Glasshouse Fund
Decision journal

AI fund decision · August 21, 2026

Maintain a neutral, defensive posture: execute no buys or sells, retain cash above 25%, hold relative-strength core positions, and do not average down in weaker names. This follows the higher-priority memory constraint to avoid overtrading and manage persistent concentration risk.

Outlook: NEUTRALPortfolio $1,034,302Cash 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 (13 outperform, 21 underperform); 0 became trades. These are the fund's calibration record.

beat SPYlag SPYAAPL: underperform at 62%AMD: underperform at 70%AMZN: outperform at 64%APLD: underperform at 61%ASML: underperform at 57%AVGO: underperform at 63%BE: underperform at 72%CAT: underperform at 69%CORZ: underperform at 76%CRWV: underperform at 66%GLW: underperform at 65%GOOGL: underperform at 60%HD: underperform at 58%INTC: underperform at 73%IREN: outperform at 58%JNJ: outperform at 74%JPM: outperform at 59%MA: outperform at 70%META: underperform at 78%MSFT: outperform at 73%MU: outperform at 66%NVDA: outperform at 63%ORCL: outperform at 61%PG: underperform at 58%QQQ: underperform at 56%SNDK: outperform at 60%SPY: underperform at 50%TSLA: underperform at 66%TSM: outperform at 58%UNH: underperform at 72%V: outperform at 66%WMT: underperform at 77%XOM: outperform at 65%^VIX: underperform at 55%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
METAUnderperform SPY78%A -15.5% 30-day decline and -7.5% five-day decline signal severe relative weakness.
WMTUnderperform SPY77%Weak sales-growth news accompanies a -9.9% five-day and -9.1% 30-day decline.
CORZUnderperform SPY76%CORZ has sharply negative five-day and 30-day momentum, while its business remains high volatility.
JNJOutperform SPY74%Positive five-day and 30-day momentum provides a stronger and more balanced trend than SPY.
INTCUnderperform SPY73%INTC's -10.1% five-day and -3.1% 30-day returns indicate substantial relative weakness.
MSFTOutperform SPY73%MSFT's +22.4% 30-day gain is exceptional versus SPY, despite a recent pullback.
BEUnderperform SPY72%BE's -11.9% five-day and -5.8% 30-day performance show pronounced downside momentum.
UNHUnderperform SPY72%UNH's -9.7% 30-day and -4.2% five-day declines show sustained relative weakness.
AMDUnderperform SPY70%AMD's -8.7% five-day and -5.3% 30-day returns indicate persistent relative weakness.
MAOutperform SPY70%MA has positive returns over both periods and materially stronger 30-day momentum than SPY.
CATUnderperform SPY69%CAT is down over both periods, including -7.2% over 30 days, without a supplied offsetting catalyst.
CRWVUnderperform SPY66%Despite strong 30-day gains, a -14.7% five-day decline and debt/dilution concerns create unfavorable near-term risk.
MUOutperform SPY66%MU's +14.8% 30-day return and stable five-day result support continuing relative strength.
TSLAUnderperform SPY66%A small five-day gain does not reverse its -9.4% 30-day decline.
VOutperform SPY66%V is positive over both five and 30 days, supporting modest relative strength versus SPY.
GLWUnderperform SPY65%Negative recent momentum combines with disclosed demand, execution, customer, litigation, and regulatory risks.
XOMOutperform SPY65%XOM's +4.5% five-day and +13.5% 30-day momentum substantially exceed SPY.
AMZNOutperform SPY64%AMZN's +5.2% 30-day gain exceeds SPY despite a modest recent pullback.
AVGOUnderperform SPY63%Negative returns over both five and 30 days point to weakening semiconductor relative momentum.
NVDAOutperform SPY63%Positive 30-day momentum remains ahead of SPY, though earnings-related and short-term risks limit conviction.
AAPLUnderperform SPY62%A positive five-day bounce does not yet outweigh its -6.6% 30-day return versus positive SPY performance.
APLDUnderperform SPY61%Strong 30-day gains are offset by an -8.2% five-day decline and elevated AI-infrastructure volatility.
ORCLOutperform SPY61%ORCL's +12.4% 30-day return supports relative strength despite a -5.6% five-day pullback.
GOOGLUnderperform SPY60%Negative returns over both periods and modest unrealized loss suggest continued relative lagging.
SNDKOutperform SPY60%A +18.1% 30-day gain remains strong, although recent declines and adverse news make the call lower conviction.
JPMOutperform SPY59%Its +3.1% 30-day return exceeds SPY, although the recent five-day decline lowers conviction.
HDUnderperform SPY58%Negative five-day and 30-day momentum plus profitability-squeeze commentary temper the earnings setup.
IRENOutperform SPY58%The +26.7% 30-day gain and AI-pivot news support relative strength, though volatility is very high.
PGUnderperform SPY58%Negative returns over both periods imply weaker momentum than SPY.
TSMOutperform SPY58%TSM's +4.4% 30-day return exceeds SPY, though the recent five-day decline limits confidence.
ASMLUnderperform SPY57%Flat 30-day performance and a -5.1% five-day drop lag the broader market setup.
QQQUnderperform SPY56%QQQ's -2.8% five-day decline and only modest 30-day gain indicate weaker near-term breadth than SPY.
^VIXUnderperform SPY55%The VIX has risen over five days, but its lower 30-day level and moderate absolute reading suggest limited sustained upside versus equities.
SPYUnderperform SPY50%SPY is the comparison benchmark, so expected relative performance is effectively neutral; this forced directional call carries no informational edge.

