Glasshouse Fund
Decision journal

AI fund decision · August 12, 2026

Maintain all current positions and retain $245,108 of cash. The market backdrop is constructive, but portfolio concentration and uneven breadth make new purchases unattractive. Hold strong leaders, do not average into AAPL, GOOGL, or UNH, and avoid weak watchlist cyclicals and semiconductors. Sources used include 5-day and 30-day return data, market and symbol news, benchmark data, and memory IDs risk_lesson:run_20260803T191113Z_b5f11a59:report_risk, risk_lesson:run_20260729T184341Z_54100baf:report_risk, trade:run_20260807T183247Z_07bc8571:AAPL:BUY, trade:run_20260803T191113Z_b5f11a59:NVDA:SELL, and report_summary:run_20260810T195022Z_25113609.

Outlook: NEUTRALPortfolio $1,031,369Cash 23.8%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: underperform at 66%AMD: underperform at 70%AMZN: outperform at 64%APLD: underperform at 56%ASML: outperform at 62%AVGO: outperform at 63%BE: underperform at 58%CAT: underperform at 72%CORZ: outperform at 57%CRWV: outperform at 61%GLW: underperform at 74%GOOGL: underperform at 67%HD: outperform at 56%INTC: underperform at 71%IREN: outperform at 55%JNJ: underperform at 56%JPM: outperform at 65%MA: outperform at 61%META: outperform at 54%MSFT: outperform at 71%MU: underperform at 69%NVDA: outperform at 60%ORCL: outperform at 55%PG: underperform at 54%QQQ: underperform at 53%SNDK: underperform at 53%SPY: outperform at 50%TSLA: underperform at 68%TSM: underperform at 59%UNH: underperform at 64%V: outperform at 58%WMT: underperform at 52%XOM: outperform at 67%^VIX: underperform at 63%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
GLWUnderperform SPY74%A 13.5% 30-day drop alongside disclosed customer, financing, margin, and execution risks argues for continued lagging performance.
CATUnderperform SPY72%The 10.9% 30-day decline and 3.2% five-day decline indicate pronounced cyclical weakness.
INTCUnderperform SPY71%A 20-billion-dollar equity offering and 11.4% 30-day decline create dilution and momentum concerns.
MSFTOutperform SPY71%A 31.4% 30-day gain and 3.4% five-day rise make MSFT the portfolio's strongest confirmed momentum leader, though extended valuation risk remains.
AMDUnderperform SPY70%Its 8.3% 30-day decline and negative 5-day return indicate weak semiconductor relative momentum.
MUUnderperform SPY69%Negative 5-day and 8.5% 30-day returns show semiconductor weakness relative to SPY.
TSLAUnderperform SPY68%A 15.5% 30-day decline outweighs the recent 3.5% bounce and signals weak intermediate relative momentum.
GOOGLUnderperform SPY67%Negative 5-day and 30-day momentum, plus a loss versus portfolio cost, outweigh long-term Waymo optionality over this horizon.
XOMOutperform SPY67%Strong 5.4% five-day and 13.2% 30-day momentum support continued relative strength.
AAPLUnderperform SPY66%Negative 5-day and 30-day momentum versus a positive SPY backdrop, with downgrade and memory-cost concerns, imply continued relative pressure.
