STOCK_MARKET
AI Analysis
Live Data

Claude Agents Beat SPY - Buy NOW & ICE, Sell APO & GD

Claude agents outperform SPY (+2.68%). Social sentiment ~53% supportive. Portfolio adds NOW & ICE, exits APO & GD. CPI print acts as a tailwind, helping multiples.

@theaiportfoliosposted on X

Breaking: Claude just bought two new stocks bc of the March CPI print Last week we gave Claude agents $50,000 to see how well they do at picking in stocks So far, they've already outperformed the SPY Today, they just bought two new stocks: 🟢 1. "BUY $NOW ServiceNow — New Position at 8% ServiceNow is the portfolio's first direct entry into enterprise workflow SaaS, and we're initiating because the market just handed us a gift wrapped in a category error. On April 8, Anthropic launched Claude Managed Agents, a cloud-hosted AI agent platform for enterprise. The market read this as "AI will replace SaaS" and sold NOW down 7.56% to $89.53, a 52-week low. Down 58% from its high of $211. What the selloff missed: ServiceNow is an Anthropic design partner. Claude is the default model powering the ServiceNow Build Agent platform. This company is not a victim of the AI agent buildout. It is infrastructure for it. The valuation: 24x forward P/E against a 5-year average of 50 to 55x. That's a 50%+ discount to its own history. Still guiding roughly 20% subscription growth, 32% operating margins, 36% FCF margins. This is a strong business at an irrationally cheap multiple. Street consensus PT: $185, which is +107% from our entry. The risk that matters: Q1 guide cut to below 19% subscription growth would break the thesis and push the stock to $75-80. At 8% weight, that full bear outcome costs roughly 1.6% of portfolio. The base case delivers +2.4%. Today's CPI makes this entry even better. Core came in cool at +0.2% MoM and +2.6% YoY, below consensus. That's a direct tailwind for long-duration SaaS multiples. Rate cut odds improve on this core read." 🟢 "2. BUY $ICE Intercontinental Exchange — New Position at 7% ICE is the NYSE parent, the largest energy futures exchange, plus mortgage tech and fixed income data. About 55% of revenue is recurring subscriptions. March 2026 set the all-time monthly volume record: 428.9M contracts, +88% average daily volume. Oil futures up 85%. Rates futures up 140%. The Iran/Hormuz crisis is printing money for ICE's transaction business. Here's what makes ICE special in this portfolio: it wins in BOTH macro scenarios. Hormuz stays impaired? Energy vol stays elevated, transaction fees keep printing. Hormuz reopens cleanly? Fed gets room to cut, rate futures volume surges, mortgage tech recovers on the refi wave. There is no macro scenario where ICE loses. Valuation: 22.98x forward vs peer CME Group at 25.41x. Similar business, more diversification (mortgage tech, data), yet trades at a discount. UBS raised their Q1 EPS estimate to $2.26 versus street consensus of $2.00. That's a 13% beat baked into one analyst's model. Apr 30 Q1 earnings is the catalyst. The risk: $20.3B debt from the Black Knight acquisition. If rates stay higher for longer, the interest burden bites. And a clean Hormuz reopen would normalize energy volumes back to baseline. The record March was crisis-driven, not structural. Today's CPI is perfect for ICE. Gasoline +21.2% MoM (largest since 1967) plus a split headline/core narrative = maximum confusion in the rates market = maximum trading volume. ICE gets paid on the vol, not the direction." 🔴 "3. SELL $APO Apollo Global — Full Exit Apollo entered the portfolio as a high-quality alternative asset manager at a reasonable valuation. The thesis broke when a securities class action arrived and escalated fast. The case is Feldman v. Apollo, filed in SDNY. CEO Marc Rowan is directly named in Epstein discovery documents. This is categorically different from the 2020 Leon Black matter, which Apollo survived by installing new leadership. Here, Rowan IS the leadership. May 1 is the lead plaintiff deadline, and the recruitment phase is peaking. As of yesterday, 10+ law firms are actively soliciting plaintiffs. Goldman cut PT $169 to $134 on Apr 7. Piper Sandler cut $165 to $146. Barclays cut $131 to $125. Three bulge bracket cuts in 48 hours. Stanford/Cornerstone settlement math: 3-8% of the $12B February decline = $360M to $960M settlement range. Claude deep research estimates roughly 55% probability this tail is real and currently unpriced in consensus EPS. Three-week expected value runs negative 4% to negative 15%, skewed to the downside." 🔴 "4. SELL $GD General Dynamics — Full Exit General Dynamics was a defense prime with a $118B backlog, Columbia-class submarines, and the G800 ramp. The thesis was defense spending supercycle plus best-in-class execution. Three broker downgrades in one week. Deutsche Bank cut to Hold on Apr 7. Jefferies cut to Hold at $380 the same week. Citi had already cut to Neutral at $380 on Apr 2. All three cite the identical thesis: Q1 consensus revenue growth of +4% is roughly 300 basis points too high. Then the insiders. CEO Novakovic plus two EVPs sold $18.1M of stock on March 11, six weeks before the Apr 22 earnings print. When three analysts say the quarter will miss and the C-suite is dumping shares, you listen. BNP Paribas raised their PT to $430 on the same day Deutsche downgraded. The bull case exists. But it requires a fifth consecutive earnings beat that three of the most active defense desks now explicitly model as a miss. Expected 12M return: +4.3% probability-weighted. Below our portfolio hurdle. Firm score 82, the weakest tier among our holdings. The structural defense story (NATO 5% GDP, Columbia subs, Gulf stream backlog) is not dead. It's just 2-3 quarters away from showing in the numbers. We can re-enter at a better price after the Apr 22 print if the thesis repairs." New updated portfolio: $VST | 10.3% $TMO | 8.9% $LLY | 8.1% $NOW | 7.6% $AVGO | 7.3% $CI | 7.1% $GLD | 7.1% $ICE | 6.8% $HALO | 6.2% $BAH | 6.0% $OKTA | 5.7% $DVN | 5.6% $MA | 4.9% $AU | 4.4% $MSFT | 4.1% Performance since inception: Claude: +2.68% SPY: -0.25% As a reminder, this is a public long term project to see how well Claude does We have 0 idea nor 0 expectation on how this will do, but we'll be sharing all updates here publicly and consistently no matter how good or bad Claude does See following tweet for information on how to invest alongside

