> ## Documentation Index
> Fetch the complete documentation index at: https://docs.openstrat.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Build Your Own Trading System

> A complete guide to systematically building a personal trading framework

## Overview

Most people trade like this:
**Spot an opportunity → feel something → place an order → lose money → conclude “bad luck.”**

But the workflow of systematic trading should be:

> Define rules → test rules → execute rules → iterate rules based on data.

A “trading system,” simply put, is:

* In **which market**
* Using **which logic**
* With **what buy/sell rules**
* Combined with **money management and risk control**
* A **verifiable, reproducible process** executed repeatedly over the long run

It is not a “secret manual that never loses money,” but a framework that:

* Helps you **think clearly before acting**;
* Gives you a relatively definite way of operating in an uncertain market;
* Prevents you from being dragged around by emotions, news, and short-term noise.

The goal of this section is to help you go from 0 to 1:

> How to turn “scattered trading ideas” into a personal trading system that is **rule-based and testable, data-driven and observable, and iteratively improvable**.

***

## The Elements of a Trading System

Structurally, a complete trading system must answer at least five questions:

1. **What do you trade?** (market selection)
2. **When do you buy?** (entry rules)
3. **When do you sell?** (exit rules)
4. **How much do you buy?** (money management)
5. **What if you’re wrong?** (risk control)

### Market Selection

**Market selection = choose the “battlefield” first, then talk tactics.**

You need to clarify:

1. **Instrument / market**

   * Stocks: single names, indices, ETFs
   * Futures: equity index, commodities, rates, FX, etc.
   * FX / crypto: 24-hour trading, leverage commonly high
   * Options and other derivatives: complex structure, higher demands on risk management

   For most individual investors, it’s better to start with:
   **stocks + index funds + a small amount of futures/FX (if experienced)**, and expand gradually.

2. **Trading horizon / timeframe**

   * Ultra-short-term (intraday, T+0): extremely high demands on attention, execution, and cost control;
   * Short-term (hold days to weeks): higher demands on rhythm and reaction speed;
   * Medium-term (weeks to months): more emphasis on trend and fundamentals/thesis;
   * Long-term (years): closer to asset allocation and value investing.

   Simple guidance:

   * If you have a day job and can’t watch the screen: prefer a **medium/long-term system**;
   * If you trade full-time and are experienced: explore short-term or intraday systems.

3. **Personal constraints**

   * Trading time: can you watch the market? for how long?
   * Capital size: small vs large capital differs greatly in instrument choice and liquidity constraints;
   * Psychological traits: do you prefer fast pace or slow pace? how much volatility can you bear?

> Choose a market and timeframe that fit you first,
> then refine the strategy within that “arena”—far more effective than “randomly trying everything.”

***

### Entry Rules

**Entry rules = when to enter + why enter here.**

Key requirements: **quantifiable, executable, repeatable.**
Not “looks about right” or “feels like it will go up.”

Common entry logic can come from:

1. **Trend following**

   * For example:

     * Price breaks above a multi-day high (e.g., 20-day/55-day high);
     * Breaks above a key resistance level with volume confirmation;
     * Bullish moving-average alignment (short-term MA above long-term MA).

   Example rule (simplified):

   * If the close breaks above the highest high of the last 20 days, and volume > 1.5× the 20-day average volume → buy at next day’s open.

2. **Mean reversion**

   * Suitable for range-bound or clearly bounded markets;
   * Examples:

     * Price deviates significantly from a moving average and rebounds after being oversold;
     * Reverse at extremes of indicators (classic example: RSI oversold rebound).

3. **Fundamental / event-driven**

   * Fixed earnings triggers (e.g., earnings improve substantially);
   * Specific event theses (restructuring, dividends, policy catalyst, etc.), but you must quantify the “trigger conditions.”

No matter which approach, you must turn “vague ideas” into **specific rules that others can understand and execute**, such as:

* What exactly counts as a “breakout”—above which price?
* Close price or intraday price?
* Buy the same day, or buy at the next day’s open?
* What is the volume threshold?

***

### Exit Rules

**Exit rules = when to sell + how to sell.**

A system that only knows how to buy but not how to sell is basically no system.

Exits should include at least three parts:

1. **Stop-loss exit** (admit you’re wrong)

   * Price stop: break below a level/support/MA;
   * Condition stop: a premise no longer holds (e.g., fundamental deterioration, major negative news).