JPMOutperform SPY65%Its 9.5% 30-day gain and positive five-day momentum demonstrate clear relative strength.
AMZNOutperform SPY64%An 11.8% 30-day gain materially exceeds SPY, though the flat recent five-day performance limits conviction.
UNHUnderperform SPY64%A 5.5% 30-day decline, negative five-day return, and medical-cost concerns indicate weakening relative performance.
AVGOOutperform SPY63%A 7.0% 30-day gain exceeds SPY despite a modest five-day consolidation.
^VIXUnderperform SPY63%The 9.6% 30-day decline and cooling-inflation risk-on backdrop favor lower volatility relative to equities.
ASMLOutperform SPY62%Positive 5-day and 30-day momentum supports relative strength versus SPY.
CRWVOutperform SPY61%Reported revenue doubling and accelerating AI-infrastructure demand support relative upside, tempered by missing return history.
MAOutperform SPY61%An 8.2% 30-day advance supports outperformance despite a recent 1.6% pullback.
NVDAOutperform SPY60%A 6.6% 30-day gain and AI-funding headline support relative strength, although the negative five-day return and sector breadth reduce confidence.
TSMUnderperform SPY59%The positive five-day move is insufficient to offset a negative 30-day return and broad semiconductor weakness.
BEUnderperform SPY58%Mixed valuation commentary and no supplied momentum history leave insufficient evidence for a durable relative-strength case.
VOutperform SPY58%Its 4.6% 30-day return exceeds SPY despite a modest five-day decline.
CORZOutperform SPY57%Data-center expansion headlines provide a catalyst, but missing return history makes this a low-confidence call.
APLDUnderperform SPY56%AI infrastructure interest is constructive, but absent usable 5-day or 30-day data and overvaluation commentary make risk-adjusted relative performance uncertain.
HDOutperform SPY56%Positive 5-day and 30-day returns support relative strength, but the CEO medical leave limits confidence.
JNJUnderperform SPY56%Positive five-day movement and pipeline news are offset by a negative 30-day return that trails SPY.
IRENOutperform SPY55%Positive AI-cloud-related news supports upside potential, but absent return history makes this highly speculative.
ORCLOutperform SPY55%Positive five-day and 30-day momentum supports a modest relative-strength view.
METAOutperform SPY54%The position remains profitable versus cost and participates in large-cap digital advertising and AI themes, but no return history limits confidence.
PGUnderperform SPY54%Slightly negative five-day and 30-day returns provide limited evidence of near-term relative leadership.
QQQUnderperform SPY53%QQQ's 1.0% 30-day return trails SPY's 2.83%, while its flat five-day action does not show renewed leadership.
SNDKUnderperform SPY53%No supplied momentum history or catalyst supports a confident case for beating SPY.
WMTUnderperform SPY52%Near-flat 30-day performance trails SPY, despite modestly positive recent momentum.
SPYOutperform SPY50%As the comparison benchmark, SPY is expected to track itself; the required directional label carries no active relative-view conviction.

