EN IT
Trading Desk · Method
About My Method — AiTrading67's operating framework
A systematic trading method, built on weekly signals, active trade management on a two-axis time + value framework, and no automatic closures. The model provides references; the choice always remains the trader's.
3 June 2026 — Fabio Gentili · 14 min read MethodEducation
Introduction

A systematic trading method rests on a simple premise: markets show recurring structures that can be identified, measured and exploited in a repeatable way — provided you have a clear process, coherent data and rules that don't change halfway through the game.

What you'll find here is not a forecasting system, nor a magic formula. It is an operating framework built on years of personal work on a structured desk, validated on thousands of real signals and continuously updated based on empirical results.


Core philosophy

Two-axis method: time + value

An open position never closes automatically.

The whole system rests on one simple idea: the model provides references, never automations. Every target in a trade's roadmap carries two complementary pieces of information:

  • the when — the week in which it could statistically make sense to consider trimming part of the position
  • the how much — the price value you can reasonably expect the move at that point

The two axes live together: the "Operating Levels" box on stocks and the "Profit-taking areas" card on indices are two views of the same roadmap.

The choice to close — partially or in full — or to stay in is always the individual trader's, who acts on their own judgement and risk profile. The system does not close trades for you; it gives you the map. This doctrine runs through every layer of the process described below.

Architecture

The system's components

The process is organised in six sequential layers. Each layer performs a specific function and produces a precise output that feeds the next one.

L0
Layer 0
Signal
Starting point
The weekly BUY signal — a proprietary Inversion Point

The system fires off a proprietary indicator that combines:

  • An adaptive trailing stop built on actual volatility (ATR)
  • An ultra-fast regime-change trigger

When price crosses upward through the dynamic reference level, a weekly BUY signal is issued — a point event marking a potential structural shift from bearish to bullish.

Why the weekly timeframe
  • Each candle carries more weight
  • Noise is lower than on daily or intraday frames
  • False signals are far less frequent

The signal is the starting point — not an automatic buy order. It is the invitation to begin a structured analysis.

Monitored universe
  • 330+ global assets
  • US large- and mid-cap stocks
  • European stocks (Italy, Germany, France, United Kingdom)
  • US sector ETFs
  • Main US and European macro indices

The universe is constantly evolving: tickers come in and out based on liquidity, technical quality and macro relevance.

L1
Layer 1
Data
Data foundation
The data foundation — two timeframes, one complete picture
Operating structure
  • Daily — operational: generates setups, defines levels
  • Weekly — structural validator: no signal is traded against the weekly structure

All baseline and proprietary indicators are rebuilt in-house with standardised Python code in a single pass — full methodological control, full reproducibility.

Indicators — Structure and Trend
  • EMA 20 and EMA 50 on both timeframes
  • Full five-component Ichimoku Kinko Hyo
  • Golden Cross and Death Cross with statistically validated distinction between fresh (dangerous) and mature (favourable) events
Indicators — Momentum and Volumes
  • MACD with histogram, classified into four editorial states: Improvement, Deterioration, Bullish cross, Decay
  • RSI (with qualitative zones beyond the classic 70/30 levels)
  • ATR (Average True Range)
  • OBV (price/flow divergences)
  • Buy pressure (BuyVolume%)
  • VWMA (volume-weighted support/resistance)
  • Rolling highs and lows over 7, 30 and 90 days (anchor objective structural levels)
Weekly ATR — the model's unit of measurement

The ATR on the weekly timeframe is the cardinal unit with which the system sizes everything: the width of the No-Trade Zone (the non-operating band), the scale of the NTZ graphic widget, the ranges of the profit-taking areas on the indices, the average weekly increment of the roadmap targets. Working in multiples of the weekly ATR means automatically adapting the model's metrics to the typical volatility of each instrument.

L2
Layer 2
Process
Decision process
Eight sequential steps — from alert to documented decision

Step 0 — Market context

  • Regime of S&P 500 and Nasdaq (directional combination SPY × QQQ)
  • BPSPX breadth indicator
  • MOVE Index bond volatility
  • State of the SPDR sector ETFs

The context feeds the final decision as qualitative information — it does not automatically block trades. It does enter, however, as a risk modulator inside the Entry Gate (Step 2).