View original tweet on X →

Community Sentiment Analysis

Real-time analysis of public opinion and engagement

Sentiment Distribution

66% Engaged
53% Positive
Positive
53%
Negative
13%
Neutral
34%

Key Takeaways

What the community is saying — both sides

Supporting

1

ServiceNow is being bought as the AI distribution layer

many replies argue $NOW is mispriced because Claude (the model) sits inside ServiceNow’s agent platform, so the company is the “landlord” of AI workflows rather than a competitor to models.

2

AI-managed portfolios beat human managers

commenters praise the system as cheaper, emotionless, and already outperforming the S&P, framing AI as replacing portfolio managers first, not workers generally.

3

Transparent, fully autonomous process

several replies emphasize the setup is public on Autopilot, the system runs the Russell 1000 universe, scores ~990 stocks, and constructs a 15‑position, liquid portfolio so followers can verify performance.

4

Tail risks can override fundamentals

the Apollo example is cited: legal/PR shocks (Epstein discovery, mass plaintiff recruiting) turned a valuation debate into a quantifiable tail risk and justified an exit.

5

Model choice matters

experiments comparing Grok, Gemini, ChatGPT and others show materially different returns, implying results depend on which LLM/agent architecture is used.

6

AI fixes active management’s core problems

supporters argue the bet is not against active management per se but against its cost and emotional flaws; removing those via automation could make active work.

7

Risk controls and sizing are deliberate

replies note limits (15 positions, max ~10% per name, liquid universe) and that worst‑case exposure is small (examples like 8% of portfolio), underscoring deliberate risk management.

Opposing

1

Contradiction or hedge:

Critics point out the oddity of Claude claiming to "kill" ServiceNow while buying $NOW, asking whether the purchase is a hedge or exposes a flawed narrative.

2

Bad timing / losing money:

Several replies note the purchase came as the stock was down (~8%), framing it as poor timing and a sign of weak decision-making.

3

Not stock-picking — just beta:

Multiple voices argue the strategy simply rides momentum and outperforms only in bull markets, essentially tracking market beta rather than selecting winners.

4

Risky technical play:

Traders warn Claude is "catching a falling knife" on $NOW — the name may be oversold across timeframes and should prove support before buying.

5

Opaque portfolio / paywall:

Users complain they can't view the claimed positions without paying the app subscription, raising questions about transparency and verifiability.

6

Accusations of fakery:

Some replies flatly call the portfolio or posts "fake", suggesting the account might be staged or misleading.

7

Index-heavy holdings:

Observers note the picks are all S&P 500 tickers, implying the model favors large-cap, mainstream names rather than unique or niche opportunities.

Top Reactions

Most popular replies, ranked by engagement

T

@theaiportfolios

Supporting

For transparency sake, here's the initial $50K it's managing The portfolio is listed on Autopilot where you can easily see the full performance & positions https://t.co/UzTbIfreSB Regardless, follow for more updates to come.

27
2
22.6K
E

@Eyuskant

Supporting

laude $50,000. Claude is already beating the S&P 500. While human fund managers are charging 2 and 20 to underperform. The most expensive thing on Wall Street has always been human ego. AI does not panic. Does not revenge trade. Does not check Twitter before buying. It just read

21
2
9.0K
0

@0xtechquity

Supporting

Claude is buying $NOW 🥹

13
1
3.2K
B

@BMCapitale

Opposing

Claude buying $NOW because Claude is killing $NOW business model

8
1
1.9K
S

@STF12016

Opposing

Claude is meant to kill service now. Is he hedging?

1
1
873
Q

@Quantrarian52

Opposing

Wait.. Claude bought this and it’s down 8%?

1
1
185

This article was AI-generated from real-time signals discovered by PureFeed.

PureFeed scans X/Twitter 24/7 and turns the noise into actionable intelligence. Create your own signals and get a personalized feed of what actually matters.

Report an Issue

Found something wrong with this article? Let us know and we'll look into it.