Glasshouse Fund
Decision journal

AI fund decision · September 18, 2026

Maintain a neutral, low-turnover posture. Hold core relative-strength leaders without adding, retain the elevated cash reserve, and monitor AMZN and other weakening holdings for persistent deterioration rather than reacting to a single session or chasing watchlist declines.

Outlook: NEUTRALPortfolio $1,045,195Cash 27.2%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 (16 outperform, 18 underperform); 0 became trades. These are the fund's calibration record.

beat SPYlag SPYAAPL: outperform at 70%AMD: outperform at 56%AMZN: underperform at 61%APLD: underperform at 67%ASML: underperform at 66%AVGO: underperform at 63%BE: outperform at 60%CAT: underperform at 61%CORZ: underperform at 69%CRWV: underperform at 75%GLW: underperform at 66%GOOGL: outperform at 57%HD: underperform at 68%INTC: outperform at 57%IREN: underperform at 55%JNJ: outperform at 65%JPM: underperform at 57%MA: underperform at 53%META: outperform at 68%MSFT: outperform at 56%MU: outperform at 54%NVDA: underperform at 56%ORCL: outperform at 52%PG: outperform at 55%QQQ: underperform at 56%SNDK: underperform at 54%SPY: outperform at 50%TSLA: outperform at 56%TSM: outperform at 55%UNH: underperform at 65%V: outperform at 56%WMT: underperform at 64%XOM: outperform at 54%^VIX: underperform at 53%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 SPY75%The 24.11% 30-day loss and 10.24% five-day decline indicate severe negative momentum despite backlog-related bullish commentary.
AAPLOutperform SPY70%Strong 30-day and 5-day momentum plus current product-launch attention support relative strength versus SPY, though concentration limits portfolio adding.
CORZUnderperform SPY69%Despite AI-pivot interest, the 15.06% 30-day decline and negative 5-day return signal unfavorable momentum versus SPY.
HDUnderperform SPY68%The 10.07% 30-day decline and negative five-day trend show persistent weakness versus SPY.
METAOutperform SPY68%The strongest positive 5-day and 30-day momentum among current holdings supports outperformance, although the stock may be extended.
APLDUnderperform SPY67%The 15.45% 30-day decline and lack of confirmed stabilization leave the risk-reward unfavorable versus SPY.
ASMLUnderperform SPY66%Materially negative 5-day and 30-day momentum signals continued semiconductor-equipment weakness versus SPY.
GLWUnderperform SPY66%A 10.79% 30-day and 11.18% five-day decline outweighs the positive mark versus cost and leaves execution risks elevated.
JNJOutperform SPY65%Positive 5-day and 30-day returns in a declining broad market support defensive relative strength versus SPY.
UNHUnderperform SPY65%The 6.03% 30-day decline and negative five-day performance indicate persistent weakness versus SPY.
WMTUnderperform SPY64%The 7.13% 30-day decline and negative five-day return indicate weaker defensive-consumer momentum versus SPY.
AVGOUnderperform SPY63%AI growth commentary is constructive, but the stock's 11.63% 30-day decline and recent weakness argue for near-term underperformance risk.
AMZNUnderperform SPY61%Negative 5-day and 30-day returns indicate weaker near-term momentum than SPY despite the company's strategic initiatives.
CATUnderperform SPY61%Negative 5-day and 30-day performance suggests cyclical momentum remains weaker than SPY.
BEOutperform SPY60%The 22.10% 30-day gain and data-center-related catalyst support relative strength, offset by elevated volatility.
GOOGLOutperform SPY57%Positive five-day momentum and a relatively resilient 30-day result versus a declining SPY support modest outperformance odds, though conviction is limited.
INTCOutperform SPY57%Positive 5-day and 30-day momentum supports potential relative strength, but the semiconductor group remains volatile.
JPMUnderperform SPY57%Negative 5-day and 30-day momentum, alongside rising-yield uncertainty, modestly favor underperformance versus SPY.
AMDOutperform SPY56%Positive 5-day and 30-day momentum supports modest relative-strength potential versus SPY, but semiconductor volatility keeps confidence low.
MSFTOutperform SPY56%Modestly positive 5-day and 30-day returns are better than SPY's 30-day decline, supporting a narrow relative-strength edge.
NVDAUnderperform SPY56%The negative 30-day return and uneven semiconductor breadth outweigh the modest positive five-day move.
QQQUnderperform SPY56%Negative 30-day performance and rising-yield pressure on technology leave QQQ modestly vulnerable versus SPY.
TSLAOutperform SPY56%The positive 6.99% 30-day return supports relative strength versus SPY, though the five-day trend is nearly flat.
VOutperform SPY56%Positive 30-day performance and nearly flat five-day action support a modest relative-strength outlook versus SPY.
IRENUnderperform SPY55%Slightly negative short- and medium-term returns provide limited evidence for outperformance versus SPY.
PGOutperform SPY55%Positive five-day and 30-day returns support modest defensive outperformance potential versus SPY.
TSMOutperform SPY55%Slightly positive 30-day performance and limited recent weakness support a modest edge versus SPY.
MUOutperform SPY54%Slightly positive short- and medium-term momentum supports a marginal relative-strength case versus SPY.
SNDKUnderperform SPY54%Negative 5-day and 30-day returns indicate slightly weaker momentum than SPY.
XOMOutperform SPY54%Positive 30-day performance provides a modest relative-strength advantage versus SPY despite a weak five-day trend.
MAUnderperform SPY53%Near-flat but negative recent momentum offers only a slight underperformance signal versus SPY.
^VIXUnderperform SPY53%With the VIX down over five days and volatility currently contained, a modest normalization lower is more likely than sustained relative strength versus SPY.
ORCLOutperform SPY52%Near-flat but slightly positive momentum is marginally better than SPY's negative 30-day performance.
SPYOutperform SPY50%SPY is the comparison benchmark and is expected to track itself; confidence is neutral by definition.

