Glasshouse Fund
Decision journal

AI fund decision · October 6, 2026

Maintain all current positions and retain cash. The constructive index trend and strong AI leadership argue against broad selling, while narrow breadth, sizable concentration, and recent AAPL trimming argue against new purchases or further immediate turnover. Reassess deployment only if breadth improves or a pullback creates more favorable entries.

Outlook: NEUTRALPortfolio $1,050,616Cash 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 (21 outperform, 13 underperform); 0 became trades. These are the fund's calibration record.

beat SPYlag SPYAAPL: outperform at 58%AMD: outperform at 68%AMZN: underperform at 59%APLD: underperform at 56%ASML: outperform at 61%AVGO: underperform at 57%BE: outperform at 55%CAT: outperform at 59%CORZ: outperform at 54%CRWV: outperform at 55%GLW: outperform at 60%GOOGL: outperform at 56%HD: underperform at 70%INTC: outperform at 60%IREN: outperform at 54%JNJ: underperform at 64%JPM: underperform at 62%MA: underperform at 57%META: outperform at 66%MSFT: outperform at 65%MU: outperform at 58%NVDA: outperform at 67%ORCL: outperform at 58%PG: underperform at 55%QQQ: outperform at 63%SNDK: outperform at 57%SPY: outperform at 50%TSLA: outperform at 61%TSM: outperform at 68%UNH: underperform at 58%V: underperform at 55%WMT: underperform at 58%XOM: underperform at 56%^VIX: underperform at 71%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
^VIXUnderperform SPY71%Falling 5-day and 30-day VIX readings alongside record equity prices imply continued relative underperformance of volatility versus SPY, subject to event-shock risk.
HDUnderperform SPY70%HD's -2.39% 5-day and -11.43% 30-day returns are clear relative-momentum deterioration versus SPY.
AMDOutperform SPY68%AMD's 3.98% 5-day and 37.45% 30-day gains indicate exceptional relative strength, though the advance is extended and vulnerable to semiconductor profit-taking.
TSMOutperform SPY68%TSM's 6.32% 5-day and 17.66% 30-day gains establish strong semiconductor leadership versus SPY, albeit after a sharp run.
NVDAOutperform SPY67%NVDA's 5.15% 5-day and 9.99% 30-day gains, plus AI-demand news, support outperformance despite memory-cost and crowded-trade risks.
METAOutperform SPY66%META's 28.34% 30-day gain and positive 5-day momentum demonstrate powerful relative strength, although the rapid advance raises valuation and profit-taking risk.
MSFTOutperform SPY65%MSFT's positive 5-day and 30-day momentum, alongside coverage citing solid AI demand, supports continued outperformance.
JNJUnderperform SPY64%JNJ's -5.47% 5-day and -6.73% 30-day returns show material relative weakness despite remaining above portfolio cost basis.
QQQOutperform SPY63%QQQ's 6.97% 30-day gain versus SPY's 1.97% reflects continued large-cap technology leadership, though this leadership is narrow.
JPMUnderperform SPY62%JPM has negative returns over both 5 and 30 days, materially lagging SPY despite supportive longer-term AI-bank commentary.
ASMLOutperform SPY61%ASML's positive 5-day and 11.69% 30-day returns support continued semiconductor-equipment leadership, tempered by execution concerns in supplied coverage.
TSLAOutperform SPY61%TSLA's positive 5-day and 30-day momentum, supported by better-than-expected delivery coverage, favors relative outperformance but with high volatility.
