The GNG Platform

    Tools built by analysts, sharpened by software.

    An end-to-end research bench. A custom agentic AI on top of every dataset we own. Eleven coordinated surfaces, one opinionated workflow.

    GNG AnalystChapter 01

    An agentic AI built on top of every dataset we own.

    GNG Analyst is not a chatbot bolted on the side of a fintech app. It is the research bench, talking back. It reads your portfolio live, queries twenty years of historical prices, runs backtests and Monte Carlos through our own tools, and remembers everything you have worked on across sessions.

    Built in-house, not a wrapper.

    Custom orchestration over OpenAI and Anthropic with our own tool layer, memory layer, and routing logic. No chatbot SDK underneath.

    Powered by GNG's full database.

    Twenty years of price history, every metric we compute, every research post our analysts publish, all retrievable in-conversation.

    Custom RAG memory per user.

    Persistent memory across every session. Upload notes, share files, build out an investing knowledge graph that grows with you.

    Build your own analysts.

    A marketplace of starter agents plus a builder. Spin up a dividend specialist, a deep-value scout, an options Greeks reader, each with its own tools and memory.

    ai.gngresearch.com
    gpt-5
    Run a 10-year backtest on my retirement portfolio vs SPY, and project dividend income out three years.
    GNG Analyst
    I will delegate the historical backtest and the dividend projection to the specialist tools, then summarize both.
    delegate_to_subagent>backtester
    delegate_to_subagent>dividend_forecaster
    emit_widget>equity_curve
    Backtest, 10y, monthly rebalance
    +212.4%
    CAGR
    12.1%
    Max DD
    -18.7%
    Sharpe
    1.04
    The retirement sleeve beat SPY by 38 bps annualized with lower max drawdown. Forward dividend income lands near $14,200 in year three at a 6.1% blended growth rate. Want me to stress-test the tail with a Monte Carlo run?
    Reply, or call /summon to load a subagent...
    Send
    Research TerminalChapter 02

    Screen 4,500 companies on the metrics that actually move price.

    Filter the entire US market on 100+ financial fields plus our proprietary Safety, Quality, and Valuation scores. Save filter sets, compare side by side, export to CSV. The whole index, none of the noise.

    • Proprietary Safety, Quality, and Valuation scores updated daily
    • Twenty years of historical fundamentals on every ticker
    • Saved screens and column presets per user
    • CSV, Excel, and PDF exports for offline work
    Screener/Safety > 75, P/E < 22, Yield > 1.5%
    217 results
    TickerPriceSafetyQualityValue
    JPM$198.42889271
    MSFT$415.16959764
    KO$62.84918476
    XOM$113.27798182
    Showing 4 of 217, full table inside
    BacktesterChapter 03

    Test the strategy on twenty years before you risk a dollar.

    Define a basket, a rebalance rule, and a benchmark. We replay daily data back to 2005, compute drawdowns, rolling Sharpe, factor exposures, and a full trade ledger. Compare runs side by side until the thesis holds up.

    • Fixed-share and percentage-weight rebalancing
    • Custom rebalance cadence: monthly, quarterly, annual, or rule-based
    • Side-by-side comparison of multiple runs against a benchmark
    • Full per-trade ledger with realized and unrealized P&L
    Quality Compounders · 20Y
    Final+487.6%
    CAGR
    9.2%
    Max DD
    -22.4%
    Sharpe
    0.84
    vs SPY
    +138 bps
    Monte CarloChapter 04

    A thousand futures. One portfolio. Honest probability bands.

    Stress-test any portfolio with 1,000 to 10,000 simulated paths. We sample from rolling historical returns and volatility, so the tails reflect real regime behavior, not a textbook normal curve. Spot the 5th-percentile outcome before you find it the hard way.

    • 5th, 50th, and 95th percentile probability bands
    • Sample from historical regime windows, not assumed distributions
    • Withdrawal modeling for retirement planning scenarios
    • Compare two simulations side by side with shared assumptions
    Monte Carlo · 30Y horizon · 10,000 paths
    Median +298%
    Year 0Year 10Year 20Year 30
    5th %ile
    +72%
    Median
    +298%
    95th %ile
    +811%
    Dividend ForecasterChapter 05

    Know what next year's check looks like.

    Project forward dividend income on every holding, ten years out. We blend declared rates, payout ratios, free cash flow trajectory, and historical growth to model what each position pays you. DRIP toggle on or off.