The debate

Bull, bear and risk cases: bull 0.75 · bear 0.91 · risk 0.97
bull case · conviction 0.75

Buy/hold the portfolio with emphasis on the holdings demonstrating clear 30-day relative strength. Despite a choppier near-term tape, the S&P 500 is up only 1.9% over 30 days, while MSFT, NVDA, JNJ, MA, and AMZN have materially outperformed. MSFT is the standout momentum leader, up 22.4% over 30 days, with current commentary arguing the market is mispricing the stock and that its real price is cheaper than it appears. AAPL offers a higher-upside, recovery-oriented catalyst: its 5-day momentum has turned positive, news reports a fresh record and HSBC sees another 10% upside, while its Houston manufacturing-center opening adds a visible strategic development. The evidence supports maintaining exposure to quality leaders and selectively adding to the strongest relative-strength names, while recognizing that elevated volatility and several weaker holdings argue against indiscriminate buying.

  • MSFT is the strongest holding on momentum: up 22.4% over 30 days versus 1.9% for the S&P 500, despite a recent 5-day pullback; current news says Dan Ives believes the market is mispricing the shares, and another report argues the stock is cheaper than it looks.
  • NVDA is up 6.9% over 30 days, outperforming the benchmark and retaining positive medium-term momentum despite a 3.7% five-day decline; its upcoming earnings discussion is a potential catalyst, although the supplied evidence does not provide specific earnings details.
  • JNJ has delivered balanced momentum, rising 2.7% over five days and 5.7% over 30 days, materially outperforming the S&P 500 over the month; the recent news flow confirms active investor attention even though one report notes a larger one-day fall than the market.
  • MA is up 5.6% over 30 days and 0.8% over five days, while V is also positive over both periods at 2.2% and 0.4%, respectively; both payment holdings are showing relative strength rather than participating in the broader short-term weakness.
  • AMZN is up 5.2% over 30 days, nearly three times the benchmark's return, providing positive medium-term momentum even after a modest 1.0% five-day decline.
  • AAPL has lagged over 30 days, down 6.6%, but its 5-day return has turned positive at 1.8%. The reported fresh-record high, HSBC's stated 10% upside view, and the opening of its Houston manufacturing center provide identifiable catalysts for a momentum reversal.
  • The portfolio has meaningful winners relative to cost: based on the supplied average costs and current prices, MSFT, JNJ, MA, V, JPM, AMZN, and NVDA are above cost, giving the committee a reason to hold rather than sell strong positions solely because of short-term weakness.
  • The market backdrop is not risk-free: the VIX is up 12.4% over five days, and market news highlights high valuations, elevated yields, and a potentially weak global-stock week. This supports staged buying and concentration in the strongest relative-strength names rather than aggressive across-the-board additions.
bear case · conviction 0.91

The clearest bearish case is that several holdings and candidates are materially losing momentum while the S&P 500 gained 1.9% over the same 30-day period. AAPL, META, UNH, and GOOGL combine weak or negative performance with deteriorating momentum, and META's decline is especially severe. The candidate list is even more vulnerable, with sharp 30-day losses in CORZ, META, CAT, TSLA, WMT, BE, AMD, and UNH. This creates a meaningful risk of adding to relative laggards before momentum stabilizes. GLW also has both fading momentum and an extensive set of disclosed operational, demand, customer, technology, litigation, regulatory, and execution risks.

