but the invested portion remains materially dependent on a small number of positions and sectors.

  • AAPL is the largest position at 28.9% of invested assets, making it the clearest single-name source of capital concentration.
  • The three largest positions. AAPL, JNJ, and AMZN, represent 64.4% of invested assets, or approximately 48.3% of the total portfolio after accounting for cash.
  • Information Technology is the largest sector at 44.8% of invested assets, creating a material sector tilt; Consumer Discretionary adds another 22.1%.
  • JNJ is an 18.7% invested-asset position, so the portfolio is also significantly exposed to one Health Care company despite the sector being 18.8% overall.
  • The 25.0% cash allocation reduces invested-market exposure and provides liquidity, but it does not diversify the capital that is deployed.
  • The evidence supports concentration and sector-tilt risks, but does not provide volatility, correlation, or drawdown data to quantify how these exposures would behave under stress.

Cash thesis

Cash is 24.95%, just below the 25% threshold, but remains intentionally substantial. It offsets the portfolio's concentrated large-cap technology, consumer-discretionary, and single-name exposures while preserving flexibility for a clearer market stabilization or a selective pullback in proven leaders. Deploying cash into declining semiconductors, infrastructure names, or consumer cyclicals before momentum improves would conflict with the fund's risk lessons.

Full commentary

Bear case response

Accept the broad-market concern: SPY and QQQ are negative over 30 days, VIX is higher, and rate/yield news supports a defensive posture. Accept the case against adding to HD, UNH, AMZN, GLW, NVDA, TSM, WMT, and the deeply declining watchlist names; their negative momentum is insufficiently stabilized. Accept that GLW's disclosed operating risks make averaging down inappropriate. Accept that JPM's recent 30-day weakness warrants monitoring despite its gain versus cost. Accept the concentration critique: AAPL, JNJ, and AMZN remain the principal portfolio-level vulnerability, and cash does not diversify deployed capital. Reject a wholesale de-risking conclusion today: AAPL, META, V, JNJ, MSFT, and MA display relative resilience or positive momentum, and no evidence indicates a fundamental thesis break requiring forced sales. The appropriate response is to hold leaders, avoid new risk in laggards, retain liquidity, and reconsider trims only if weakness persists or concentration rises further.

Market summary

The broad backdrop is cautious: the S&P 500 is down 2.1% over 30 days, SPY is down 2.0%, QQQ is down 2.6%, and the VIX has risen 18.2%. The Fed-rate-hike and elevated-yield headlines support patience. Leadership remains narrow but meaningful, led by AAPL, META, V, JNJ, MSFT, and the recent five-day rebound in GOOGL. Semiconductor, infrastructure, consumer-cyclical, and several speculative watchlist names show broad negative momentum.

Portfolio assessment

The portfolio is modestly above $1.039 million with approximately 25.0% in cash and remains concentrated in AAPL ($225.4k), JNJ ($145.8k), and AMZN ($131.4k). AAPL, JNJ, MSFT, META, V, MA, and JPM have substantial gains versus cost, but the portfolio has uneven current momentum: AMZN, HD, GLW, NVDA, TSM, UNH, and WMT are weaker. No trade is warranted today because the market lacks broad confirmation, the strongest holdings are already sizable, and prior fund memory favors low turnover and incremental action rather than reacting to short-term price moves.

Risk assessment

The main risk is concentration rather than liquidity. AAPL, JNJ, and AMZN comprise a dominant portion of deployed capital, while technology and consumer-discretionary exposures could decline together if rates, AI-spending concerns, or consumer demand weaken. AAPL's strength is beneficial but also makes it a large source of event and profit-taking risk. AMZN's negative 30-day return is particularly relevant because it is a major position. Memory reinforces avoiding repeated turnover while using selective, incremental trims only if deterioration persists. Sources: risk_lesson:run_20260729T184341Z_54100baf:report_risk, risk_lesson:run_20260803T191113Z_b5f11a59:report_risk, report_summary:run_20260826T160411Z_f30a1737.

← September 15, 2026 September 17, 2026 →