AITrader

Performance

Strategy model

AIT-1

AIT-1 Daneel

active
Top performing

Inception: Feb 17, 2026 · 7 weekly runs

Model details

Weekly Top-20 Nasdaq-100 portfolio: stocks ranked by AI, equal weight, rebalanced every week, with trading costs included.

Outperformance

vs Nasdaq-100 (cap-weight)

Outperformance

vs S&P 500 (cap-weight)

Beta (β)

Signal — score vs next-week return (latest week)

-0.0061

Positive means higher-rated names tended to outperform next week.

Signal strength details

Selected portfolio

Configuration for the selected portfolio

Key metrics

Performance Overview

BalancedTop 20 · Weekly · Equal

Simulated growth of $10,000 from inception, net of trading costs, versus benchmarks.

Showing the model's published weekly track until your selected portfolio finishes loading.

Metrics at-a-glance

CAGR

-48.1%

tap to explain

CAGR

Annualized compound growth rate. If the strategy grew at this exact pace every calendar year since inception, this is the annual return you would have seen.

tap to flip back

Total return

-7.1%

tap to explain

Total return

How much the $10,000 starting capital has grown in total since inception. This is the raw cumulative gain, before any annualization.

tap to flip back

Max drawdown

-7.9%

tap to explain

Max drawdown

The worst peak-to-trough decline since inception. If you had invested at the peak and sold at the worst point, this is how much you would have lost. Closer to zero is better.

tap to flip back

Sharpe ratio

-2.47

tap to explain

Sharpe ratio

Return per unit of risk. It divides the strategy's average return by how much the returns fluctuate week to week. Above 1.0 is generally considered good for a stock strategy. Higher is better.

tap to flip back

% weeks outperforming Nasdaq-100

+17%

tap to explain

% weeks outperforming Nasdaq-100

Share of rebalance-to-rebalance weeks where the portfolio return beat the Nasdaq-100 cap-weighted index. Above 50% means it wins more weeks than it loses.

tap to flip back

% weeks outperforming S&P 500

+33%

tap to explain

% weeks outperforming S&P 500

Share of rebalance-to-rebalance weeks where the portfolio return beat the S&P 500 cap-weighted index. Above 50% means it wins more weeks than it loses.

tap to flip back

What you are looking at

  • We pick the top 20 stocks every week from the Nasdaq-100, ranked by AI score (selected portfolio).
  • Each stock gets equal weight — no outsized bets on single names.
  • Model batch day: Monday (weekly rebalance aligns with each batch).
  • We subtract realistic trading costs so the chart reflects what you would actually keep.
  • No retroactive edits. Once a week closes, the results are locked.

Starting capital: $10,000 simulated. Disclaimer

Top rated stocks

Returns

Total return

-7.1%

tap to explain

Total return

How much the $10,000 starting capital has grown over the full period since inception.

tap to flip back

CAGR

-48.1%

tap to explain

CAGR

Annualized compound growth rate — what the portfolio's growth would look like if it grew at this pace every year.

tap to flip back

Compared to benchmarks

All returns measured from Tue, Feb 17 to Mon, Mar 30.

Strategy / BenchmarkTotal returnCAGRMax drawdown
AIT-1 Daneel
-3.4% vs Nasdaq-100 (cap-weighted, cumulative)
-7.1%-48.1%-7.9%
Nasdaq-100 (cap-weighted)-3.7%-28.2%-5.8%
Nasdaq-100 (equal-weighted)-1.1%-9.4%-4.6%
S&P 500 (cap-weighted)-4.9%-36.2%-6.4%

Cumulative returns

Total percentage return from inception.

Weekly returns

Week-over-week percentage change.

CAGR over time

Annualized growth from inception using a $10,000 starting value.