  • AAPL is down 6.6% over 30 days despite the S&P 500 gaining 1.9%, making it the strongest large current holding evidence of broad relative deterioration.
  • META is both in the red and down 15.5% over 30 days, a severe combination of realized weakness and negative momentum.
  • UNH is down 3.1% in the portfolio and down 9.7% over 30 days, indicating weakness both on the position and recent-trend measures.
  • GOOGL is down 2.1% in the portfolio and down 1.8% over 30 days, underperforming the benchmark while already showing an unrealized loss.
  • HD is not currently in the red, but its 30-day momentum is negative at -1.3%, so its modest gain versus cost is not supported by positive recent trend evidence.
  • The candidate set contains several pronounced momentum failures: CORZ -12.2%, CAT -7.2%, TSLA -9.4%, WMT -9.1%, BE -5.8%, and AMD -5.3% over 30 days.
  • GLW's negative 30-day momentum of -2.0% is accompanied by disclosed risks involving customer demand, capital-spending pace, margin improvement execution, technology change, product performance, loss of significant customers, litigation, regulation, and customer financing or receivables.
  • The evidence does not show broad momentum support for adding to the weaker names; instead, the most visible signals cluster around relative underperformance and declining trends.
risk case · conviction 0.97

The primary capital risk is concentration: 71.2% of invested assets is concentrated in AAPL, AMZN, and JNJ, while 70.8% is allocated to Information Technology and Consumer Discretionary. This leaves the portfolio materially exposed to adverse outcomes in a small number of holdings and sectors, independent of any directional view. Cash at 25.9% provides a meaningful capital buffer, but the invested portion remains highly concentrated.

  • AAPL, AMZN, and JNJ represent 27.2%, 24.8%, and 19.2% of invested assets, respectively; together they comprise 71.2%.
  • AAPL and AMZN alone account for 52.0% of invested assets, creating substantial single-name concentration in two holdings.
  • Information Technology is 40.1% and Consumer Discretionary is 30.7% of invested assets, for a combined 70.8% sector allocation.
  • The top three sectors. Information Technology, Consumer Discretionary, and Health Care, represent 90.2% of invested assets, leaving limited sector diversification.
  • Cash is 25.9% of the portfolio, which limits the amount of capital directly exposed to position and sector losses; however, it does not reduce concentration within the 74.1% invested allocation.
  • The remaining positions are individually small, with seven holdings each below 9% of invested assets, so they do little to offset the dominant exposures.

Cash thesis

Cash is $267.7K, or 25.9% of portfolio value. This elevated reserve is intentional: it offsets the portfolio's concentration in AAPL, AMZN, JNJ, and growth/consumer exposures, protects against a broader pullback amid valuation and yield concerns, and preserves capacity to deploy only after a clearer entry setup emerges. Holding cash is preferable to forcing purchases into a weak five-day tape or adding to lagging watchlist names.

Full commentary

Bear case response

Accept: AAPL is down 6.6% over 30 days versus positive benchmark performance, so its five-day rebound is insufficient evidence for another purchase; retain but do not add. Accept: META (-15.5% over 30 days), UNH (-9.7%), GOOGL (-1.8%), and HD (-1.3%) lack relative-strength confirmation; the current small META and UNH positions should be held rather than averaged down, and no expansion is warranted. Accept: candidates including CORZ, CAT, TSLA, WMT, BE, and AMD have materially negative momentum and should not be bought merely because they have declined. Accept: GLW's negative momentum and extensive disclosed operational and execution risks make it unsuitable for a new position. Partially reject the implication that all strong holdings should be reduced: MSFT, JNJ, MA, V, AMZN, and NVDA have outperformed over 30 days and remain above cost, supporting retention. However, their strength is not treated as a valuation case or a reason to add after recent short-term weakness. The concentration concern is fully accepted and is the principal reason for retaining cash and declining new purchases.

Market summary

The broad market remains constructive over 30 days, with SPY up 2.6% and the stated S&P 500 benchmark up 1.9%, but the near-term tape has weakened: SPY is down 1.8% over five days, QQQ is down 2.8%, and the VIX is up 12.4%. News highlights elevated valuations, steady-to-higher yields, and the prospect of the weakest global-equity week since mid-July. This favors retaining proven relative-strength holdings while avoiding new exposure to falling or highly volatile names.

Portfolio assessment

The portfolio has gained to $1.034M, with substantial unrealized gains in JNJ, MSFT, AMZN, V, MA, and JPM. However, AAPL ($208.9K), AMZN ($190.3K), and JNJ ($147.5K) remain dominant positions, and technology/consumer-discretionary exposure is substantial. Recent memory shows repeated AAPL buying and trimming within a short period; no further action is appropriate without clearer confirmation. The correct action is to hold existing core positions, make no incremental purchases, and avoid selling into a generally choppy market absent a thesis break.

Risk assessment

Concentration is the primary risk rather than a lack of holdings. AAPL, AMZN, and JNJ represent approximately 71% of invested assets, leaving performance dependent on a few names. AAPL, GOOGL, META, UNH, and HD have negative 30-day momentum, while AAPL and AMZN are also the largest equity exposures. The portfolio is additionally vulnerable to a correlated reversal in AI, cloud, and large-cap technology through AAPL, MSFT, NVDA, AMZN, and GOOGL. The cash reserve moderates total-portfolio drawdown risk but does not solve concentration within invested capital.

← August 20, 2026