GLWOutperform SPY60%GLW has a 9.49% 30-day gain and remains above SPY momentum, though customer concentration, demand, and execution risks limit confidence.
INTCOutperform SPY60%INTC's 30.59% 30-day gain supports an outperform call, although the turnaround and foundry narrative remain execution-sensitive.
AMZNUnderperform SPY59%AMZN's 30-day return is negative while SPY advanced, and available coverage references underperformance amid agentic-AI concerns.
CATOutperform SPY59%CAT's positive 5-day and 8.85% 30-day returns show stronger momentum than SPY, consistent with cyclical leadership.
AAPLOutperform SPY58%AAPL has positive 5-day and 30-day momentum and remains a market leader, but its very large portfolio weight and moderate relative strength versus AI leaders constrain conviction.
MUOutperform SPY58%MU's 13.98% 30-day gain supports relative strength, but the supplied memory-price-test headline creates a meaningful cyclical risk.
ORCLOutperform SPY58%ORCL's strong 5-day gain and modestly positive 30-day return support a measured relative-strength view.
UNHUnderperform SPY58%UNH's positive 5-day rebound does not offset its -3.88% 30-day return and continued healthcare-specific uncertainty.
WMTUnderperform SPY58%WMT is below portfolio cost and has negative returns over both 5 and 30 days, indicating persistent relative weakness.
AVGOUnderperform SPY57%AVGO has positive short-term momentum but negative 30-day performance, leaving its relative-strength signal weaker than the broader AI leadership group.
MAUnderperform SPY57%MA's nearly flat 5-day return and -2.84% 30-day return suggest weaker near-term relative performance than SPY.
SNDKOutperform SPY57%SNDK's 10.89% 30-day return supports an outperform call, although negative 5-day momentum signals a possible near-term consolidation.
APLDUnderperform SPY56%APLD has negative 5-day momentum and only modest positive 30-day performance, providing limited evidence it can exceed SPY.
GOOGLOutperform SPY56%GOOGL's positive 5-day and 30-day returns suggest modest improvement versus SPY, supported by long-term Waymo expansion optionality but constrained by the position being below cost.
XOMUnderperform SPY56%XOM's short-term rebound is not enough to overcome its slightly negative 30-day return and below-cost position.
BEOutperform SPY55%BE's 34.18% 30-day advance and raised-revenue-outlook coverage support relative strength, but its volatility and large prior run keep confidence modest.
CRWVOutperform SPY55%CRWV's positive 5-day and 30-day momentum supports a modest relative-strength call, with limited additional evidence.
PGUnderperform SPY55%PG's negative 5-day and slightly negative 30-day performance provide insufficient evidence of relative strength versus SPY.
VUnderperform SPY55%V has positive 5-day momentum but a slightly negative 30-day return, leaving relative performance less compelling than SPY.
CORZOutperform SPY54%CORZ has positive returns on both horizons, but the modest magnitude and limited fundamental evidence warrant low conviction.
IRENOutperform SPY54%IREN's positive 30-day return exceeds SPY, but negative 5-day momentum and limited supporting evidence reduce conviction.
SPYOutperform SPY50%SPY is the benchmark itself; the forced directional call carries neutral-equivalent confidence because it should track its own benchmark.