2. **Take-profit exit** (take money when you’re right)

   * Fixed targets: sell partially/fully at an expected return or technical target;
   * Trailing exits: raise the stop as price rises (e.g., trail a moving average or channel).

3. **Time stop / invalidation exit**

   * For example:

     * If price doesn’t move as expected within X days after entry, exit;
     * If held beyond a time limit with no trend, rotate to another instrument.

Example exit rules (trend-following system):

* Initial stop: 10% below entry or at a key support level;
* Once price gains more than 20%:

  * raise the stop above cost to ensure you don’t lose;
* If price closes below the 20-day moving average with confirmation → sell all.

**Key: write exit rules before entry to avoid “changing your mind on the fly” at critical moments.**

***

### Money Management

**Money management = how much to buy.**

It determines whether the same strategy results in **“steady gains”** or **“violent booms and busts.”**

Classic approaches:

1. **Fixed risk / fixed fraction**

   * e.g., maximum loss per trade ≤ 1%–2% of account equity;
   * compute per-share risk using the stop, then back out the position size.

2. **Tiered position sizing**

   * initial positions are typically not full size;
   * keep some “ammo” for with-trend adds or defense.

3. **Portfolio-level controls**

   * cap maximum weight per instrument (e.g., ≤ 15%–20% of total equity);
   * cap total exposure of highly correlated positions, e.g., a single sector ≤ XX%.

A simple example:

* Account equity 100,000, max per-trade risk 2% (2,000);
* Entry 10, stop 9 → per-share risk 1;
* Max shares = 2,000 ÷ 1 = 2,000.

> Money management isn’t to “make you earn less,”
> but to ensure you “don’t die during a streak of being wrong.”

***

### Risk Control

**Risk control = how to stop out + how to prevent “ruin risk.”**

More important than “how to make more” is:

> “In the worst case, how much could I lose? Can I accept that outcome?”

Key dimensions:

1. **Per-trade risk control**

   * Use price stops + position sizing to cap maximum loss per trade.

2. **Overall drawdown control**

   * Set a “maximum drawdown threshold” (e.g., 20%);
   * When drawdown hits the threshold, automatically reduce exposure or pause trading and enter “defense mode.”

3. **Leverage and liquidity risk**

   * Avoid heavy size or leverage in illiquid, highly volatile instruments;
   * For high-leverage markets like futures and FX, strict margin and stop mechanisms are essential.

4. **Black swan contingency**

   * Don’t go all-in on a single instrument, a single direction, or a single market;
   * Use moderate diversification and defensive assets (cash, bonds, etc.) to increase the safety buffer.

***

## System Testing

Writing a system down does not mean it works.
It must go through the full pipeline: **historical backtest → paper trading → small-capital live trading → iterative optimization**.

### Historical Backtesting

**Historical backtesting = use historical data to verify whether the system had an “edge” in the past.**

Basic approach:

1. Define the test window

   * Cover multiple regimes: up, down, and sideways;
   * Avoid selecting only periods favorable to the system (otherwise you’re fooling yourself).

2. Record core metrics:

   * total return, annualized return;
   * maximum drawdown;
   * win rate, payoff ratio, expectancy;
   * number of trades, holding-time distribution, etc.

3. Watch for common pitfalls:

   * **Overfitting (over-optimization)**:

     * keep tuning parameters until the historical curve looks “perfect,” as if it can profit everywhere;
     * live results then diverge badly.
   * **Survivorship bias**:

     * backtesting only instruments that “survived to today,” ignoring delisted/blown-up ones;
     * overestimates real performance.

A simple mindset:

> Backtests aren’t to find a “perfectly rising curve,”
> but to understand: **roughly how good the system is, and which regimes fit it best.**

***

### Paper Trading

**Paper trading = run “fake money, real rules” for a period.**

It serves three main purposes:

1. Verify whether rules are truly “executable”

   * Do you frequently need subjective judgment to apply the rules?
   * Is there too much ambiguity (leading to very different results between traders)?

2. Observe slippage, fees, and liquidity impact

   * Backtests are idealized; paper trading is closer to reality;
   * For short-term/high-frequency strategies, costs can dominate outcomes.

3. Practice execution and workflow

   * Run the system daily—from signal generation to order placement and logging;
   * See whether you can follow the rules consistently without the “reward stimulation” of real profits.