Step 1 — Anagraphics and Screening

The freshly issued signal is placed in its context: the sector it belongs to, the reference sector ETF, its position relative to the current macro regime. Fresh weekly signals are separated from trades already in progress.

Step 2 — Extended entry gate (4 dimensions)

It answers a single question: is this signal worth analysing, and under which conditions?

The gate combines three independent dimensions to reach a nuanced operability call:

1. Technical Quality band of the setup, five levels
  • High — full-quality setup, conditions aligned
  • Medium — ordinary setup, some compromise
  • Reduced — marginal setup, to be opened with a tighter stop if at all
  • Watchlist — setup not immediately operational, kept under observation pending confirmations
  • Exclusion — setup not worth opening due to adverse technical or macro conditions
2. Opening policy

Open in standard mode, open with a tighter stop loss in case of marginal quality, or do not open.

3. Macro-risk modulator

Reduces or zeroes the sizing in adverse market regimes: macro regime under stress, VIX/MOVE/BPSPX gauges in critical zone, sector of belonging deteriorating.

The result is not "trade discarded yes/no" but "openable, and in which mode" or "not openable, and why". The gate's calibration dataset consists of thousands of historic signals on stocks and a statistically validated W+1 probability band for EU/UK indices.

Step 3 — Category filters

Category-specific rules: proximity to all-time highs in industrial stocks, differentiated behaviour of the CD1D indicator with inverted semantics on indices versus stocks, dedicated W+1 probability band for EU/UK indices.

Step 4 — Entry timing

The classic trigger is the breakout above the upper edge of the No-Trade Zone (NTZ) for bullish setups, or the break of the lower edge for bearish setups. An early intra-week entry on breakout or pullback to the structural level is also possible (RPDB rule).

Steps 5–8 — Full technical analysis, sizing and documentation

Described in the following layers (L3 Analysis, L4 Phase & Sizing, L5 Active trade management).

L3
Layer 3
Analysis
Technical analysis
Full Two-Side analysis — two phases, five mandatory elements

Phase 1 — Five mandatory elements

1. Distance Score Overhead (DSO, 0–100)

Actual operating room before the main resistance, normalised in ATR multiples. Five reading bands: DSO > 80 wide running room; 60–80 sufficient room; 30–60 approaching recent highs; DSO < 30 hugging the highs; DATH negative price exploring above the ATH (uncharted territory).

2. Entry orientation (RPDB)

How many times price has already tested the key level. First attack = caution and possible pullback entry (RPDB rule). Repeated tests = heavily contested level, direct breakout is risky.

3. Pre-Event Tension (TPE, 0–10)

When a high-impact event (earnings, FOMC, macro data) is imminent, TPE rises. Above threshold, new directional entries are suspended or downsized.

4. ADX with directional spread (+DI/−DI) and daily Cumulative Delta

Trend strength and direction clarity. Cumulated buy versus sell pressure beneath the surface of price.

5. Signal Strength + IQS Phase — two complementary readings

The system exposes two distinct indicators that must be read together:

  • Signal Strength (0–100) — measures how convincing a setup is in its directional orientation. For bullish trades it is the overall technical quality of the long scenario; for bearish trades it is the cumulated downward pressure on price. It is recalibrated on per-category statistical models (US stocks, EU/UK stocks, US indices, EU/UK indices).
  • IQS Phase (0–100) — describes the technical-cycle phase the instrument is in: compressed / ordinary / tight / mature. It is not a quality judgement on the setup: a high IQS Phase means "already-stretched setup, caution", a low IQS Phase means "compression awaiting release, active attention".

The two readings together produce a well-qualified directional scenario: a setup with high Strength and low IQS Phase is a convincing directional scenario at the start; one with high Strength and mature IQS Phase is a convincing directional scenario already extended that demands attention to timing.