The debate

Bull, bear and risk cases: bull 0.84 · bear 0.87 · risk 0.98
bull case · conviction 0.84

The evidence supports buying or holding a concentrated group of technology and AI beneficiaries rather than reducing overall equity exposure. Market momentum is constructive: the S&P 500 is up 1.31% over 30 days and has reached a new record high, QQQ is up 6.97%, volatility is falling, and several portfolio holdings are materially outperforming the benchmark. The strongest leadership is concentrated in AMD, TSM, META, NVDA, and GLW, while MSFT, AAPL, and GOOGL are also advancing. News flow adds credible catalysts around AI demand, supplier revenue, and upcoming Apple attention. The case is strongest for maintaining or adding selectively to leaders with superior relative strength, while avoiding indiscriminate buying of weaker holdings.

  • Relative strength is compelling: AMD is up 37.45% over 30 days, META 28.34%, TSM 17.66%, NVDA 9.99%, GLW 9.49%, and MSFT 4.82%, all ahead of the S&P 500's 1.31% gain. This is evidence of sustained leadership rather than merely a one-day move.
  • Near-term momentum is broad among the leaders: NVDA is up 5.15% over five days, TSM 6.32%, AMD 3.98%, MSFT 3.19%, and AAPL 1.06%. NVDA and TSM are especially strong on both five-day and 30-day horizons.
  • The AI catalyst remains visible in the news: Microsoft is described as rising on solid AI demand, Nvidia has reportedly reached a new high as a key supplier's revenue booms on AI demand, and Nvidia revenue is expected in one headline to grow 70%. The headline also flags extreme memory costs as a risk, so the catalyst is strong but not risk-free.
  • Apple has a positive setup despite more moderate momentum: AAPL is up 2.39% over 30 days, the market identifies Apple, Microsoft, and Nvidia as keeping stocks afloat, and an October 13 date is specifically highlighted in Apple-related coverage as a potential attention catalyst.
  • The broader tape is supportive for holding risk assets: the S&P 500 is at a record high, SPY is up 1.97% over 30 days, QQQ is up 6.97%, and the VIX is down 5.02% over 30 days. Falling volatility and leadership from large-cap technology favor staying invested rather than exiting strong names.
  • Alphabet offers an additional longer-term catalyst beyond current 3.48% 30-day momentum: the constructive fund memory says Waymo is expanding domestically, entering international markets, and scaling operations. That provides strategic upside optionality even though GOOGL is currently just below its listed average cost.
  • The bullish case should remain selective. AMZN, JNJ, JPM, MA, PG, UNH, V, WMT, and XOM have negative 30-day momentum, while NVDA's headline notes memory-cost pressure. Those weaker or more contested positions argue for rotation toward demonstrated leaders, not an indiscriminate increase across the entire portfolio.
bear case · conviction 0.87

The bearish case is that portfolio breadth is weakening despite a positive benchmark: the S&P 500 gained 1.31% over 30 days while several holdings and candidates posted meaningful declines. Three current holdings are already below cost, and defensive, financial, consumer, healthcare, energy, and communications names show fading momentum. This creates a risk that weakness broadens beyond the currently red positions.

  • Current losses are concentrated in GOOGL (-1.01%), WMT (-1.20%), and XOM (-2.46%), indicating the portfolio already has positions failing to participate.
  • The clearest momentum deterioration is in JNJ (-6.73%), JPM (-6.36%), UNH (-3.88%), MA (-2.84%), AMZN (-1.38%), and V (-0.79%) over 30 days, despite the S&P 500 gaining 1.31%.
  • HD is the most severe candidate warning, down 11.43% over 30 days; adding exposure while momentum is deteriorating would increase downside risk rather than improve portfolio breadth.
  • AVGO is also down 1.76% over 30 days, making it another candidate with evidence of weakening price momentum.
  • PG is nearly flat but still negative over 30 days (-0.18%), while WMT and XOM are both negative on both the momentum and cost-basis measures, suggesting persistent rather than isolated weakness.
  • GLW's cited risk disclosures identify multiple ways the recent holding gain could reverse, including loss of significant customers, customer financing problems, product or component issues, adverse litigation, regulatory developments, technology change, and difficulty pacing capital spending to demand.
  • The breadth of fading momentum across financials, healthcare, consumer names, energy, and retail argues against treating the weakness as confined to one sector or one company.
risk case · conviction 0.98

The largest capital risk is concentration: 61.0% of invested assets are in AAPL, JNJ, and AMZN, while Information Technology represents 46.7% of invested assets. The 25.9% cash allocation reduces total-portfolio exposure and provides liquidity, but leaves the invested capital materially dependent on a small number of positions and sectors.