> The goal of paper trading isn’t “how much you make,”
> but to validate: **rules are clear and usable + you can actually follow them.**

***

### Live Optimization

After paper trading, start live trading with **small capital**, entering the “run and refine” stage.

Key points:

1. **Start with small size**

   * Use a small portion of capital you can “fully accept losing”;
   * Adapt to emotional swings and execution pressure.

2. **Strictly separate “in-system” trades from “out-of-system” trades**

   * Mark every trade in the log: system-compliant or impulsive;
   * Often losses aren’t because the system is bad, but because you “couldn’t resist.”

3. **Review and adjust periodically**

   * e.g., monthly/quarterly:

     * compute return, drawdown, win rate, payoff ratio;
     * compare to backtests and see if it deviates materially;
   * Follow the “**small steps, slow tuning**” principle:

     * change only a few parameters or one or two rules at a time;
     * after changes, re-backtest + re-paper trade—don’t “swap the whole system” frequently.

4. **Avoid constantly resetting everything**

   * Many beginners scrap a system after a few losses and switch again;
   * They remain forever in “new land exploration” and never truly execute any system.

> Good systems aren’t perfect from day one—
> they are built through continuous iteration **with clear rules as the foundation**.

***

## Core Concepts

When building a trading system, several ideas are especially critical:

1. **Rule-based / mechanical**

   * Rules should be clear enough that:

     > If someone else follows them, they can make roughly the same decisions.
   * “About right” and “feels like a good opportunity” are not a system.

2. **Positive expectancy**

   * Use expectancy to measure whether the system has a long-run edge:

     > Expectancy = win rate × average win − loss rate × average loss
   * Positive expectancy ≠ win every time. It means:

     > Over a sufficiently large sample, the overall outcome tends to be positive.

3. **Prepare for “random outcomes” as if they’re inevitable**

   * Losing streaks, extreme regimes, black swans—treat them as “they will happen sooner or later”;
   * System design and money management must reserve room for them in advance.

4. **A system ≠ indicators only**

   * Many so-called “systems” are just stacks of technical indicators;
   * A real system also includes:

     * money management;
     * risk control;
     * execution workflow;
     * psychological contingency plans.

5. **Fit between the person and the system**

   * Even a great system, if it doesn’t fit your personality, time, and risk preference,
     → you’ll likely “rewrite the rules yourself” in live trading;
   * The right system is the one you **are willing and able to execute for the long term**.

***

## Practical Application

Below is a simplified example showing how to build a basic “trend-following system” from scratch.

### Example: Broad Index ETF Trend-Following System (Simplified)

**1. Market selection**

* Instrument: a broad index ETF (good liquidity, high diversification);
* Timeframe: daily chart, medium-term;
* Suitable for: people with a day job who can spend 10–30 minutes a day checking the market.

**2. Entry rules (when to buy)**

* Consider buying when all of the following are true:

  1. Close breaks above a 60-day high;
  2. Close is above the 20-day moving average;
  3. Volume is not lower than the 20-day average volume.

* Execution:

  * Don’t rush to buy on the signal day—buy at next day’s open;
  * Initial position: 30% of total account equity.

**3. Exit rules (when to sell)**

* Initial stop:

  * If price closes below the 20-day moving average with confirmation → sell all;
* Take-profit / exit:

  * If after making a new high, there are 3 consecutive candles whose closes make new lows, and price breaks below the 20-day MA → sell;
  * If the decline from the most recent high exceeds 15% → forced exit (take-profit/stop-loss).

**4. Money management (how much to buy)**

* Max per-trade risk ≤ 2% of total equity;
* If the stop distance is large, reduce position size;
  For example:

  * If stop distance is 5% from entry, max position ≈ 2% ÷ 5% = 40%;
  * Conservatively, you can fix it at 30%–40%.

**5. Risk control**

* Only trade this one ETF at a time:

  * portfolio risk is simple and clear;
* If maximum drawdown from peak exceeds 15%:

  * pause all new entries;
  * check whether you followed the rules strictly;
  * if needed, pause the system and reassess/backtest.