Phase 2 — Options analysis (for high-quality setups)

  • Open interest and volume per strike reveal concentration clusters
  • Market makers run hedges on those clusters, defending those price levels
  • Supports and resistances form that are not visible on classical technical analysis
  • An options cluster coinciding with a rolling high or a key moving average carries far greater weight
L4
Layer 4
Phase & Sizing
Conviction & sizing
Three complementary indicators — distinct readings, integrated decision

After the full technical analysis, the system exposes three distinct indicators, each with a precise function:

1. Signal Strength (0–100) — driver of the sizing

Five editorial reading bands: convincing (≥70), medium (50–69), medium-weak (30–49), weak (10–29), muted (<10). The Strength is remodulated each week based on price behaviour: a high Strength fading in the following weeks signals loss of momentum, a Strength building up signals consolidation.

2. IQS Phase (0–100) — cycle-phase descriptor

Complementary to Strength. Four bands: compressed/accumulation (<30), ordinary (30–65), tight (65–80), mature (≥80). It is to be read as "where price sits in the cycle", not as "how good the setup is". It is explicitly disconnected from sizing.

3. Sizing class — derived from Signal Strength

Position-sizing class, derived from Strength via fixed thresholds:

  • Strength ≥ 70 → Full
  • Strength 50–69 → Standard
  • Strength 30–49 → Reduced
  • Strength < 30 → Minimal

To be read as a maximum cap: you can always go below it, never above.

Explicit Strength ↔ IQS Phase decoupling

A setup in a compressed phase (low IQS Phase) can have a high Strength and thus a Full Sizing class, and vice versa a setup in a mature phase can have medium Strength and thus a Standard Sizing class. Strength and IQS Phase describe two different dimensions of the setup; keeping them separate allows for more nuanced readings of "directional conviction" and "position in the cycle".

TradeLog — the continuous fine-tuning

Every analysed signal — including those discarded or kept under observation — is recorded in an internal database. The TradeLog is the basis for the continuous fine-tuning of the method, the recalibration of the per-category Signal Strength models, and the historical statistics on the Extended entry gate.

L5
Layer 5
Active trade management
Roadmap, not automations
The roadmap of the open trade — references, not automations
Once a trade is open, the system does not run it for you: it gives you a roadmap built on the Two-axis time + value method.

Target Roadmap — three two-dimensional references (TP1 / TP2 / TP3)

Each Take Profit on the roadmap combines a time gate (the week in which it could statistically make sense to consider trimming) and a price value (how much you can expect the move at that point). The gates differ by instrument category:

  • Stocks — W6, W8, W10 (6, 8 and 10 weeks from trade entry) with indicative fractions of 34% / 33% / 33%
  • US indices (SPY / QQQ / IWM) — W6, W11, W17, with price values computed dynamically off the weekly ATR
  • EU/UK indices (CAC / DAX / MIB / ISF) — static table with levels +4%, +7%, +10% off the Entry, over the same median W6 / W11 / W17 timings
The 3 TPs are never automatic closure thresholds: they are hypothetical closure thresholds computed by our model. The choice to close — partially or in full — or to stay in is always the individual trader's, who acts on their own judgement and risk profile.

Coverage complete

When all three Take Profit targets have been reached by price, the system declares Coverage complete: the model has exhausted the time + value references it proposes for that trade. The trade continues at the trader's discretion until the signal reverses (PI SELL). It is not a closing recommendation.

Structural stop loss (SL1 / SL2)

The main stop loss (SL1) is anchored to the current weekly Inversion Point — a dynamic reference that updates each week as the trade progresses. SL2 is a secondary active-management stop, anchored to a previous historic Inversion Point (lookback up to four weeks back, on the protective side relative to SL1).

Alternative SHORT — service reference inside LONG trades

In posts dedicated to a bullish trade, alongside the full Long operating scheme an alternative Short setup is also shown. It is not an active operating call: it describes what the system would set up if the Inversion Point were broken downward and a bearish reversal kicked in. The Entry coincides with the current weekly Inversion Point; the bearish targets and stops are computed with dedicated coefficients off the weekly ATR. It gives the reader a ready alternative reference in case the main setup collapses unexpectedly.

