Glasshouse Fund
Decision journal

AI fund buys AMD, ORCL, HD, sells AAPL · October 1, 2026

Maintain a neutral, low-turnover posture. Execute one incremental AAPL trim to reduce the portfolio's largest concentration, hold other positions, retain the elevated cash reserve, and avoid chasing semiconductor or AI leaders until broader market confirmation improves.

Outlook: NEUTRALPortfolio $1,036,629Cash 25.1%Grounding: ok

Trades

SELL 75 AAPL78% · filled at $330

Trim the largest position by roughly $24,700 to reduce single-name concentration. AAPL remains above cost and has outperformed over 30 days, but its 5-day return is negative, the portfolio already has substantial technology exposure, and fund memory prioritizes incremental concentration reduction over adding to large winners.

Risks the AI named

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 (17 outperform, 17 underperform); 5 became trades. These are the fund's calibration record.

beat SPYlag SPYAAPL: outperform at 60%AMD: outperform at 56%AMZN: underperform at 59%APLD: underperform at 70%ASML: outperform at 62%AVGO: underperform at 55%BE: outperform at 58%CAT: outperform at 53%CORZ: underperform at 73%CRWV: underperform at 54%GLW: outperform at 52%GOOGL: outperform at 53%HD: underperform at 69%INTC: outperform at 57%IREN: underperform at 68%JNJ: underperform at 55%JPM: underperform at 64%MA: underperform at 63%META: outperform at 55%MSFT: outperform at 66%MU: outperform at 61%NVDA: outperform at 63%ORCL: underperform at 68%PG: outperform at 52%QQQ: outperform at 56%SNDK: outperform at 57%SPY: underperform at 50%TSLA: underperform at 62%TSM: outperform at 68%UNH: underperform at 72%V: underperform at 61%WMT: underperform at 60%XOM: outperform at 57%^VIX: underperform at 54%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
CORZUnderperform SPY73%Sharp negative 5-day and 30-day performance signals continued high-risk underperformance.
UNHUnderperform SPY72%tradedNegative returns from cost and across both 5-day and 30-day periods show the clearest held-position deterioration.
APLDUnderperform SPY70%Deep negative 5-day and 30-day returns indicate materially weaker momentum than SPY.
HDUnderperform SPY69%tradedMaterially negative 5-day and 30-day momentum points to persistent weakness versus SPY.
IRENUnderperform SPY68%The steep 5-day decline and only flat 30-day performance indicate unfavorable momentum.
ORCLUnderperform SPY68%tradedNegative momentum over both measured periods signals sustained relative weakness.
TSMOutperform SPY68%Positive returns over both 5 and 30 days, alongside semiconductor strength, provide the clearest relative-strength setup.
MSFTOutperform SPY66%Positive 5-day acceleration, positive 30-day performance and favorable analyst commentary support relative strength.
JPMUnderperform SPY64%Negative 5-day and 30-day returns indicate weakening financial-sector relative performance.
MAUnderperform SPY63%The negative 30-day trend and weak 5-day return point to underperformance despite the position's positive unrealized gain.
NVDAOutperform SPY63%Positive 5-day momentum, a positive 30-day result and the reported buyback authorization support relative strength.
ASMLOutperform SPY62%Positive 5-day and 30-day momentum supports continued semiconductor-equipment relative strength.
TSLAUnderperform SPY62%The sharp negative 5-day return and flat 30-day result point to unfavorable relative momentum.
MUOutperform SPY61%Strong 30-day semiconductor momentum outweighs the modest negative 5-day move.
VUnderperform SPY61%tradedNegative 5-day and 30-day returns indicate weak payments-sector momentum versus SPY.
AAPLOutperform SPY60%Its 30-day gain materially exceeds SPY despite modest 5-day softness, supporting relative strength but not an additional purchase.
WMTUnderperform SPY60%The negative 5-day return and price below cost outweigh its modestly positive 30-day result.
AMZNUnderperform SPY59%Negative 30-day momentum and AI-spending concerns leave Amazon weaker than SPY until price confirmation improves.
BEOutperform SPY58%The strong 30-day advance and positive near-term momentum favor relative outperformance, albeit with high volatility.
INTCOutperform SPY57%The very strong 30-day gain supports relative momentum, although the negative 5-day move makes the signal volatile.
SNDKOutperform SPY57%The strong 30-day advance supports outperformance, though the sector's volatility and recent pullback reduce confidence.
XOMOutperform SPY57%Positive returns over both 5 and 30 days favor modest relative strength despite the position remaining below cost.
AMDOutperform SPY56%tradedStrong semiconductor momentum and a 28.3% 30-day gain support relative strength, though the move is extended and volatile.
QQQOutperform SPY56%Its positive 30-day return supports a modest advantage over SPY despite nearly flat 5-day momentum.
AVGOUnderperform SPY55%Its negative 30-day return and only marginal 5-day improvement provide insufficient evidence of durable relative leadership.
JNJUnderperform SPY55%Negative 5-day momentum and a slightly negative 30-day return suggest modest relative weakness despite gains from cost.
METAOutperform SPY55%Its exceptional 30-day gain supports relative strength, but the 6.7% 5-day decline raises profit-taking risk.
CRWVUnderperform SPY54%Negative 5-day momentum with only a marginal 30-day gain offers weak relative-strength evidence.
^VIXUnderperform SPY54%A rising volatility index reflects uncertainty, but a sustained risk-on recovery would likely restrain VIX relative performance.
CATOutperform SPY53%Stable positive 5-day momentum and a nearly flat 30-day result modestly compare favorably with weak SPY momentum.
GOOGLOutperform SPY53%Positive 5-day and 30-day returns are slightly better than SPY, though the position remains below cost and execution concerns remain.
GLWOutperform SPY52%Flat recent trading and a positive 30-day return modestly favor SPY, but disclosed execution and customer risks cap conviction.
PGOutperform SPY52%Near-flat short-term action and a positive 30-day return modestly favor it over SPY, with low conviction.
SPYUnderperform SPY50%The supplied 30-day SPY signal is negative, reflecting a weak and uneven broad-market backdrop.

