Source: YouTube URL: https://www.youtube.com/watch?v=aj248dbT-ys Date: 2025-10-08 Duration: 9:45
US stocks continue to hit new highs in 2025, defying historical September decline patterns with Q4 showing 90% probability of gains averaging 4.7%. The video analyzes the ongoing AI investment boom and introduces Cathay 00980T ETF—a balanced portfolio of 70% top-10 US tech stocks (weighted by market cap) and 30% US Treasury bonds (3-10 year maturity) with automatic monthly rebalancing. This strategy aims to capture AI-driven growth while managing volatility through bond allocation.
The US stock market has entered a euphoric phase where traditional valuation metrics have lost predictive power for short-term movements. September 2025’s defiance of historical weakness patterns (typically a down month) signals exceptional market strength. When rare events like this occur, Q4 historically delivers 90% win rates with 4.7% average returns.
Bubble Dynamics: The presenter acknowledges this is a bubble-formation process but emphasizes the unpredictability of timing. The key strategy isn’t predicting when bubbles burst, but having proper portfolio construction to capture gains while protecting against eventual corrections.
Trump administration’s “summit meetings” with tech giants (Apple, Microsoft, Meta) focused on securing commitments for massive AI infrastructure investments. This capital flows into: - Data center construction - NVIDIA GPU procurement - Networking equipment (Broadcom) - Cloud computing capacity expansion
AI Perpetual Motion Machine: The September 23 announcement of NVIDIA investing in OpenAI created a fascinating circular flow: 1. NVIDIA invests in OpenAI 2. OpenAI places large orders with Oracle 3. Oracle purchases more NVIDIA chips 4. All three companies see market cap expansion
While this loop appears sustainable as long as AI subscription demand continues growing, the risk emerges if consumer/enterprise adoption slows.
A critical insight: P/E ratios show zero correlation with 1-year forward returns (essentially random distribution on scatter plots). High valuations today don’t predict next year’s performance, though they may indicate lower 5-year annualized returns.
This data-driven finding contradicts common bearish arguments that “high P/E = imminent decline.” Short-term market direction remains unpredictable regardless of valuation levels.
The presenter advocates for defensive positioning despite market strength: - 3-10 year US Treasuries: Optimal duration balance - Not too volatile (20+ year bonds) - Not too low yield (1-3 year bonds) - Provides meaningful downside protection during recession/rate cuts - 70/30 Stock/Bond Split: Aggressive enough to capture growth, defensive enough to cushion declines
Rebalancing Challenge: Individual investors often lack discipline to rebalance when needed—either adding bonds at market peaks or buying stocks after crashes. Automated solutions solve this behavioral problem.
00980T (Cathay US TOP Stock-Bond Balanced ETF): - Allocation: 70% top-10 US tech (market-cap weighted) + 30% US Treasury (3-10Y) - Rebalancing: Monthly automatic - Distribution: Semi-annual dividends - Weighting Method: Market-cap weighted (higher weight to larger companies) - Best For: Investors wanting AI exposure with downside protection, willing to sacrifice some upside for reduced volatility
00757 (統一尖牙 FAANG ETF): - Allocation: 100% top-10 US tech stocks (equal-weighted ~10% each) - Distribution: Non-distributing (accumulating) - Weighting Method: Equal weight (reduces impact of any single stock) - Best For: Pure tech bulls comfortable with high volatility, no bond buffer
00662 (富邦納斯達克 Nasdaq ETF): - Allocation: 200+ Nasdaq stocks - Distribution: Non-distributing (accumulating) - Diversification: Highest diversification but diluted returns from mega-cap winners - Best For: Tech sector exposure with maximum diversification, lower single-stock risk
NVIDIA’s OpenAI investment announcement triggered AI stock rallies: - 00980T: +0.68% (despite 30% bond allocation) - 00662: +0.52% (200+ holdings diluted gains) - 00757: -0.13% (equal weighting reduced NVIDIA impact)
This demonstrates market-cap weighting’s advantage during concentrated rallies—00980T’s higher NVIDIA allocation drove outperformance despite its 30% defensive bond position.
