How your five analysts work

Five AI trading algos, one portfolio — switch between them anytime from the bar up top. Here's exactly what each one is for, how it works, what its track record actually is (with the caveats, because that's the whole point), and how to use it.

Chasing Giants US small & mid-cap swing trades Open →

What it's for

It tells you exactly which US small and mid-cap stocks to buy at the next market open, and when to sell. You place every trade yourself at your own broker — it never touches your money.

How it works

Every market day it scans roughly 3,400 names for one setup: a confirmed uptrend pulling back to its moving average. When a name qualifies it prints a sized order for the next open. When more names qualify than it can hold, it ranks them by nine-month momentum and takes the strongest; the book holds at most two names at a time. Each position then exits on one of two rules — 5 straight bearish EMA closes, or a 30% trailing stop — checked daily until it closes.

Before any of that, it checks the weather. A small-cap trend filter (the Russell 2000 against its 50- and 200-day averages) decides the regime. In a bull it takes setups and deploys the whole bankroll across them. In a bear, with small-caps below those averages, it stops buying entirely and sits in cash, parked in a Treasury-bill fund like SGOV so it keeps earning interest while it waits, letting held names exit on their normal rules. Sitting out downturns is the single biggest reason the drawdowns stay shallow.

The track record

+223%Return, 2016 to mid-2026 (survivorship-free)

+337%S&P 500, same decade

+21%Out-of-sample 2022 to mid-2026 (S&P +65%)

−33%Max drawdown

66%Average share parked in T-bills

Survivorship-free backtest of the deployed daily-scan engine, delisted names included, with idle cash earning T-bill interest. We will be blunt: on return it trails the S&P 500. Over the full decade the honest figure is +223% against +337% for the index, and out-of-sample (2022 to mid-2026) it made +21% against +65%. Because the engine re-scans daily, the result does not depend on which day your clock starts — every start converges to the same book. It held only two names at a time and sat about two-thirds of the decade in T-bills, yet the max drawdown was still around −33%. From 2020 to mid-2026 it made +109% against the index's +150% — the price of all that defense. Limits: one cycle, the regime thresholds were tuned with these years visible, and the universe still omits pre-2019 delistings. Most days the radar reads FLAT or CASH. That is the system working, not a fault.

How to use it

Set your bankroll on the TODAY tab so order cards size real share counts.
Read TODAY each morning before the open; place any orders at your broker.
If TODAY reads CASH · DEFENSIVE, the regime is bearish: place no new buys, park the idle cash in SGOV (a 0-3 month Treasury-bill ETF) so it earns while you wait, and just manage what you hold. Sell SGOV the day buy orders fire. This is the system protecting you, not idling.
Log what you buy under MY HOLDINGS so the exit monitor watches it and flags the sell.

Pathfinder AI Roth IRA planner Open →

What it's for

An AI analyst pointed at a Roth IRA. It runs a model retirement book and shows you its ranked picks and target allocation, so you can mirror the discipline in your own account. Research only — you execute at your broker.

How it works

One rule, applied without discretion: it ranks every US stock of $2 billion and up by 12-month price momentum and holds the 20 strongest, equal-weight at 5% each, refreshed quarterly. No forecasts, no stock stories. The book stays fully invested (so cash reads $0), and the dashboard shows every holding and each quarterly refresh.

The track record

+32%/yrBacktest CAGR, 2019 to 2026 (S&P +16%)

6 of 8Years it beat the S&P 500

1.4Beta: swings harder than the market

−21%Backtest max drawdown (mild sample)

Backtested 2019 to 2026, survivorship-free, point-in-time, net of trading costs. We will be blunt: it beats the S&P's raw return by taking more risk, not proven skill. At a beta around 1.4 it swings harder than the market both ways, and 2019 to 2026 was a strong stretch for momentum. The −21% worst drop is a mild sample: real momentum crashes have historically run −40% to −65%, and those years are outside our data, so size for a −50% stress. There is no live track record yet; the first snapshot was locked June 2026 and zero quarters have been scored. A backtest is not money earned. Paper-trade the approach before mirroring it with real money.

