Glasshouse Fund
Decision journal

AI fund decision · August 17, 2026

Maintain all current holdings and the 26.1% cash reserve. The portfolio should preserve exposure to MSFT, NVDA, AMZN, JPM, and payment leaders while avoiding additions to concentrated or lagging positions. No new buys or sells are justified today because the evidence is mixed and recent AAPL activity argues for trading discipline.

Outlook: NEUTRALPortfolio $1,026,706Cash 26.1%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 (10 outperform, 24 underperform); 0 became trades. These are the fund's calibration record.

beat SPYlag SPYAAPL: underperform at 66%AMD: underperform at 56%AMZN: outperform at 59%APLD: underperform at 60%ASML: outperform at 63%AVGO: underperform at 57%BE: underperform at 55%CAT: underperform at 67%CORZ: underperform at 64%CRWV: underperform at 58%GLW: underperform at 68%GOOGL: underperform at 63%HD: underperform at 55%INTC: underperform at 54%IREN: underperform at 57%JNJ: underperform at 56%JPM: outperform at 67%MA: outperform at 64%META: underperform at 70%MSFT: outperform at 72%MU: outperform at 61%NVDA: outperform at 69%ORCL: outperform at 64%PG: underperform at 59%QQQ: underperform at 54%SNDK: underperform at 56%SPY: outperform at 50%TSLA: underperform at 68%TSM: underperform at 53%UNH: underperform at 66%V: underperform at 53%WMT: underperform at 55%XOM: outperform at 58%^VIX: underperform at 70%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
MSFTOutperform SPY72%Azure revenue momentum and a 26.7% 30-day gain support relative strength despite a short-term pullback.
METAUnderperform SPY70%A 10.2% 30-day decline and slightly negative five-day performance indicate pronounced relative weakness.
^VIXUnderperform SPY70%The VIX is down over both periods, and a constructive equity backdrop favors continued volatility compression versus SPY.
NVDAOutperform SPY69%Positive returns over both periods and OpenAI ecosystem news support AI-infrastructure leadership.
GLWUnderperform SPY68%The 9.3% 30-day decline and documented customer, capex, margin, regulatory, and execution risks outweigh the recent bounce.
TSLAUnderperform SPY68%A positive five-day move does not overcome a 13.3% 30-day decline and continuing execution uncertainty.
CATUnderperform SPY67%Its 7.9% 30-day decline materially trails SPY despite a modest recent rebound.
JPMOutperform SPY67%Positive five-day momentum and an 8.5% 30-day gain show clear relative strength beyond technology.
AAPLUnderperform SPY66%Negative 5-day and 30-day momentum, plus component-cost concerns, leave AAPL behind SPY absent a new catalyst.
UNHUnderperform SPY66%Negative five-day and 30-day momentum, with the shares below cost, indicate relative weakness.
CORZUnderperform SPY64%A 3.8% five-day gain does not repair the 11.1% 30-day decline and high-risk profile.
MAOutperform SPY64%Positive returns over both periods and a 5.9% 30-day gain modestly exceed SPY.
ORCLOutperform SPY64%A 14.4% 30-day gain and stable five-day action indicate sustained relative strength.
ASMLOutperform SPY63%Positive 5-day and 30-day semiconductor-equipment momentum supports relative performance.
GOOGLUnderperform SPY63%Negative 5-day and 30-day returns versus a rising SPY indicate weak relative momentum despite supportive ownership news.
MUOutperform SPY61%Tightening memory-pricing news and strong five-day momentum support continued relative strength, albeit after a sharp advance.
APLDUnderperform SPY60%Strong short-term and monthly momentum is tempered by commentary that the stock appears fully valued after a large run.
AMZNOutperform SPY59%AWS AI-demand support and a positive 30-day return favor relative strength, though the 5.6% five-day drop reduces conviction.
PGUnderperform SPY59%Negative returns over both measured periods are likely to lag SPY in the current market regime.
CRWVUnderperform SPY58%Exceptional momentum is offset by concerns that the AI-spending-dependent model may be fragile.
XOMOutperform SPY58%Oil-related geopolitical risk and a 10.8% 30-day gain support relative strength.
AVGOUnderperform SPY57%A 6.9% five-day decline despite positive monthly performance signals weakening near-term momentum.
IRENUnderperform SPY57%Strong AI-related momentum is counterbalanced by financing concerns and a volatile, high-beta profile.
AMDUnderperform SPY56%The strong 5-day rebound is offset by a negative 30-day return and no supplied fresh catalyst.
JNJUnderperform SPY56%Stable defensive performance is positive, but its 1.0% 30-day gain is likely to lag SPY in a constructive risk backdrop.
SNDKUnderperform SPY56%The 32.6% five-day surge is highly extended while the 30-day return remains slightly negative.
BEUnderperform SPY55%The sharp five-day rebound is not enough to offset a negative 30-day return and elevated volatility.
HDUnderperform SPY55%Upcoming earnings and a price-target increase help, but negative five-day momentum and only marginal monthly gains trail SPY.
WMTUnderperform SPY55%Recent strength is positive, but its 0.4% 30-day gain materially trails SPY.
INTCUnderperform SPY54%The five-day recovery is constructive, but the negative 30-day return remains below SPY's gain.
QQQUnderperform SPY54%QQQ has positive momentum but its 2.7% 30-day gain trails SPY's 4.6% gain.
TSMUnderperform SPY53%Positive short-term momentum is constructive, but its 1.1% 30-day return lags SPY.
VUnderperform SPY53%Positive momentum supports stability, but its 2.0% 30-day gain is below SPY's 4.6% gain.
SPYOutperform SPY50%SPY is the comparison benchmark; confidence is neutral because its expected relative result is inherently market-matching.