The debate

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

Buy or hold with a bias toward the portfolio’s strongest relative-strength leaders. The evidence supports a constructive risk environment: the S&P 500 gained 2.83% over 30 days, the VIX fell 9.6%, and cooling CPI data is supporting higher equities and AI-related shares. Several portfolio holdings are materially outperforming that benchmark on a 30-day basis, while company-specific catalysts reinforce the strongest trends. The case is strongest for MSFT, AMZN, JPM, MA, NVDA, HD, and V, while weaker momentum in AAPL, GOOGL, and UNH argues for selectivity rather than indiscriminate buying.

  • MSFT is the clearest momentum leader: up 31.4% over 30 days and 3.4% over five days, far ahead of the S&P 500’s 2.83% 30-day return. Recent coverage says Microsoft cleared a major hurdle, and post-earnings coverage provides a potential catalyst.
  • AMZN is up 11.8% over 30 days, substantially outperforming the benchmark, with only a modest 0.1% five-day decline. The combination of strong intermediate momentum and a pullback-freeze rather than a major breakdown supports holding and adding selectively.
  • JPM (+9.5% over 30 days), MA (+8.2%), NVDA (+6.6%), HD (+5.4%), and V (+4.6%) all show positive 30-day relative strength versus the S&P 500. JPM also gained 0.8% in the latest five days, and HD gained 0.4%, indicating that several financial, payments, semiconductor, and home-improvement exposures remain supported.
  • NVDA has a tangible AI-related catalyst: news reports Nvidia confirmed a $500 billion AI funding deal, while the broader market is seeing a jump in AI-related shares. Its positive 30-day momentum supports maintaining exposure despite a mild 0.8% five-day pullback.
  • MA’s 8.2% 30-day advance is strong despite a 1.6% five-day dip, which can be viewed as consolidation within an established upward trend. V is also positive over 30 days, supporting the payments allocation as a relative-strength theme.
  • JNJ is a more defensive hold: its five-day momentum is positive at 0.9%, and coverage cites advancement of seltorexant in major depressive disorder. That provides a company-specific development even though its 30-day return is slightly negative.
  • HD’s CEO medical leave is a clear near-term uncertainty, but two executives were selected to oversee operations. The stock still shows positive 30-day momentum and a positive five-day move, suggesting the market has not produced evidence of a sustained technical breakdown.
  • GOOGL’s long-term evidence remains constructive despite weak recent price action: Alphabet’s filing describes Waymo expanding to more domestic cities, entering international markets, and scaling operations. That is a credible optionality catalyst, although the stock’s 30-day decline of 5.0% warrants patience rather than aggressive new capital.
  • The portfolio is not dependent on one narrow winner: exposure spans AI and software leadership, consumer internet, financials, payments, healthcare, and home improvement. The favorable market backdrop and multiple outperformers support holding overall exposure while reallocating incremental capital toward the strongest momentum names.
  • The bullish case is selective, not universal. AAPL is down 2.6% over 30 days, GOOGL is down 5.0%, and UNH is down 5.5%; HD also faces leadership news. These positions weaken the aggregate case and should be monitored rather than averaged up solely because they are already owned.
bear case · conviction 0.86

The portfolio has limited downside protection because several holdings and leading candidates are already losing momentum while the S&P 500 gained 2.8% over the same 30-day period. AAPL, GOOGL, JNJ, and UNH are weakening relative to the benchmark, with GOOGL and UNH already below cost. The candidate list shows broader deterioration across semiconductors, industrials, and consumer-facing names, suggesting that adding to cyclically or technologically exposed positions could compound weakness rather than diversify it.

  • AAPL is only modestly above its average cost at 300.77 versus 298.69, but its 30-day return is -2.6% while the S&P 500 returned +2.8%; the position therefore has little realized cushion against continued weakness.
  • GOOGL is already down 2.8% versus average cost and has declined 5.0% over 30 days, materially lagging the benchmark. This combination of negative absolute performance and deteriorating momentum is a clear downside signal.
  • UNH is slightly below cost and down 5.5% over 30 days, indicating that the position is weakening rather than stabilizing.
  • JNJ remains above cost, but its 30-day return is -1.4%, showing that its recent performance is deteriorating despite the unrealized gain.
  • The candidate set is broadly weak: TSLA is down 15.5%, GLW 13.5%, INTC 11.4%, CAT 10.9%, MU 8.5%, AMD 8.3%, and TSM 3.4% over 30 days. These declines argue against treating these names as attractive momentum additions.
  • GLW's disclosed risk factors specifically include customer losses, customer financing and receivables risk, inability to pace capital spending with demand, product or component problems, litigation, regulatory developments, and difficulty improving margins. Those risks are especially concerning alongside GLW's 13.5% 30-day decline.
  • The weakness is not confined to one company: it spans large technology holdings, healthcare, industrials, semiconductors, and several candidates. That breadth raises the risk that current price deterioration reflects a wider loss of support rather than an isolated issue.
  • The portfolio's strongest recent evidence is not sufficient to offset the bearish signals: several holdings have gains versus cost, but the available momentum data show meaningful recent deterioration in key positions and candidates.
risk case · conviction 0.98

The primary capital risk is concentration: AAPL and AMZN represent 53.7% of invested assets, while the top three positions. AAPL, AMZN, and JNJ, represent 71.7%. This is reinforced by a 72.3% invested-asset allocation to Information Technology and Consumer Discretionary. Cash at 23.8% of the portfolio provides a meaningful offset, but the invested portion remains heavily dependent on a small number of holdings and sectors.

