Thesis challenged · not confirmed

Find Undervalued Stocks Faster With AI Screening + DCF Valuation

Rank S&P 500 companies by quality, value, growth, and risk. Build sector-aware DCF scenarios, compare intrinsic value, and let deterministic AI agents challenge your thesis before you commit.

See Plans & Pricing
No card required. Free to analyze.
Start Free Analysis
Divergia / Quantitative Engine
AssetRankQualitySignal (Agents)
NVDA94A+Top ranked
MSFT91AHigh signal
META87B+Monitor
AAPL85A-Monitor
AMZN82BMonitor
S&P 500 universe · T-90 LagT-90 Lag Backtest
S&P 500S&P 500 universeS&P 500 base
+6.70ppAudited PIT alphaDE_ALPHA vs SPY · 2019–2025 backtest · measurement-base re-pin, not an engine gain
T-90T-90 backtest lagNo look-ahead
3SEC + FRED + SimFinSEC · FRED · SIMFIN
Data from SEC EDGAR, FRED, and SimFinStrict T-90 methodology: the model cannot see data before it would have been available.

Confirmation bias is expensive

Most research tools help you collect evidence for the thesis you already like. Divergia keeps the math deterministic, then forces the thesis through disagreement: valuation assumptions, factor rank, sector model, risk context, and agent critique.

The goal is not a louder buy signal. It is a clearer map of where your thesis can break.

From S&P 500 universe to disagreement map

A structured loop for moving from candidate discovery to valuation, critique, and portfolio context without hiding the assumptions.

1

Filtering and Relative Ranking

We don’t use fixed thresholds. The full S&P 500 screening universe (~300+ eligible names post sector filter) is evaluated to return the top snapshot across Quality, Value, Growth, and Risk factors.

2

Sector Valuation (DCF & Bounded Models)

The system detects business nature. It applies Gordon Growth for financials, FFO for REITs, and dynamic WACC for the rest. Immediate results in 3 sensitivities (Bull, Base, Bear).

3

Quantitative Audit by 8 AI Agents

The AI here is not a chat. They are personas (Risk, Macro, Devil's Advocate, Quant) deterministically analyzing the ticker and forcing a solid contradictory thesis for your review.

4

Portfolio Construction (Sizing)

The Portfolio Engine consolidates the top assets by score and adjusts portfolio weights using inverse volatility optimization against the current market regime.

A research loop built for challenge

The product separates deterministic calculations from AI critique, so the story never outruns the evidence.

Ranked Evidence

Compare companies against peers

Compare companies mathematically instead of using static filters like P/E. See where a business sits across quality, value, growth, and risk factors.

  • Clusters by technical opportunity
  • Factor Rarity for anomalies

Sector Modeling

Build sector-aware valuation scenarios

Run interactive valuation models with institutional rigor. Adjust parameters to test Bull, Base, and Bear cases with dynamic WACC and automated cash flow margins.

  • Real-time capital costs
  • 3 Sensitivities with open parameters

Adversarial Review

AI agents that challenge your thesis

Get a structured second opinion from autonomous AI agents. They stress-test your thesis against macro trends, balance sheet risks, and news without reinforcing your biases.

  • 100% visible assumptions exposure
  • Antagonistic perspectives (Devil's Advocate)

Deterministic math. Adversarial review.

Divergia is not a generic bot fed with superficial prompts. The math is computed first; AI critique comes after.

Proprietary Quantitative Model, not a Data Wrapper

Our back-end logic does not make generic API calls to show empty dashboards. We process raw data right from financial statements, execute modeling, and generate the signal before you interact with the screen.

Transparent Backtest vs Hidden Disclaimer

The industry shies away from showing historicals because their logic doesn't hold up. Our system simulates a 90-day lag delay on SEC data availability, ensuring reported returns never assume information the investor didn't have in real-time.

Built for the Analyst, not the Gambler

If you are looking for magic alerts or intraday day-trading strategies, Divergia is not for you. This tool is built entirely to consolidate algorithmic Value Investing and structural review of solid businesses over the medium to long term.

Guarantees & Methodology

Critical answers before you trust your time to the engine.

Is this regulated financial advice?

No. Divergia is SaaS (Software as a Service) analytical infrastructure. The software does not recommend a 'Buy' to an individual without knowing their risk aversion. It provides the same structured research you would see on an institutional desk so you can make the final decision.

How do you ensure the backtest has no look-ahead bias?

Every historical data point is locked out for 90 days after the corporate accounting period closes. Under the official 2026-08-10 re-pin (production-parity base plus a determinism contract: strict freeze, lake-pinned regime inputs, and a historized risk-free rate), DE_ALPHA records +6.70pp daily-reconstructed, annualized point-in-time alpha versus SPY. That number is published with provenance: it originates in a measurement-base change built on coherence repairs, not a new engine-improvement claim.

Where do you pull raw fundamental data from?

The engine is fed by primary SEC EDGAR 10-K and 10-Q filings, normalized fundamental data via SimFin, pricing via market APIs, and macro environment via the Federal Reserve (FRED).

Doesn't AI run the risk of hallucinating data?

We have strictly separated our Deterministic Domain Layer from AI. All math, DCF, margins, and scores are purely quantitative. The AI is fed these static numbers and restricts its output to qualitative critique. The AI audits the math, it never generates it.

Is it really free?

Yes — analyze individual tickers free, no card required. Pro starts at $29/mo for the full S&P 500 screener, Portfolio Engine, and 2019–2025 T-90 backtests. Cancel anytime.

Stop asking only whether you are right

Run the math, expose the assumptions, and see where opposing evidence can weaken your thesis before the market does it for you.

Start Free Analysis

No card required. Free to analyze.