  • AAPL is the dominant position at 25.7% of invested assets, approximately 19.0% of the total portfolio including cash.
  • JNJ and AMZN add 17.8% and 17.5% of invested assets, respectively; the three largest holdings comprise 61.0% of invested assets, creating substantial single-name concentration.
  • Information Technology is the largest sector at 46.7% of invested assets, with MSFT, NVDA, AMD, TSM, ORCL, and GLW adding to the sector tilt beyond AAPL.
  • Health Care and Consumer Discretionary are also concentrated at 19.6% and 17.5% of invested assets, respectively, leaving only 16.2% across the remaining listed sectors.
  • Cash is 25.9% of the total portfolio and is a meaningful loss mitigant, but concentration remains high within the 74.1% invested portion.

Cash thesis

Cash of $272,449.26, or 25.9% of portfolio value, is intentional. It offsets substantial concentration in AAPL, JNJ, AMZN, and technology-linked holdings, preserves flexibility for a broader-market pullback or improved breadth, and avoids chasing AI and semiconductor leaders after strong 30-day advances. The opportunity cost is real in a record market, but current evidence supports selective patience rather than broad deployment.

Full commentary

Bear case response

Accept: market breadth is uneven despite positive index returns. JNJ, JPM, UNH, MA, AMZN, V, WMT, XOM, and PG have lagged SPY over 30 days, so broad new buying is not warranted. Accept: GOOGL, WMT, and XOM are below cost, but reject selling solely because of modest unrealized losses; none has a sufficiently new thesis-breaking fact in the supplied information to justify forced liquidation. Accept: JNJ's -6.73% 30-day return, JPM's -6.36%, UNH's -3.88%, and other lagging cyclical/defensive positions require monitoring and no additions. Reject an immediate portfolio-wide risk reduction: the portfolio already holds 25.9% cash, several core leaders retain strong momentum, and indiscriminate selling would risk turning temporary relative weakness into realized losses. Accept: HD's -11.43% 30-day decline and AVGO's -1.76% decline make them unsuitable additions. Accept: GLW's filing-based customer, demand, regulatory, and execution risks mean its 30-day strength should not be chased. Accept: strong AI performance is backward-looking and could invite profit-taking; therefore, no additions are made to NVDA, AMD, TSM, MSFT, META, or AAPL. The decision also respects recent AAPL trimming and prior risk lessons favoring low turnover, disciplined concentration management, and avoiding repeated trades.

Market summary

The market backdrop is constructive but narrow: SPY is up 1.97% over 30 days, QQQ is up 6.97%, the S&P 500 is at a record high, and VIX is 15.07 after declining over both 5 and 30 days. Leadership is concentrated in AI, semiconductors, and selected megacaps, while healthcare, financials, consumer defensives, energy, and several portfolio laggards have weak relative momentum. This supports retaining core exposure but not deploying cash aggressively into an extended and concentrated leadership group.

Portfolio assessment

The portfolio value is $1,050,615.72, with 25.9% in cash. AAPL ($200,155.85), JNJ ($138,296.67), and AMZN ($136,166.98) remain the central concentration issue, together representing roughly 45.2% of total portfolio value and a substantially larger share of invested assets. The portfolio has meaningful gains in AAPL, MSFT, NVDA, META, TSM, GLW, and AMD, while GOOGL, WMT, and XOM are below cost. Recent AAPL trimming on October 1 appropriately reduced single-name exposure; further immediate turnover is not justified by the available evidence.

Risk assessment

The primary risk is concentrated exposure rather than inadequate equity participation. AAPL alone is about 19.0% of total portfolio value, while JNJ and AMZN are each near 13%. Technology and AI-related exposures across AAPL, MSFT, NVDA, AMD, TSM, ORCL, GLW, GOOGL, META, and AMZN are correlated to a common growth, AI-spending, valuation, and earnings-execution narrative. A narrow rally could reverse if AI expectations, memory costs, yields, regulation, or guidance disappoint. The cash reserve moderates total-portfolio drawdown, but it does not remove losses from the concentrated invested sleeve. This conclusion follows the high-priority concentration and incremental-trimming discipline in risk memory.

← October 2, 2026