The debate

Bull, bear and risk cases: bull 0.77 · bear 0.90 · risk 0.98
bull case · conviction 0.77

The strongest evidence supports holding, and selectively adding to, the portfolio's relative-strength leaders, especially AAPL, META, JNJ, and GOOGL. This is a favorable stock-selection environment for these names: the S&P 500 declined 0.91% over 30 days, while AAPL rose 10.16%, META 15.68%, JNJ 4.30%, and GOOGL 0.48%. AAPL and META are showing particularly strong positive momentum despite a market backdrop marked by rising yields and a lower benchmark. Near-term catalysts further support AAPL and JNJ, while Alphabet has a documented long-term growth option in Waymo. The case is strongest for holding existing leaders rather than aggressively buying every name, because the evidence does not include valuation, earnings, or detailed fundamentals.

  • AAPL is the clearest large-cap momentum leader: up 10.16% over 30 days and 1.42% over five days, outperforming the S&P 500 by roughly 11 percentage points over 30 days.
  • AAPL has an immediate product catalyst: the iPhone 18 Pro began selling today, and a Barron's headline cites potential for a 13% rise if demand persists. That forecast is conditional, but it provides a concrete catalyst aligned with already-positive price action.
  • META has the strongest momentum among the listed portfolio holdings: up 15.68% over 30 days and 5.29% over five days, outperforming the benchmark by roughly 16.6 percentage points over 30 days.
  • JNJ combines positive momentum with defensive relative strength: it gained 4.30% over 30 days and 1.75% over five days while the benchmark fell. Recent coverage specifically notes an 18% gain in three months, reinforcing the trend.
  • GOOGL has improving short-term relative strength: up 2.61% over five days and roughly flat over 30 days while the S&P 500 declined. The constructive fund evidence identifies Waymo as an expanding business scaling into more domestic cities and entering international markets.
  • Additional holdings are also outperforming the benchmark over 30 days, including V, MA, TSM, XOM, PG, and MSFT on the supplied data, supporting a hold bias for the portfolio's stronger franchises rather than broad risk reduction.
  • The market backdrop is not uniformly bullish: rising bond yields, AI-related scrutiny, and a 0.91% benchmark decline argue for selectivity. The evidence favors concentrating incremental exposure in demonstrated relative-strength names, not chasing the many weak performers such as HD, AMZN, UNH, WMT, or GLW.
bear case · conviction 0.90