How to use it

Open Pathfinder from the bar up top.
Start in the Simple view — a plain overview plus what the book holds.
Flip to Expert view (top-right) for conviction scores, charts, and the research deck.

Moonshot Bitcoin & Ethereum cycle advisor Open →

What it's for

An honest cycle advisor for Bitcoin and Ethereum only. It aims to flag the bottom of a correction (buy low) and the parabolic top (sell high), so you sell high and accumulate low. It never places trades — you act at your own exchange.

How it works

A 150-day trend filter decides risk-on vs risk-off, and it is the only piece allowed to move money — the layered signals below did not hold up out of sample. On top sits a regime-adaptive Upside Signal blending RSI, EMA and Bollinger Bands (momentum-led in a bull, mean-reversion-led in a bear). BTC also gets a capitulation/rainbow bottom model; ETH gets an honest asset-agnostic version, never the BTC machinery.

The track record

1.29BTC Sharpe (hold 1.09)

−58%Max drawdown (hold −84%)

$171Grew $1 since 2015 (holding made $241)

~60%Time in market

Backtested on Bitcoin, Jul 2015 to Jun 2026, on the shipped 150-day trend filter. We will be blunt: it does not beat buy-and-hold on return. Holding BTC turned $1 into about $241; this turned it into about $171. What you get for that is a smoother ride, not more money: a higher risk-adjusted return (Sharpe 1.29 vs 1.09) and a drawdown about a third shallower than holding (−58% vs −84%), by sitting out roughly 40% of the time. A handful of cycle tops drive the figures, so read them as indicative. Crypto can lose nearly everything.

How to use it

Open Moonshot from the bar up top.
Read the plain verdict per asset, the Upside Signal score, and the buy-low / sell-high callouts when they fire.
Expand "For technical folks" for the full backtest evidence. You trade at your own exchange.

Ascension Large caps becoming mega caps Open →

What it's for

It shows you the ten biggest names ascending the hardest right now — the $20–400B stocks climbing toward mega-cap. You hold each through the blow-off and step off when it breaks. The names that built the Mag-7 all passed through this exact window on the way up. You place every trade yourself.

How it works

It ranks every name between $20B and $400B by 12-month price momentum, drops the parabolic data-glitches (anything past +200% — they revert hardest), and holds the top ten equal-weight. Each position rides until it breaks 15% off its own peak, then the slot rotates into the next-best mover. The book is never in cash.

The track record

+310%Full return '19–'26 (S&P +157%)

+148%Last two years, out of design (S&P +42%)

−36%Max drawdown

+69%Across '21–'24, holding the 2022 crash (S&P +28%)

This is aggressive on purpose and high-beta: it roughly doubled the S&P by taking more risk in a bull market, not by skill a regression can prove. The 200% momentum cap is load-bearing — without it the headline was an inflated +680% built on data glitches and parabolic names that reverted hardest; filtering them cut the return in half but lowered the drawdown and turned a flat 2022 into +69%. It is also one market cycle (2019–2026, no true momentum crash in the sample), so size for a −50% stress, not the −36% backtest.

How to use it

Open Ascension from the bar up top.
Read today's ten ascending names — ticker, sector, market cap, 12-month momentum.
Mirror what you like at your own broker, sized for the drawdown, and step off a name when it breaks 15% off its high.

A few more questions

Do I need to connect my brokerage or exchange?

No. None of the three ever connect to your accounts, hold your money, or place trades. They publish research and you execute yourself, wherever you already trade.

Is this investment advice?

No. Everything here is a research tool. It knows nothing about your finances, taxes, or risk tolerance. You decide what to trade, and you own the outcome.

Why are the caveats so blunt?

Because the honesty is the product. A backtest is not a promise, in-sample is not out-of-sample, and a few good trades can flatter any record. We'd rather you trust us than be impressed by us — so the unflattering parts ship right alongside the numbers.

Can I cancel anytime?

Yes. Manage or cancel from the billing portal; access runs to the end of the period you paid for.