The debate

Bull, bear and risk cases: bull 0.76 · bear 0.87 · risk 0.98
bull case · conviction 0.76

The strongest buy-and-hold case is to maintain exposure to the portfolio's leading technology, semiconductor, consumer, and payments positions while selectively adding to demonstrated relative-strength winners. The market backdrop remains constructive: the S&P 500 gained 4.6% over 30 days, QQQ gained 2.7%, and the VIX declined 17.0%, indicating improving risk appetite despite current geopolitical and AI-valuation concerns. Within the portfolio, MSFT, NVDA, AMZN, JPM, and MA are outperforming the S&P 500 on a 30-day basis, with identifiable business catalysts supporting continued momentum. Conviction is tempered because AAPL, GOOGL, HD, UNH, and META are lagging and market news includes warnings of a possible technology correction.

  • MSFT is the clearest relative-strength leader: its 30-day momentum is +26.7% versus +4.6% for the S&P 500. News reports that Azure annual revenue surpassed $100 billion and that the stock jumped 18% in a week, providing a fundamental catalyst behind the move. The recent five-day pullback of 2.1% may represent consolidation rather than a broken trend.
  • NVDA has strong, broad momentum, rising 10.6% over 30 days and 3.5% over five days, substantially ahead of the S&P 500. The reported OpenAI tenant deal is an additional AI-demand catalyst and reinforces the strategic importance of Nvidia's infrastructure ecosystem.
  • AMZN is up 6.2% over 30 days, outperforming the S&P 500 despite a recent five-day decline of 5.6%. That combination supports a buy-and-hold argument if the pullback is viewed as a reset within a still-positive monthly trend rather than evidence of a reversal.
  • JPM is delivering both positive short-term and medium-term momentum: +0.8% over five days and +8.5% over 30 days. Its 30-day performance is nearly twice the S&P 500's, making it a relative-strength diversifier beyond the technology complex.
  • MA is also outperforming, up 5.9% over 30 days and 1.1% over five days. V is positive over both periods as well, at +2.0% over 30 days and +1.0% over five days, supporting continued exposure to payments-sector momentum.
  • HD has a near-term catalyst in upcoming earnings, and Stifel raised its price target citing platform strength. Although HD is down 3.4% over five days, its 30-day return remains slightly positive, making weakness potentially actionable if earnings and platform execution validate the constructive view.
  • The portfolio already has meaningful unrealized gains in several leaders based on current prices versus average costs, including MSFT, NVDA, JPM, V, MA, and AMZN. Holding these winners preserves exposure to proven portfolio-level strength rather than selling solely because of short-term volatility.
  • The case is not uniform: AAPL is down 3.5% over 30 days, GOOGL is down 1.9%, META is down 10.2%, and UNH is down 6.4%. AAPL also faces a reported memory-chip-related product-price increase, while market commentary warns of a possible technology correction. These are reasons to be selective, not reasons to abandon the stronger holdings.
bear case · conviction 0.87

The clearest risk is broad-based momentum deterioration relative to a strong benchmark: the S&P 500 gained 4.6% over 30 days, while several holdings and candidates fell materially. The portfolio is especially exposed to names with either negative absolute returns or sharply weakening momentum, making further downside plausible if the relative weakness persists.