Risk

Max drawdown

-7.9%

tap to explain

Max drawdown

The largest peak-to-trough decline in portfolio value. A drawdown of -20% means the portfolio fell 20% from its peak before recovering. Closer to 0% is better.

tap to flip back

Sharpe ratio

-2.47

tap to explain

Sharpe ratio

Return per unit of risk. Average weekly return divided by the standard deviation of weekly returns, then annualized. Above 1.0 is generally considered good for a stock strategy.

tap to flip back

Drawdown over time

Drawdown from rolling peak for each series. Tap chips to show or hide lines.

Deeper troughs = larger losses from peak. 0% means at the all-time high for that window.

Consistency

% weeks outperforming Nasdaq-100 (cap-weighted)

+17%

tap to explain

% weeks outperforming Nasdaq-100 (cap-weighted)

Share of weeks where the portfolio beat the Nasdaq-100 cap-weighted benchmark. Above 50% means it wins more weeks than it loses.

tap to flip back

% weeks outperforming S&P 500 (cap-weighted)

+33%

tap to explain

% weeks outperforming S&P 500 (cap-weighted)

Share of weeks where the portfolio beat the S&P 500 cap-weighted benchmark. Above 50% means it wins more weeks than it loses.

tap to flip back

Cumulative outperformance

How much AIT-1 Daneel is ahead of (or behind) each benchmark over time. Above zero = AI is winning.

Research validation

Beyond portfolio returns, we track whether the AI scores actually predict which stocks will outperform across all 100 Nasdaq-100 stocks, not just our top picks. This layer is tied to the strategy model (AI ratings engine), not the portfolio.

Quintile analysis

Stocks split into 5 equal groups by AI score. Q1 = lowest rated, Q5 = highest rated. If the model has real signal, Q5 should consistently outperform Q1.

Q5 outperformed Q1 in 2 of 6 weeks (33%)

Above 50% means top-rated stocks outperform bottom-rated stocks more often than not.

Q1

-1.83%

21 stocks

Q2

-1.55%

20 stocks

Q3

-1.75%

20 stocks

Q4

-2.56%

20 stocks

Q5

-5.20%

20 stocks

Q5 outperformed Q1 by -3.37% that week. A positive spread means higher-rated stocks outperformed lower-rated ones.

Signal strength

Does the AI score actually predict which stocks will do better next week?

Quick read: Beta tells you if higher AI scores lead to higher next-week returns, R² tells you how strong that relationship is, and Alpha is weekly market backdrop (not AI skill).

Beta (β) — the signal

Weak

-0.0061

Extra next-week return per +1 on the AI score. Positive means the model is working — higher-rated stocks outperform lower-rated ones.

Good: > 0. Strong: > 0.002.

R² — fit quality

Good

0.0509

AI score explains about 5.1% of cross-stock next-week return differences.

Meaningful: 0.01–0.05. Exceptional: > 0.05.

Alpha (α) — market backdrop

Context

-0.0270

Predicted return at AI score = 0. This mostly reflects weekly market direction.

Down-market backdrop of about 2.7%. Alpha is context, not AI skill.

Measured on Mon, Mar 23 · n=101 stocks

Full calculation details

The scientific basis: Peer-reviewed research (Ko & Lee; Pelster & Val, Finance Research Letters) shows AI ratings correlate with future stock returns. We test this hypothesis live. See how this model works

Reality checks

Includes trading costs

Each time we rebalance, we deduct 15 basis points (0.15%) per unit of portfolio turnover. For example, if 30% of the portfolio changes in a given week, the cost is 0.30 × 0.15% = 0.045% deducted from that week's return. This models real-world trading friction.

No retroactive edits

Once a week closes, the data is locked. We do not revise history when the model is updated. Each strategy model version is tracked separately.

Rules-based system

Every decision is deterministic. Same inputs produce the same outputs. No human discretion, no cherry-picked dates, no post-hoc adjustments.

Returns shown are pre-tax. Your actual returns will depend on your tax situation and jurisdiction. Tax treatment of investment gains varies by country and individual circumstances.

Want to understand how this model works?

See the full methodology, AI model configuration, prompt design, and scientific grounding.

Model details

Not investment advice. Past performance does not guarantee future results. Full disclaimer