    • Per-position forward yield and ten-year income projection
    • Growth rates pulled from free cash flow trajectory, not just declared history
    • DRIP modeling that compounds reinvested shares forward
    • Side-by-side scenarios: status quo vs trim-and-rotate
    Forward dividend income · DRIP off
    $26,780 / yr by Y10
    $8.2kY1
    $9.1kY2
    $10.3kY3
    $11.8kY4
    $13.6kY5
    $15.5kY6
    $17.8kY7
    $20.4kY8
    $23.3kY9
    $26.8kY10
    Year 1
    $8,200
    Year 5
    $13,580
    CAGR
    12.6%
    Portfolio OptimizerChapter 06

    Find the weights you should have been running.

    Plot the efficient frontier on any basket you own. Maximize Sharpe, target a volatility, or run risk parity. Constraints for minimum positions, sector caps, turnover. Apply the optimized weights with one click.

    • Mean-variance, max-Sharpe, min-volatility, and risk-parity engines
    • Constrained optimization: position caps, sector caps, turnover limits
    • Frontier comparison between current allocation and optimized weights
    • One-click apply on manual portfolios with full pre-trade diff
    Efficient frontier · 8 holdings
    Sharpe target 1.10
    Lower riskHigher return >
    Current
    Optimized
    Vol drop
    -3.4%
    Return lift
    +1.8%
    Sharpe
    1.10
    Top PicksChapter 07

    The analysts' best ideas, written down with a date stamp.

    Each GNG analyst publishes a live top-picks list with conviction, target price, and full thesis. Public track record, no after-the-fact editing. Follow the ones whose style matches yours.

    • Live picks per analyst with conviction, target, and entry zone
    • Public, time-stamped track record per author
    • One-click jump to the underlying thesis article
    • Model portfolios assembled from analyst picks
    Top Picks · Glenn Ford
    Updated this week
    ASML
    $712High conv

    Monopoly on EUV. Pricing power through the cycle.

    +18%
    since pick
    BRK.B
    $425High conv

    Cash mountain plus insurance float, optionality on every deal.

    +11%
    since pick
    TSM
    $162Med conv

    Foundry leadership; geopolitics priced in.

    +24%
    since pick
    Research PostsChapter 08

    Long-form analysis from the desk, not a content farm.

    Every article on GNG comes from a working analyst with a public track record. Investment theses, sector deep dives, post-mortems on the picks that did not work. Filter by author, sector, or tag. Save for later, share with the community.

    • Investment theses with linked data, charts, and model assumptions
    • Sector deep dives and earnings post-mortems on cadence
    • Author profiles with full publication history and pick performance
    • Save, share, and react inline with the rest of the community
    Sector deep dive/May 12

    Why the semiconductor floor has shifted

    Glenn Ford·12 min read
    Thesis/May 09

    The unloved insurance compounder

    Adam Galas·8 min read
    Post-mortem/May 06

    What I got wrong on regional banks

    Connor Graham·6 min read
    Community ChatChapter 09

    A room with the analysts, not a comment section.

    A live community chat where subscribers and GNG analysts trade ideas in real time. Earnings reactions as they print, position changes, the questions you would want to ask the team. Quiet most days, loud when it matters.

    • Live chat rooms by sector, strategy, and event
    • Analysts and subscribers in the same space, no segregated tiers
    • Direct messages and pinned reading lists per room
    • Hosted on Rocket.Chat under your GNG login, no third-party signup
    # msft-earnings
    24 online
    Glenn FordGNG9:31

    MSFT print at 4. Watching segment splits on Azure, especially AI-attached.

    Adam GalasGNG9:42

    Read your prep note. Are we still anchored to the +30% Azure scenario or has guide changed your priors?

    Glenn FordGNG9:44

    Held, but I want to see capex commentary first. If FY guide moves, I will post a revised model in #msft tonight.

    Maria L.9:46

    Trimmed 20% into the print. Sleeping easier.

    Message #msft-earnings...
    PricingChapter 10

    A plan for every stage. Fourteen-day trial on every paid plan.

    Start free, or run a paid trial with no card friction. Cancel from your billing page in two clicks.

    Two ways in

    Open a trial, or stay on the free notebook.

    Fourteen days of full platform access on any paid tier, no card friction. Or stay on the free read-only tier and pick up the weekly note. Both paths are honest defaults.

    Free forever on the read-only tier. No card. Cancel anytime.