  • AAPL is only modestly above its average cost while showing a 30-day decline of 3.5%, a poor combination of limited cushion and deteriorating momentum.
  • META is approximately flat versus average cost but has fallen 10.2% over 30 days, indicating a sharp reversal and meaningful risk of further capital loss.
  • UNH is down 2.6% versus average cost and down 6.4% over 30 days, showing both realized weakness and accelerating negative momentum.
  • GOOGL is down 2.6% versus average cost and down 1.9% over 30 days, while the benchmark gained 4.6%; this is clear relative underperformance without a positive momentum offset.
  • The candidate list contains more severe downside signals: TSLA is down 13.3% over 30 days, CORZ 11.1%, GLW 9.3%, and CAT 7.9%. These are not isolated mild pullbacks but substantial short-term deterioration.
  • AMD, BE, INTC, PG, and SNDK are also posting negative 30-day returns, suggesting weakness spans multiple sectors rather than being confined to one industry.
  • GLW carries an especially broad risk set in the supplied fund memory, including customer losses or financing problems, inability to pace capital spending with demand, margin-execution risk, technology change, litigation, product-performance issues, regulatory developments, and tax-related risks. Its 9.3% 30-day decline makes those risks more urgent.
  • The portfolio's positive holdings do not eliminate the issue: several names with gains versus cost can still be vulnerable if momentum is fading, while the evidence provides no corresponding positive momentum signal for the weaker holdings.
risk case · conviction 0.98

The largest capital risk is concentration: 70.9% of invested assets are in AAPL, AMZN, and JNJ, while 90.1% are concentrated in Information Technology, Consumer Discretionary, and Health Care. Cash at 26.1% provides a meaningful buffer, but does not eliminate the dependence on a small number of holdings and sectors.

  • AAPL is the largest position at 27.0% of invested assets, followed by AMZN at 25.1% and JNJ at 18.8%.
  • The top three positions represent 70.9% of invested assets, creating material single-name concentration.
  • Information Technology, Consumer Discretionary, and Health Care comprise 90.1% of invested assets; Financials are only 2.1%, limiting sector diversification.
  • AAPL and AMZN together account for 52.1% of invested assets, making them the most direct sources of capital concentration risk.
  • Cash is 26.1% of the total portfolio, so approximately 73.9% is invested; this reduces overall exposure but leaves the invested portion highly concentrated.

Cash thesis

Cash at 26.1% is intentional. It offsets concentrated exposure to AAPL, AMZN, JNJ, and AI-linked equities; preserves optionality for a more favorable entry after pullbacks; and protects against a technology-led correction, geopolitical escalation, or rising-yield shock. Although cash creates opportunity cost in a rising market, current evidence does not justify deploying it broadly into names that are either extended or lagging SPY.

Full commentary

Bear case response

Accept: AAPL's 3.5% 30-day decline, limited gain versus cost, and large position size warrant caution; the recent August 12 trim was appropriate, and no add is justified. Accept: META's 10.2% 30-day decline and near-flat gain versus cost indicate meaningful momentum risk, although its small portfolio weight limits immediate portfolio damage. Accept: UNH and GOOGL are below cost and lag SPY over 30 days; both should be held rather than averaged down absent a new catalyst. Accept: weak 30-day performance in TSLA, CORZ, GLW, CAT, and several other candidates argues against initiating thematic watchlist positions. Accept: GLW's documented operating, customer, capital-spending, regulatory, and execution risks reinforce avoidance despite its recent five-day rebound. Partially reject a broadly bearish conclusion: MSFT, NVDA, AMZN, JPM, MA, and V retain positive relative-strength evidence, so wholesale selling would sacrifice credible leadership. However, accept that MSFT's sharp advance and AMZN's 5.6% five-day decline elevate reversal risk, which is why the portfolio will hold rather than add. The bear's concentration argument is fully accepted; cash and no new purchases are the appropriate response.

Market summary

The broad backdrop is constructive but selective: SPY is up 4.6% over 30 days, QQQ is up 2.7%, and the VIX has declined 17.0%, yet today's news flow highlights AI-valuation concerns, rising yields, and Iran-related geopolitical risk. Leadership is narrow. MSFT, NVDA, JPM, MA, and AMZN have favorable 30-day relative performance, while AAPL, GOOGL, META, and UNH have lagged materially.

Portfolio assessment

The portfolio is defensively positioned at 26.1% cash but remains concentrated in AAPL, AMZN, and JNJ, which represent about 52.4% of total portfolio value and 70.9% of invested assets. No transaction is warranted today: recent AAPL trades included an August 7 purchase followed by an August 12 trim, so further action without a new catalyst would risk repeating reactive trading. Retain profitable leaders, monitor the weaker holdings, and avoid adding to already-concentrated technology and consumer exposures.

Risk assessment

Primary risk is concentration rather than immediate market-wide stress. AAPL, AMZN, and JNJ dominate invested capital, while technology, consumer discretionary, and health care account for most equity exposure. A reversal in AI/cloud spending or consumer demand could affect AAPL, AMZN, MSFT, NVDA, and GOOGL simultaneously. AAPL, META, GOOGL, and UNH have negative 30-day momentum relative to a strong SPY benchmark. Retaining cash and avoiding fresh concentration are consistent with prior risk lessons.

← August 14, 2026