The presenter emphasizes: 1. No Market Timing: Accept inability to predict bubble bursts 2. Systematic Rebalancing: Automated discipline beats discretionary decisions 3. Asset Allocation Over Stock Picking: Portfolio construction matters more than individual selections 4. Volatility Management: Bonds aren’t just for bears—they enable staying power during inevitable corrections
“九是韭菜的韭” (September is the month of ‘chives’—slang for retail investors getting harvested)
“利空也已經不再是利空，利多是更大的利多” (Negative news is no longer negative, positive news becomes even more positive)
“本益比跟未來一年股價漲幅幾乎是呈現隨機性的關係” (P/E ratios show essentially random correlation with next year’s stock performance)
“泡泡可以吹得很久也可以吹得很短，重點不是去預測泡沫什麼時候破裂，重點是泡沫破裂之後我們資金可不可以適當去做調整” (Bubbles can last long or short; the key isn’t predicting when they burst but whether we can adjust our capital appropriately after they do)
“00980T把成長與穩健做結合” (00980T combines growth with stability)
2025年美股持續創新高，打破歷史上9月回檔慣性，第四季上漲機率高達9成、平均報酬4.7%。影片分析AI投資熱潮持續，並介紹凱基00980T ETF——結合70%美國前十大科技股（市值加權）與30%美國公債（3-10年期），每月自動再平衡，讓投資人既能參與AI成長又能透過債券配置降低波動風險。
美股市場進入亢奮期，傳統估值指標對短期走勢失去預測能力。2025年9月打破歷史上常見的回檔慣性（傳統弱勢月份），這種罕見現象預示第四季強勁表現——歷史上出現此現象時，第四季上漲機率達9成、平均報酬4.7%。
泡沫動態：主講人承認這是吹泡泡過程，但強調時機點難以預測。關鍵策略不是預測泡沫何時破裂，而是透過適當的資產配置，既能獲取漲勢又能在回檔時保護資金。
川普政府與科技巨頭（蘋果、微軟、Meta）舉行的「鴻門宴」，要求各公司承諾對美國AI基礎建設的投資金額。這些資金流向： - 數據中心建設 - 輝達GPU採購 - 網路設備（博通） - 雲端運算擴容
AI永動機：9月23日輝達宣布投資OpenAI，形成有趣的資金循環： 1. 輝達投資OpenAI 2. OpenAI向甲骨文下大單 3. 甲骨文回頭向輝達採購更多晶片 4. 三家公司市值同步暴增
這個循環只要AI訂閱需求持續成長就能維持，但風險在於若消費者/企業應用需求下滑，永動機將會失效。
關鍵洞察：本益比與未來一年報酬呈現零相關（散佈圖呈現隨機分布）。今天的高估值無法預測明年表現，儘管可能暗示未來5年年化報酬較低。
這項數據發現推翻常見的空頭論點「高本益比=即將下跌」。不論估值高低，短期市場走向依然無法預測。
主講人在市場強勢時仍建議防禦性佈局： - 3-10年期美國公債：最佳存續期間平衡 - 不會太波動（20年期以上） - 收益率不會太低（1-3年期） - 經濟衰退/降息時提供有意義的下檔保護 - 股七債三配置：既能積極參與成長，又有足夠的防禦緩衝
再平衡挑戰：個人投資者往往缺乏紀律執行再平衡——可能不敢在高點加碼債券，也可能不敢在股災時買股票。自動化解決方案可克服這種行為偏誤。
00980T（凱基美國TOP股債平衡ETF）： - 配置：70%前十大美國科技股（市值加權）+ 30%美國公債（3-10年期） - 再平衡：每月自動執行 - 配息：每半年配息一次 - 加權方式：市值加權（大公司權重較高） - 適合對象：想要AI成長但希望降低波動，願意犧牲部分上檔換取下檔保護
00757（統一尖牙FAANG ETF）： - 配置：100%前十大美國科技股（等權重約各10%） - 配息：不配息（累積型） - 加權方式：等權重（降低單一股票影響） - 適合對象：純科技股多頭，可承受高波動，不需債券緩衝
00662（富邦納斯達克ETF）： - 配置：200多檔納斯達克成分股 - 配息：不配息（累積型） - 分散程度：最高分散但稀釋巨型股漲幅 - 適合對象：想要科技產業曝險但要最大分散，降低單一個股風險
輝達宣布投資OpenAI當天，觸發AI類股大漲： - 00980T：+0.68%（儘管有30%債券配置） - 00662：+0.52%（200多檔持股稀釋漲幅） - 00757：-0.13%（等權重降低輝達影響）
這展示市值加權的優勢——當市場熱點集中時，00980T的高輝達配置推動績效，甚至打敗100%股票配置的其他ETF。
主講人強調： 1. 不做市場擇時：接受無法預測泡沫破裂時機 2. 系統化再平衡：自動化紀律勝過主觀判斷 3. 資產配置勝過選股：投資組合建構比個股挑選重要 4. 波動管理：債券不只是空頭保護——它讓你在修正期間有餘裕繼續持有
“九是韭菜的韭”（9月是「韭菜月」——散戶被收割的俚語）
“利空也已經不再是利空，利多是更大的利多”（壞消息不再是壞消息，好消息變成更大的好消息）
“本益比跟未來一年股價漲幅幾乎是呈現隨機性的關係”（本益比與次年股價漲幅呈現隨機相關）
“泡泡可以吹得很久也可以吹得很短，重點不是去預測泡沫什麼時候破裂，重點是泡沫破裂之後我們資金可不可以適當去做調整”（泡沫可長可短，重點不是預測何時破裂，而是破裂後能否適當調整資金）
“00980T把成長與穩健做結合”（00980T結合成長與穩健）
#美股 #USStocks #AI投資 #AIInvesting #ETF #00980T #資產配置 #AssetAllocation #股債平衡 #StockBondBalance #輝達 #NVIDIA #科技股 #TechStocks #美國公債 #USTreasury #投資策略 #InvestmentStrategy #市值加權 #MarketCapWeighted #再平衡 #Rebalancing #川普 #Trump #AI基礎建設 #AIInfrastructure #本益比 #PERatio #泡沫 #Bubble #OpenAI #凱基ETF #CathayETF