**6. Testing and optimization**

* Backtest on 5–10 years of historical data:

  * record return, drawdown, win rate, payoff ratio;
* Paper trade for a few months to ensure the rules are executable in real conditions;
* Start live with small capital and observe:

  * whether false breakouts happen too often;
  * whether to adjust 60-day/20-day parameters;
  * whether to add additional filters.

> This example is not a “recommended strategy,”
> but a demonstration of:
> **how to decompose an idea into a system structure that can be written, tested, and executed.**

***

## FAQ

### Q1: The system is written, but I always want to change rules on the fly—what do I do?

This is extremely common—human nature resists loss and uncertainty.

How to handle it:

1. **Log “in-system” vs “out-of-system” trades separately**

   * Tag each trade: fully compliant or impulsive;
   * After a period, compare performance—you’ll often see that
     impulsive “out-of-system” trades drag down results.

2. **Set a “rule-change window”**

   * For example: only evaluate and tweak once per month/quarter during review;
   * The rest of the time, execute only—no ad-hoc rule changes.

3. **Test new ideas in paper trading or with small capital first**

   * You can experiment, but don’t let it directly affect the main system’s exposure;
   * Prevent “emotion + new rules” from together destabilizing the account.

***

### Q2: If the market regime changes and the system stops working, what should I do?

First distinguish:

* **Short-term mismatch** (the strategy performs poorly in the current phase but still has a long-run edge);
* **Structural breakdown** (the market structure the strategy relies on has changed).

Suggestions:

1. Compare to historical backtests:

   * Is current drawdown/performance still within historical variation?
   * Have you seen similar phases before, followed by recovery?

2. Diversify strategies and markets:

   * Don’t bet all capital on one logic or one market;
   * Combine different styles (trend, mean reversion, value, event-driven).

3. Tiered response:

   * If it’s a mismatch, reduce size but keep the system running;
   * If it’s confirmed structural failure (e.g., market mechanism changes materially),
     → cut exposure or exit, then redevelop or switch strategies.

> A system is not “set and forget.”
> **A system also needs to be “managed systematically.”**

***

### Q3: Can beginners directly use someone else’s trading system?

You can reference it, but **don’t copy it blindly**.

Reasons:

1. Their system is based on their:

   * capital size;
   * risk preference;
   * time and energy;
   * personality and execution capability.

2. Even if you copy the rules, you:

   * often won’t execute as strictly as they do;
   * are more likely to doubt and quit halfway during drawdowns.

Better approach:

* Treat others’ systems as **templates and inspiration**:

  * learn the structure: how entries, exits, money management, and risk control are designed;
  * then simplify and customize for your reality;
* At minimum, you should:

  * personally run a basic backtest;
  * know in what regimes the system tends to work, and where it tends to struggle.

> The only system you can use long-term
> is the one you **understand, trust, and can execute**.

***

## Summary

* Building a trading system means moving from “trading by feel” to trading with **rules, data, and contingency plans**;
* A complete system must clarify five core elements:

  1. **Market selection**: what to trade, which timeframe, under what constraints;
  2. **Entry rules**: when to buy and why;
  3. **Exit rules**: when to sell and how;
  4. **Money management**: how much to buy each time and in total;
  5. **Risk control**: worst-case loss, and how to avoid blow-ups and massive drawdowns.
* Any system must go through:

  * **historical backtest → paper trading → small-capital live trading → data-driven optimization**;
* The ultimate purpose of a trading system is not “never losing,”
  but to help you, in a noisy and emotional market:

  * know what you’re doing;
  * accept the logic behind the outcomes;
  * survive long-term and continuously improve your odds.

***

## Further Reading

* Related resources links:

  * Strategy and systematic-trading topics under “investor education/quant investing/programmatic trading” on major broker and fund-company websites;
  * Series articles and open courses on “systematic trading,” “strategy backtesting,” and “money management” in quant and trader communities;
  * Teaching docs and sample strategies from backtesting and paper-trading platforms (e.g., common quant platforms).

* Recommended books or articles:

  * Van K. Tharp, *Trade Your Way to Financial Freedom* — a classic introduction to system construction, expectancy, and money management;
  * *Way of the Turtle* — shows how to write rules clearly and execute a complete system with discipline;
  * Mark Douglas, *Trading in the Zone* — helps you understand systematic trading psychologically and execute rules with probabilistic thinking;
  * System trading books by Perry J. Kaufman and others (many translated editions) — more technical and quantitative approaches to system design and testing.