Profit-taking areas (indices) with ±Weekly ATR range

On Broad indices in open trades, the three price areas of the roadmap are presented as a price band (lower range – upper range, with width equal to the weekly ATR) rather than as a hard value. The band recognises that the statistical level is not an exact point but a zone where the level may materialise, accounting for the typical volatility of the instrument. Reached areas are frozen at their first-touch value.

Index exit gate — three-week observation

When a sell signal (PI SELL) fires on an index with an open long trade, the system does not close the coverage immediately: it opens a three-week observation window. If in the following three weeks price builds up a confirming drawdown (indicative threshold around 6%) or a technical condition reinforcing the bearish signal kicks in, the exit is confirmed; otherwise the signal is treated as a false positive and the trade continues. On stocks a PI SELL signal directly opens the Decline phase.


Three pillars

Why it works

Three structural pillars — and what the data actually say.

We do not publish aggregate return statistics: an average number across thousands of trades says little without knowing how they were closed, in which market regime, with what sizing. The system does not try to predict the future. It identifies the conditions in which the past shows a statistically more favourable distribution — and concentrates capital there.

PILLAR 1
Active selection of favourable conditions

The Extended entry gate identifies the conditions that historically produce the most negative trades and those that historically produce the most winning ones. Some evidence from the historic data:

  • Fresh weekly Golden Cross (≤ 3 candles from the crossover): almost doubles the incidence of quick failures versus baseline
  • Fresh daily Death Cross (five sessions after the crossover): triples the incidence of quick failures
  • Mature Golden Cross (6–13 candles from the crossover) with price compressed on the 50-week moving average: produces a quick-failure incidence equal to one third of baseline — the single best cell documented in the dataset

Baseline metrics (on raw signals, no filter):

  • Baseline quick-failure incidence: 18.6% on stocks, 50.8% on indices
  • Best cell (mature GC + EMA50 compression): 3.3% over 30 historic cases

The gate exists to tilt the portfolio toward the right side of this distribution.

PILLAR 2
Active prevention of quick losses

The Quick Failure — a trade closing in the red in its first week — is the most damaging loss category: it burns capital before the setup has time to develop.

The system intercepts upstream the conditions that historically produce high Quick Failure rates:

  • Proximity to all-time highs in certain sectors
  • Directional volume weakness
  • Pre-event tension
  • Weekly momentum deteriorating
  • Signal Strength muted or fading in the following weeks

These signals get reclassified or moved into observation — reducing exposure at the moment of maximum vulnerability.

PILLAR 3
Intelligent capital allocation

A system that sizes every trade the same way wastes the informational edge built in the prior steps.

The Sizing class derived from Signal Strength tilts the portfolio toward the setups with the strongest directional conviction of the week. Not aggressive concentration — rational distribution of capital based on measured quality, not on intuition.


A personal note
Why I share it

This system was born for personal use. I am an individual trader — I operate on indices, stocks and equities from a structured desk — and over the years I built this framework to bring rigour and reproducibility to my operating decisions. Every document, every rule, every version is the result of ongoing observation, backtesting and review of my own real trades.

I share it publicly for two reasons.

Educational — the systematic approach to trading is under-represented in the Italian debate, which is often dominated by intuitive analyses or signals without a declared methodology.

Community — building a community of traders who share the same methodological rigour: people to compare notes with, improve the system together and grow alongside one another.

I don't sell dreams. I don't manage third-party capital. I publish weekly analyses, share the method and make available the tools I use every week. If you appreciate methodological transparency and want to follow the process in real time, this is the right place.

Methodological note. Personal trading methodology shared for educational purposes and to build a community. Not financial advice. Individual trader — not a fund, not a course. The statistical findings refer to backtest data and do not guarantee future results.
Copyright — Texts, analysis methodology and proprietary indicators (IQS Lite, IQS Phase, sub-score, NTZ, CD1D, DSO, CFI, VRI, SRI, CCP, RPDB, TPE, PI, Signal Strength, Entry Protocol) published on aitrading67.com are © Fabio Gentili. Reproduction, even partial, in any form and for any purpose, is forbidden without the author’s written authorisation.
Share WhatsApp Telegram Gmail LinkedIn