The strongest bearish case is that portfolio breadth and momentum are deteriorating at the same time. The S&P 500 is down 0.9% over 30 days, while several current holdings are materially weaker than the benchmark and multiple candidates show sharper declines. This is not isolated to one position: weakness spans consumer, healthcare, financial, technology, semiconductor, and communications-related names. The portfolio also has limited evidence of near-term leadership in several holdings, while recent winners remain exposed to a broader pullback in risk appetite.

  • HD is a clear deterioration signal: the position is down 8.4% from cost and has fallen 10.1% over 30 days, substantially underperforming the 0.9% benchmark decline.
  • UNH is down 7.1% from cost and 6.0% over 30 days, indicating continuing weakness rather than a merely temporary mark-to-market loss.
  • AMZN is down 4.4% over 30 days despite being modestly above its average cost, creating a risk that the current gain is being eroded by deteriorating momentum.
  • WMT is down 7.1% over 30 days and 1.9% from cost, while PG is slightly below cost; even traditionally defensive consumer exposure is not providing consistent strength.
  • GLW is up from cost but down 10.8% over 30 days, making the gain vulnerable to reversal. The available company risk disclosures specifically identify customer demand, customer financing and receivables, loss of significant customers, technology change, component performance, litigation, regulation, and the ability to improve margins as risks.
  • JPM and MA are both losing momentum, down 3.7% and 0.6% over 30 days respectively. Their positive marks versus cost do not remove the risk that recent performance is rolling over.
  • NVDA is down 2.5% over 30 days, and the broader candidate list shows sharper weakness in ASML (-11.6%), AVGO (-11.6%), SNDK (-1.6%), and other technology-linked names. This argues against assuming semiconductor-related exposure will automatically lead a recovery.
  • The candidate set contains especially severe momentum deterioration: CRWV is down 24.1%, APLD 15.4%, CORZ 15.1%, ASML 11.6%, AVGO 11.6%, and CAT 6.8% over 30 days. Buying into these declines would risk treating falling momentum as an opportunity without evidence of stabilization.
  • GOOGL, PG, WMT, and XOM are all below average cost, while HD and UNH are materially below cost. Losses are therefore distributed across several holdings rather than concentrated in a single outlier.
  • The benchmark itself is down over the same 30-day period, so the portfolio is operating in a weakening market backdrop. In that environment, positions with negative absolute performance and negative recent momentum have a higher risk of further drawdown before any recovery is confirmed.
risk case · conviction 0.98

The primary capital risks are concentrated single-name exposure and a pronounced technology tilt. AAPL, JNJ, and AMZN comprise 66.8% of invested assets, while the top five positions comprise 82.8%. Information Technology is 46.3% of invested assets, creating material dependence on one sector. Cash at 27.2% reduces total portfolio market exposure, but does not diversify the invested equity book; the deployed capital remains heavily concentrated.

  • AAPL is the largest position at 29.8% of invested assets, making it the clearest single-name concentration risk.
  • JNJ and AMZN add 19.3% and 17.7%, respectively; together with AAPL they represent 66.8% of invested assets, or approximately 48.6% of the total portfolio after accounting for cash.
  • The five largest positions. AAPL, JNJ, AMZN, MSFT, and GOOGL, represent 82.8% of invested assets, leaving limited diversification among deployed capital.
  • Information Technology is 46.3% of invested assets, the largest sector exposure; Consumer Discretionary and Health Care add 19.7% and 19.4%, respectively.
  • The portfolio has limited exposure outside its dominant sectors: Consumer Staples 3.1%, Financials 2.1%, and Energy 1.5% of invested assets.
  • Cash is 27.2% of the portfolio and provides a buffer against losses in the invested book, but the remaining capital is concentrated rather than broadly diversified.
  • The evidence does not establish a directional view; the key risk is that adverse outcomes in a small number of holdings or in Information Technology could have an outsized effect on portfolio capital.

