Can AI Beat the S&P 500? 1-Year Update on My Tracking Portfolio

When I launched this AI-driven investment portfolio a year ago, the goal wasn’t to find a magic shortcut to wealth. It was an audit.

Every market cycle creates a new form of modern alchemy. In 2020, it was speculative meme stocks and zero-interest-rate euphoria. Today, it is the promise that artificial intelligence can out-analyze Wall Street, eliminate human error, and generate market-beating returns without the friction of traditional due diligence.

As someone who spends forty to sixty hours a week auditing financials and reconciling ledger data, I don't buy hype. I look at raw numbers, risk-adjusted returns, and operational reality. When you strip away the marketing promises of automated investing, you are left with a fundamental question: Can an algorithm construct a balanced, resilient portfolio capable of beating a standard low-cost index fund, or is it merely taking on higher volatility for similar gains?

Here is the full, transparent breakdown of how the AI-selected portfolio performed over the past 12 months, how it stacked up against the S&P 500 benchmark, and what this experiment reveals about human discipline versus algorithmic execution.

The Original Thesis & Setup

In September 2025, I prompted a large language model to construct a diversified, ten-stock equity portfolio designed for long-term compound growth. I instructed the system to avoid penny-stock speculation, look beyond obvious mega-cap momentum, and balance high-growth technology with defensive healthcare, energy, and retail.

The AI selected ten distinct holdings across four major sectors:

AI Portfolio Lineup:
├── Technology & Semiconductors: NVDA, TSM, MSFT, AAPL, FTNT
├── Healthcare & MedTech: LLY, ISRG
├── E-Commerce & Consumer Discretionary: SHOP, BOOT
└── Energy & Infrastructure: CCJ

At first glance, the algorithm created what appeared to be an institutional-grade growth basket. It balanced the world’s most dominant chip designer (NVIDIA) with its primary fabrication monopoly (Taiwan Semiconductor), layered in enterprise cybersecurity (Fortinet), added high-margin healthcare innovators (Eli Lilly and Intuitive Surgical), and diversified into nuclear energy (Cameco) and specialty retail (Boot Barn).

I locked in the purchase baselines, tracked cost bases and market valuations, and committed to letting the strategy run without emotional human intervention.

The Benchmark: AI Portfolio vs. The S&P 500

In public accounting, performance without a benchmark is meaningless. A portfolio that returns 15% in a vacuum looks great until you realize the passive market returned 18% with half the volatility. To evaluate this experiment rigorously, the portfolio's equal-weighted performance was measured directly against the SPDR S&P 500 ETF Trust (SPY) over the identical 12-month timeframe.

+-------------------------------------------------------------------------------+
|                       1-YEAR BENCHMARK COMPARISON AUDIT                       |
+--------------------------+---------------------+------------------------------+
| Metric                   | AI Growth Portfolio | S&P 500 Index (SPY Benchmark)|
+--------------------------+---------------------+------------------------------+
| 12-Month Total Return    | ~22.4%              | ~17.5%                       |
| Outperformance (Alpha)   | +4.9% vs. Benchmark | Baseline (0.0%)              |
| Maximum Drawdown         | -14.2%              | -8.6%                        |
| Concentration (Top 3)    | 30.0% (Equal Weight)| ~21.0% (Cap-Weighted Tech)   |
| Primary Return Drivers   | NVDA, TSM, CCJ, LLY | Mega-Cap Tech & Financials   |
+--------------------------+---------------------+------------------------------+

The Core Takeaway on Returns

The AI portfolio generated approximately +4.9% of alpha over the S&P 500 over the 12-month period, driven almost entirely by explosive runs in semiconductors (NVDA, TSM) and nuclear energy (CCJ).

However, that outperformance carried a real psychological cost: higher volatility and steeper pullbacks. During quarterly pullbacks, the AI portfolio suffered drawdowns nearly double the magnitude of the broad index. The index fund offered a far smoother ride for passive investors who lack the stomach to watch single-stock holdings swing double digits in a single week.