The debate

Bull, bear and risk cases: bull 0.70 · bear 0.87 · risk 0.98
bull case · conviction 0.70

The strongest evidence supports buying or holding a concentrated group of large-cap technology and semiconductor leaders, particularly AAPL, MSFT, TSM, NVDA, and META. Leadership is evident despite a nearly flat S&P 500 over the past 30 days: AAPL gained 5.9%, TSM 7.0%, and META 27.1%, while MSFT advanced 1.6%. Semiconductor momentum is being reinforced by a reported rally following Micron earnings and renewed AI optimism. Company-specific catalysts include Microsoft appearing on Wells Fargo’s Q4 Tactical Ideas List with cited upside potential, Nvidia’s record $150 billion stock-buyback authorization, and Alphabet’s continued Waymo expansion. The portfolio also has substantial cash, about 28.8%, providing flexibility to add to relative-strength winners rather than being forced to sell into volatility. This is a constructive but selective bull case: market breadth is not uniformly strong, volatility is rising, and several holdings are lagging.

  • Relative strength is favorable versus the S&P 500’s 0.26% 30-day return: AAPL is up 5.86%, TSM 7.05%, META 27.08%, and XOM 4.03%; MSFT is also positive at 1.55%.
  • META is the clearest momentum leader among portfolio holdings, with a 27.1% 30-day gain despite a recent 6.7% five-day pullback.
  • TSM combines strong 30-day momentum with positive near-term follow-through: +7.05% over 30 days and +1.12% over five days.
  • MSFT has the strongest recent acceleration among the core mega-cap names listed, gaining 3.01% in five days, and has a reported Wells Fargo tactical-list catalyst with cited 41% upside potential.
  • AAPL’s 5.86% 30-day gain materially outpaces the benchmark, indicating that negative AI-disruption commentary has not yet broken its price trend.
  • NVDA’s five-day momentum is positive at 1.69%, while Nvidia’s reported record $150 billion buyback provides a concrete shareholder-return catalyst; broader semiconductor stocks are also reported to be rallying after Micron earnings.
  • Alphabet offers a longer-term optionality catalyst through Waymo’s expansion to more domestic cities, international markets, and greater operating scale, while GOOGL remains modestly above the benchmark over 30 days.
  • The case is not risk-free: the VIX rose 12.6% over 30 days, the S&P 500 was nearly flat, and AMZN, JPM, MA, UNH, and several defensive holdings showed negative 30-day momentum. That argues for selective accumulation, not indiscriminate buying.
bear case · conviction 0.87

The strongest bear case is that several holdings are already losing money or exhibiting deteriorating 30-day momentum while the broader market has been essentially flat. Weakness is concentrated in healthcare, consumer staples, payments, financials, and large-cap technology, creating a risk that recent gains in other positions mask a widening set of laggards. The evidence is strongest for UNH, GOOGL, AMZN, JPM, MA, V, and JNJ; GLW also carries a broad set of disclosed execution, customer, regulatory, technology, and capital-spending risks despite its current gain.