Cash thesis

Cash at 27.2% is intentionally high because the equity book remains concentrated in AAPL, JNJ, AMZN, MSFT, and GOOGL, while market breadth is uneven and yields are rising. The reserve reduces total-portfolio drawdown sensitivity and preserves flexibility to deploy only after broader confirmation or a more compelling risk-reward opportunity. It is not a directional market call; it is compensation for concentration risk and insufficient confirmation for new purchases.

Full commentary

Bear case response

Accepted: HD's 10.07% 30-day decline and 8.40% loss versus cost indicate real deterioration, but its approximately 1.4% portfolio weight makes an immediate sale low-impact; hold rather than repeatedly trade a small position. Accepted: UNH, WMT, and GLW show persistent negative 30-day momentum, and GLW's disclosed operational, customer, financing, and regulatory risks warrant caution; none merits an add. Accepted: AMZN is down 4.36% over 30 days despite remaining above cost, which is concerning given its large size. However, the position has already been reduced relative to the August memory snapshot, and there is no new catalyst evidence requiring another immediate trim today. Accepted: JPM, MA, and NVDA have softened recently, so prior gains should not be treated as downside protection; this supports no new purchases. Accepted: weakness in ASML, AVGO, APLD, CORZ, and CRWV argues against buying fallen technology or AI-infrastructure watchlist names without stabilization. Rejected: the breadth weakness does not justify broad liquidation because volatility is contained, the benchmark decline is modest, and AAPL, META, JNJ, MSFT, V, MA, and TSM retain relative strength. The appropriate response is to preserve cash, avoid new adds, and monitor large concentrated positions rather than make broad risk-off sales.

Market summary

The broad market is soft but not disorderly: SPY is down 1.77% over 30 days, QQQ is down 1.94%, and rising Treasury yields are pressuring equities, while the VIX at 14.9 remains relatively contained. Leadership is narrow, led by AAPL, META, JNJ, and selected technology franchises, while consumer, healthcare, semiconductor, and infrastructure-related breadth is mixed to weak. This supports patience rather than either aggressive deployment of cash or wholesale liquidation.

Portfolio assessment

The portfolio has strong absolute gains in AAPL, JNJ, MSFT, META, V, and MA, but deployed capital remains highly concentrated. AAPL, JNJ, and AMZN account for approximately 48.6% of total portfolio value including cash, while AAPL alone is roughly 21.7%. AMZN's 4.36% 30-day decline is the most important near-term monitoring issue because it remains a large position, but it is still above cost and has already been reduced from prior reported levels. Fund memory supports maintaining low turnover: recent risk lessons identify concentration as the primary risk and advise selective, incremental action rather than repeated trading. The current evidence does not justify adding to concentrated leaders or selling core positions solely because of short-term weakness.

Risk assessment

The dominant risk is concentration rather than a market-wide volatility spike. Technology represents a large share of invested capital, and correlated weakness across AI, cloud, semiconductors, and consumer-discretionary exposure could affect AAPL, MSFT, NVDA, AMZN, GOOGL, TSM, and GLW together. AAPL's strong momentum is beneficial but increases single-name dependence; AMZN's negative 30-day momentum is material because of its size. HD, UNH, WMT, and GLW show meaningful recent weakness, but their smaller weights limit their immediate portfolio impact. Holding cash mitigates overall exposure but does not diversify the invested book.

← September 17, 2026 September 21, 2026 →