Asset Breakdown: The Winners, Compounders, and Laggards

+-------------------------------------------------------------------------------+
|                      AI PORTFOLIO ASSET PERFORMANCE AUDIT                     |
+----------------------+------------+-------------------------------------------+
| Ticker & Company     | Sector     | 1-Year Performance & Role in Portfolio    |
+----------------------+------------+-------------------------------------------+
| NVDA (NVIDIA)        | Tech/AI    | Market Leader: Massive AI datacenter lift |
| TSM (TSMC)           | Tech/Chips | High Conviction: Global foundry dominance |
| CCJ (Cameco)         | Energy     | Surprise Winner: Nuclear demand re-rating |
| LLY (Eli Lilly)      | Healthcare | Core Driver: GLP-1 & pipeline execution   |
| ISRG (Intuitive)     | Healthcare | Steady Anchor: Robotic surgery volume     |
| MSFT (Microsoft)     | Enterprise | Market Match: Enterprise Azure expansion  |
| AAPL (Apple)         | Consumer   | Low Volatility: High-margin ecosystem     |
| FTNT (Fortinet)      | Security   | Volatile: Multiple compression on tech pullback|
| SHOP (Shopify)       | E-Commerce | Macro Sensitive: Consumer spending shifts |
| BOOT (Boot Barn)     | Retail     | Laggard: Discretionary retail margin drag |
+----------------------+------------+-------------------------------------------+

1. The Alpha Generators (NVDA, TSM, CCJ, LLY)

The outperformance over the S&P 500 was driven by high-conviction secular tailwinds.

• Semiconductors (NVDA & TSM): The AI captured both compute architecture and physical fabrication monopolies.
• Uranium & Infrastructure (CCJ): Cameco was the algorithm's strongest contrarian pick, capitalizing on the massive power requirements of AI data infrastructure.
• Biopharma Leadership (LLY): Demand for metabolic treatments provided non-correlated institutional inflows.

2. The Baseline Compounders (MSFT, AAPL, ISRG)

Microsoft and Apple mirrored the broader index, keeping the portfolio anchored when smaller-cap growth compressed. Intuitive Surgical proved to be a reliable healthcare compounder with steady procedural expansion.

3. The Laggards (BOOT, SHOP, FTNT)

Specialty retail and mid-tier tech highlighted where algorithms struggle most with macro headwinds. Boot Barn and Shopify experienced friction under fluctuating consumer discretionary trends, pulling overall portfolio gains closer to the benchmark.

3 Critical Lessons from 12 Months of Algorithmic Investing

Lesson 1: AI Is a Momentum Aggregator, Not an Oracle

The greatest myth in automated finance is that AI "predicts" future price action.

Large language models are pattern-matching engines. When you ask an AI to construct an equity portfolio, it aggregates consensus institutional research, financial filings, and historical momentum. It selected companies with wide moats and high return on invested capital (ROIC).

Picking great businesses is only half the battle. Surviving the holding period without panic-selling at the bottom is the other half.

Lesson 2: Index Funds Win on Efficiency and Peace of Mind

Beating the S&P 500 by ~5% is rewarding, but managing a basket of ten individual equities requires constant tracking, rebalancing considerations, tax planning, and the emotional tolerance for volatility.

For 95% of investors, a low-cost S&P 500 or total market index fund remains the superior vehicle: zero management friction, automated dividend reinvestment, and maximum tax efficiency.

Lesson 3: The Psychology of Execution Belongs to Humans

An algorithm can output a portfolio in two seconds. It cannot execute monthly dollar-cost averaging, maintain an emergency fund, or resist the urge to tinker during market corrections.

In Adlerian psychology, the Separation of Tasks defines long-term success:

• Outside your control: Market cycles, interest rates, macroeconomic headlines, and quarterly earnings surprises.
• Within your control: Your savings rate, your asset allocation, keeping your expenses lean, and your discipline to hold through drawdowns.

The Decision Blueprint: How to Use AI in Your Finances

Step 1: Ideation & Sector Screening (AI)
└── Use AI to summarize 10-K filings, calculate ROIC, and identify secular themes.

Step 2: Balance Sheet & Valuation Audit (Human)
└── Verify debt-to-equity, free cash flow yields, and price-to-earnings multiples.

Step 3: Core-Satellite Portfolio Structure (Execution)
└── Keep 80–90% of your capital in passive index funds (S&P 500 / Total Market) 
    and allocate 10–20% max to curated high-conviction satellite picks.

Step 4: The 12-Month Holding Rule (Discipline)
└── Lock in positions and eliminate emotional rebalancing for at least one full year.

Final Verdict: Can AI Replace Due Diligence?

AI is an exceptional research assistant and a terrible financial planner.

It can analyze balance sheets in seconds and highlight secular trends, but it will never understand your financial timeline, your mortgage, your career risk, or your emotional relationship with money.

Treat AI like a junior associate: let it run the queries, structure the data, and build the initial models. But the final sign-off—and the discipline to hold the line—belongs entirely to you.