  • AAPL is the largest position at 28.7% of invested assets, creating the single biggest name-specific concentration.
  • AMZN is 25.0% of invested assets; together with AAPL, the two positions comprise 53.7% of invested assets.
  • JNJ adds another 18.0%, bringing the top three positions to 71.7% of invested assets.
  • Information Technology is 41.5% and Consumer Discretionary is 30.8% of invested assets, for a combined sector exposure of 72.3%.
  • The portfolio has limited diversification outside these concentrations: Health Care is 18.2%, Communication Services 7.4%, and Financials 2.0%.
  • Cash is 23.8% of the total portfolio, reducing the amount exposed to equity concentration but also leaving the invested portfolio highly concentrated.
  • The evidence does not establish a directional performance conclusion versus the S&P 500; the risk assessment is based on exposure structure alone.

Cash thesis

Cash is 23.8%, below the mandatory 25% explanation threshold but still intentionally meaningful. It offsets concentrated equity exposure, preserves flexibility for a clearer pullback or confirmed catalyst, and avoids deploying capital into extended leaders while market breadth is mixed. The reserve also reduces the need to sell core holdings reactively if weaker positions deteriorate.

Full commentary

Bear case response

Accept: AAPL has little cushion above its $298.69 cost basis and trails SPY over 30 days; this rejects any case for adding now. Accept: GOOGL is below cost and down 5.0% over 30 days, while UNH is slightly below cost and down 5.5%; neither merits new capital without trend improvement. Accept: JNJ's -1.4% 30-day return is weaker than SPY despite its gain versus cost, so it is a defensive hold rather than a momentum leader. Accept: weakness across TSLA, GLW, INTC, MU, AMD, CAT, and TSM argues against rotating cash into cyclical or semiconductor laggards; GLW's disclosed execution, customer, financing, and margin risks reinforce that conclusion. Partially accept: MSFT, AMZN, NVDA, JPM, MA, HD, and V have stronger intermediate momentum, but that supports holding rather than chasing after advances. Reject the implication that broad weakness requires wholesale selling: SPY is positive, VIX is lower, CPI headlines are constructive, and the 23.8% cash reserve provides meaningful downside moderation. The final decision is therefore a defensive hold, not an aggressive buy or broad liquidation.

Market summary

Cooling CPI headlines, a 2.83% 30-day S&P 500 gain, and a 9.6% 30-day VIX decline support a constructive broad-market backdrop. Leadership is uneven: MSFT, AMZN, JPM, MA, NVDA, HD, V, and XOM show positive intermediate momentum, while AAPL, GOOGL, JNJ, UNH, and many semiconductor/industrial watchlist names lag SPY. The appropriate posture is to hold existing core exposure without chasing extended winners or averaging into laggards.

Portfolio assessment

Portfolio value is $1.031M with 23.8% cash. The main issue is not market-wide stress but concentration: AAPL, AMZN, and JNJ are the dominant exposures, while technology and consumer-discretionary risks remain highly correlated. MSFT and AMZN have strong momentum, but adding would increase an already concentrated growth posture. AAPL and GOOGL have weakened relative to SPY, yet no fresh fundamental evidence justifies a forced sale today. No trades are recommended to avoid reversing recent decisions or overtrading.

Risk assessment

AAPL represents roughly 21.8% of total portfolio value, AMZN roughly 19.0%, and JNJ roughly 13.7%; together they create substantial single-name risk. A reversal in AI, cloud, consumer spending, or large-cap technology could affect AAPL, AMZN, MSFT, NVDA, and GOOGL together. AAPL's weak 5-day and 30-day trend, GOOGL's negative momentum and loss versus cost, and UNH's negative trend require monitoring. Recent memory prioritizes disciplined incremental risk management and avoiding repeated trading in volatile technology exposures; the existing cash reserve partially mitigates but does not eliminate concentration risk.

← August 11, 2026