Five models.Reading the market.
Written rules watch five stock markets — the US, India, Canada, the UK and Korea. Every evening they answer one question: stay in, or step aside. We publish the answer, the full holdings, and every reason behind them. You decide what to do with it.
Free during preview · Full access, no card · Educational tool, not advice
Today · 2026-08-24
Today's posture
🇺🇸 US Stocks
Invested
Day 91 in state
🇮🇳 India Stocks
Invested
Day 46 in state
🇨🇦 Canada Stocks
Invested
Day 129 in state
🇬🇧 UK Stocks
Invested
—
🇰🇷 Korea Stocks
Invested
Day 32 in state
We don’t hold your money, place your trades, or pick your portfolio. We publish what the models see.
Why rules
Markets fall.
Rules don’t flinch.
When markets crash, index funds ride them all the way down — and gut feel sells too late, buys back too late. Written rules step aside when stress builds and step back in when it clears.
Worst drawdown · 2003–2025
BacktestThe defence rules cap the worst stretches — that’s the whole edge. Full Shield is the fullest-defence variant; the lighter shields draw down more. Hypothetical backtest, 2003–2025.
How it works
One strategy.Two layers.
Every model is one strategy with two moving parts. That’s the whole mental model.
Mostly the broad market index, plus a short list of the strongest stocks. Re-ranked on a fixed quarterly schedule, every market. What’s still strong stays; what’s faded gets swapped out. Sign up and you see every model’s actual current holdings — no teaser tier.
Full mechanics →Checked at every close against market-stress signals — trend and volatility for the US gate, a wider 13-signal panel (FX, oil, banks, momentum) for the international models. Invested: the basket is fully in. Defensive: the entire basket moves to cash equivalents. Whole basket, never half. When it flips, you get one email saying what changed and why. Most weeks: nothing moves.
What people do when it flips →The record
Tested against history.Now running in public.
The rules were run against 22 years of real history — including the three crises everyone remembers.
Sep 2008
Lehman.
Banking signals climbed for months. Defensive by mid-summer. Back in by April 2009.
Mar 2020
COVID.
Nobody saw it coming. The stress score did — by late February. Defensive in days.
2022–23
Rate-hike chop.
Multiple flips. Some saved real money. Some were false alarms. The rules don’t flinch either way.
Pre-registered. Walk-forward validated. Out-of-sample tested once.
Hypothetical backtest. Not a real investor. Not a prediction. Read the methodology →
What you get
A dashboard.A quiet inbox.
Signing up gets you exactly four things:
One dashboard.
Posture and actual holdings, all five models. Updated after every close.
Email only when it matters.
One per model, only on a posture change: what tripped, what the model now holds. Most weeks: silence.
Nothing hidden.
Every rule, threshold, holding, and signal change since June 1, 2026. Written down.
No card. No lock-in.
Full access during preview. Leave whenever.
Things people ask
Straight answers.
Is this advice?
No. We’re not registered as an investment adviser, portfolio manager, or dealer. For personal advice, talk to someone who is.
Will the model be wrong?
Sometimes. Some shifts save real money. Some are false alarms — and false alarms have a cost too (see taxes, below). The rules don’t flinch either way.
What about taxes and trading costs?
Real. In a taxable account, acting on the signal realizes gains and pays commissions — and the basket re-ranks quarterly, which is where most of the tax friction sits. The defensive switch itself flips rarely — about once a year across the 22-year backtest. In a TFSA or RRSP the tax disappears; the friction doesn’t. Backtest figures are net of trading costs but pre-tax; the dashboard shows an after-tax view at a representative ~43% marginal rate.
What does it cost?
Nothing during preview. No card. If a paid tier launches later, we’ll tell you first.
Take a look.
30 seconds to sign up. Free during preview. Quiet inbox, promised.
Foliox publishes systematic model output for educational use. Not investment advice. Not registered as an investment adviser, portfolio manager, or dealer. Hypothetical backtest results do not reflect real outcomes and do not predict future performance. Live signal record starts 2026-06-01. The live segment is model output published in real time; no client capital follows the model, so it is not the realized return of any investor.