  • UNH is the clearest deterioration signal: the position is down 9.95% from cost and has fallen 6.51% over 30 days.
  • GOOGL is down 3.68% from cost, while AMZN has declined 2.77% over 30 days, indicating weakness in two major technology holdings.
  • JPM, MA, and V show negative 30-day momentum of 6.60%, 6.80%, and 5.35%, respectively. This raises concern that financial and payment exposure is losing momentum even though JPM, MA, and V remain above cost.
  • JNJ has negative 30-day momentum of 0.39%, and its position is only supported by limited evidence of resilience rather than improving momentum.
  • PG, WMT, and XOM are already below cost by 1.99%, 3.99%, and 3.09%, respectively. Their declines broaden the weakness beyond growth-oriented holdings.
  • The S&P 500 gained only 0.26% over 30 days, while SPY's supplied momentum signal is negative 0.85%. Against that weak market backdrop, multiple holdings are materially underperforming on the supplied momentum measures.
  • GLW is up from cost, but the supplied risk memory identifies potential adverse outcomes involving customer losses or financial weakness, inability to pace capital spending to demand, margin-execution failures, technology change, product or component problems, litigation, regulation, tax developments, acquisitions, and key-personnel retention.
  • Several non-held candidates also show pronounced weakness, including HD at -12.79%, APLD at -11.19%, CORZ at -9.31%, ORCL at -9.64%, and AVGO at -5.31%. This argues against treating the broader opportunity set as uniformly healthy.
  • The portfolio has meaningful exposure to names with simultaneous price losses and/or negative momentum, so a continued deterioration in market breadth could leave the portfolio dependent on a smaller group of winners.
risk case · conviction 0.98

The primary capital risk is concentration: 67.1% of invested assets are held in just AAPL, JNJ, and AMZN, with AAPL alone representing 30.1% of invested assets. The portfolio also has a substantial Information Technology tilt at 48.0% of invested assets. Cash is a meaningful 28.9% of total portfolio value and provides liquidity, but the invested portion remains highly concentrated.

  • AAPL is the largest exposure at 30.1% of invested assets, making it the single biggest source of position-specific capital risk.
  • JNJ and AMZN add 19.1% and 17.9%, respectively; together with AAPL, the three largest positions comprise 67.1% of invested assets.
  • Information Technology is 48.0% of invested assets, led by AAPL, MSFT, NVDA, and GLW, creating a material sector tilt independent of directional views.
  • Health Care and Consumer Discretionary are the next-largest sectors at 19.2% and 17.9%, respectively, so the portfolio is concentrated across only a few major sector exposures.
  • Cash is 28.9% of the portfolio, which limits fully invested exposure but also means the risk-bearing capital is concentrated in relatively few positions.
  • The smallest holdings contribute little diversification relative to the dominant AAPL, JNJ, and AMZN positions.

Cash thesis

Cash is 28.9% of portfolio value, above the 25% threshold, and is intentional. It offsets the concentrated equity sleeve, preserves flexibility during elevated yield and volatility uncertainty, and avoids forcing purchases into a market where SPY is weak over the supplied 30-day measure and breadth is uneven. The reserve will be deployed only into confirmed relative strength or materially improved breadth, not to average down laggards.

Full commentary

Bear case response

Accept the concentration case: AAPL, JNJ and AMZN dominate risk-bearing capital, so a modest AAPL trim is appropriate despite its positive 30-day return. Accept that UNH is deteriorating, with a loss from cost and negative 30-day momentum; however, its very small position does not justify trading solely to create activity. Accept that GOOGL and AMZN have weaker recent momentum, but reject immediate sales because their declines are modest relative to the portfolio and there is not yet evidence of a decisive thesis break. Accept the warning on JPM, MA and V because all have negative 30-day momentum; their positions are small and will be held rather than churned. Accept that PG, WMT and XOM being below cost reflects broad unevenness, but their modest position sizes and limited momentum signals do not warrant reactive sales. Accept that JNJ's momentum is soft, but its positive return from cost and defensive role support holding rather than trimming simultaneously with AAPL. Accept GLW's disclosed operational, customer, regulatory and capital-spending risks; its positive return and flat recent momentum justify monitoring, not adding. Reject the implication that all technology strength should be sold: MSFT, TSM and NVDA have positive near-term momentum, but no additions will be made because the bear case on narrow leadership and elevated volatility remains valid.

Market summary

The broad market is range-bound: the S&P 500 is up only 0.26% over 30 days while SPY's supplied 30-day signal is negative and the VIX has risen. Treasury-yield volatility and uneven breadth argue against aggressive deployment. Selective AI and semiconductor leadership remains intact, notably MSFT, TSM and NVDA near term, but leadership is narrow and several financial, consumer, healthcare and internet holdings have negative 30-day momentum.

Portfolio assessment

The portfolio has appreciated to $1.04 million, but concentration remains the principal issue. AAPL, JNJ and AMZN account for a disproportionate share of invested capital, with AAPL alone worth about $222,252. The portfolio should retain core exposure to its strongest franchises but rebalance AAPL modestly rather than add to existing large positions. No new watchlist purchases are warranted until market breadth improves.

Risk assessment

AAPL, JNJ and AMZN remain the dominant position risks, while technology-linked holdings retain correlated sensitivity to AI-capex expectations, rates, regulation and momentum reversals. The risk is not insufficient exposure; it is concentrated exposure within the invested sleeve. UNH, GOOGL, AMZN, JPM, MA and V require monitoring because of weak 30-day momentum. The elevated cash allocation moderates total-portfolio drawdown but does not eliminate risk from the largest holdings. This follows the fund's prior risk lessons favoring incremental trims, low turnover and avoiding additions to weakening names.

← September 30, 2026 October 2, 2026 →