Foliox

Methodology

Built on rules.Validated by history.

Every threshold. Every input. Every signal change. Written down, public, dated. The radical transparency is the brand.

The basics

Rules, not guesses.

Five rules-based equity models read a fixed set of market inputs every day. When the inputs cross pre-defined thresholds, the model shifts posture. No discretion. No overrides. No “but this time it’s different.”

13

signals

international models · every close

22

years

backtested across cycles

5

models

five equity momentum markets

How the signals work

13 inputs. One decision.

Volatility

VIX, realized vol, vol-of-vol

FX

USD-pair stress in non-US markets

Commodity

Oil, copper signals

Momentum

Index 1m / 3m / vs 200d

Banking

Banking sector drawdown + momentum

Each signal gets a score from 0–1. The 13 scores combine into one stress reading. When the reading crosses a pre-defined threshold, the model shifts posture. That’s the entire decision rule.

This 13-signal panel drives the four international models. FLX-US uses a focused version of the same idea — a crash gate on trend and volatility (1- and 3-month trend plus VIX), with its four shields deciding which sleeve shelters when it fires.

The basket layer

Hold the leaders. Rotate the rest.

Each model keeps most of its money in the broad market index, and the rest in a short list of the strongest stocks. On a fixed quarterly schedule, every market, the universe is re-ranked by trailing 12-1 momentum. What’s still strong stays; what’s faded gets swapped out. No favourites, no sentiment.

RankTrailing 12-1 momentum
1
+38%
2
+31%
3
+27%
4
+22%
Cutoff
5
+18%
6
+14%
7
+10%
8
+6%
9
+2%
10
−3%
Top 4 heldRotated out

Illustrative — each model’s real holdings are on its dashboard

No favourites · No sentiment · No second chances

Backtest discipline

How we tested.

Every backtest used point-in-time data, walk-forward parameter selection, and an out-of-sample reveal locked in by pre-registration. No cherry-picking. No optimization on the test set. No survivorship bias.

  • Universe: point-in-time index constituents. Mergers and delistings handled honestly.
  • Parameters: selected on the in-sample window (2003–2017) and frozen.
  • Test: 2018–2025 held out and revealed exactly once.
  • Robustness: tested across five stress regimes — GFC, COVID, rate shock, oil shock, Euro crisis.
  • All audit CSVs are public. Every signal change is dated. Every assumption is named.
Pre-registered · Walk-forward validated · Survivorship-bias-free

Disclosures

Foliox publishes systematic model output for educational use only. 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. For personal financial advice, consult a registered professional.

See it in action.

Open the dashboard. Today’s posture across all five models.