\section{XAUUSD Trading AI: Technical
Whitepaper}\label{xauusd-trading-ai-technical-whitepaper}

\subsection{Machine Learning Framework with Smart Money Concepts
Integration}\label{machine-learning-framework-with-smart-money-concepts-integration}

\textbf{Version 1.0} \textbar{} \textbf{Date: September 18, 2025}
\textbar{} \textbf{Author: Jonus Nattapong Tapachom}

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{Executive Summary}\label{executive-summary}

This technical whitepaper presents a comprehensive algorithmic trading
framework for XAUUSD (Gold/USD futures) price prediction, integrating
Smart Money Concepts (SMC) with advanced machine learning techniques.
The system achieves an 85.4\% win rate across 1,247 trades in
backtesting (2015-2020), with a Sharpe ratio of 1.41 and total return of
18.2\%.

\textbf{Key Technical Achievements:} - \textbf{23-Feature Engineering
Pipeline}: Combining traditional technical indicators with SMC-derived
features - \textbf{XGBoost Optimization}: Hyperparameter-tuned gradient
boosting with class balancing - \textbf{Time-Series Cross-Validation}:
Preventing data leakage in temporal predictions - \textbf{Multi-Regime
Robustness}: Consistent performance across bull, bear, and sideways
markets

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{1. System Architecture}\label{system-architecture}

\subsubsection{1.1 Core Components}\label{core-components}

\begin{verbatim}
┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│   Data Pipeline │───▶│ Feature Engineer │───▶│   ML Model      │
│                 │    │                  │    │                 │
│ • Yahoo Finance │    │ • Technical      │    │ • XGBoost       │
│ • Preprocessing │    │ • SMC Features   │    │ • Prediction    │
│ • Quality Check │    │ • Normalization  │    │ • Probability   │
└─────────────────┘    └──────────────────┘    └─────────────────┘
                                                       │
┌─────────────────┐    ┌──────────────────┐           ▼
│ Backtesting     │◀───│ Strategy Engine  │    ┌─────────────────┐
│ Framework       │    │                  │    │ Signal          │
│                 │    │ • Position       │    │ Generation      │
│ • Performance   │    │ • Risk Mgmt      │    │                 │
│ • Metrics       │    │ • Execution      │    └─────────────────┘
└─────────────────┘    └──────────────────┘
\end{verbatim}

\subsubsection{1.2 Data Flow Architecture}\label{data-flow-architecture}

\begin{Shaded}
\begin{Highlighting}[]
\NormalTok{graph TD}
\NormalTok{    A[Yahoo Finance API] {-}{-}\textgreater{} B[Raw Price Data]}
\NormalTok{    B {-}{-}\textgreater{} C[Data Validation]}
\NormalTok{    C {-}{-}\textgreater{} D[Technical Indicators]}
\NormalTok{    D {-}{-}\textgreater{} E[SMC Feature Extraction]}
\NormalTok{    E {-}{-}\textgreater{} F[Feature Normalization]}
\NormalTok{    F {-}{-}\textgreater{} G[Train/Validation Split]}
\NormalTok{    G {-}{-}\textgreater{} H[XGBoost Training]}
\NormalTok{    H {-}{-}\textgreater{} I[Model Validation]}
\NormalTok{    I {-}{-}\textgreater{} J[Backtesting Engine]}
\NormalTok{    J {-}{-}\textgreater{} K[Performance Analysis]}
\end{Highlighting}
\end{Shaded}

\subsubsection{1.3 Dataset Flow Diagram}\label{dataset-flow-diagram}

\begin{Shaded}
\begin{Highlighting}[]
\NormalTok{graph TD}
\NormalTok{    A[Yahoo Finance\textless{}br/\textgreater{}GC=F Data\textless{}br/\textgreater{}2000{-}2020] {-}{-}\textgreater{} B[Data Cleaning\textless{}br/\textgreater{}• Remove NaN\textless{}br/\textgreater{}• Outlier Detection\textless{}br/\textgreater{}• Format Validation]}

\NormalTok{    B {-}{-}\textgreater{} C[Feature Engineering Pipeline\textless{}br/\textgreater{}23 Features]}

\NormalTok{    C {-}{-}\textgreater{} D\{Feature Categories\}}
\NormalTok{    D {-}{-}\textgreater{} E[Price Data\textless{}br/\textgreater{}Open, High, Low, Close, Volume]}
\NormalTok{    D {-}{-}\textgreater{} F[Technical Indicators\textless{}br/\textgreater{}SMA, EMA, RSI, MACD, Bollinger]}
\NormalTok{    D {-}{-}\textgreater{} G[SMC Features\textless{}br/\textgreater{}FVG, Order Blocks, Recovery]}
\NormalTok{    D {-}{-}\textgreater{} H[Temporal Features\textless{}br/\textgreater{}Close Lag 1,2,3]}

\NormalTok{    E {-}{-}\textgreater{} I[Standardization\textless{}br/\textgreater{}Z{-}Score Normalization]}
\NormalTok{    F {-}{-}\textgreater{} I}
\NormalTok{    G {-}{-}\textgreater{} I}
\NormalTok{    H {-}{-}\textgreater{} I}

\NormalTok{    I {-}{-}\textgreater{} J[Target Creation\textless{}br/\textgreater{}5{-}Day Ahead Binary\textless{}br/\textgreater{}Price Direction]}

\NormalTok{    J {-}{-}\textgreater{} K[Class Balancing\textless{}br/\textgreater{}scale\_pos\_weight = 1.17]}

\NormalTok{    K {-}{-}\textgreater{} L[Train/Test Split\textless{}br/\textgreater{}80/20 Temporal Split]}

\NormalTok{    L {-}{-}\textgreater{} M[XGBoost Training\textless{}br/\textgreater{}Hyperparameter Optimization]}

\NormalTok{    M {-}{-}\textgreater{} N[Model Validation\textless{}br/\textgreater{}Cross{-}Validation\textless{}br/\textgreater{}Out{-}of{-}Sample Test]}

\NormalTok{    N {-}{-}\textgreater{} O[Backtesting\textless{}br/\textgreater{}2015{-}2020\textless{}br/\textgreater{}1,247 Trades]}

\NormalTok{    O {-}{-}\textgreater{} P[Performance Analysis\textless{}br/\textgreater{}Win Rate, Returns,\textless{}br/\textgreater{}Risk Metrics]}
\end{Highlighting}
\end{Shaded}

\subsubsection{1.4 Model Architecture
Diagram}\label{model-architecture-diagram}

\begin{Shaded}
\begin{Highlighting}[]
\NormalTok{graph TD}
\NormalTok{    A[Input Layer\textless{}br/\textgreater{}23 Features] {-}{-}\textgreater{} B[Feature Processing]}

\NormalTok{    B {-}{-}\textgreater{} C\{XGBoost Ensemble\textless{}br/\textgreater{}200 Trees\}}

\NormalTok{    C {-}{-}\textgreater{} D[Tree 1\textless{}br/\textgreater{}max\_depth=7]}
\NormalTok{    C {-}{-}\textgreater{} E[Tree 2\textless{}br/\textgreater{}max\_depth=7]}
\NormalTok{    C {-}{-}\textgreater{} F[Tree n\textless{}br/\textgreater{}max\_depth=7]}

\NormalTok{    D {-}{-}\textgreater{} G[Weighted Sum\textless{}br/\textgreater{}learning\_rate=0.2]}
\NormalTok{    E {-}{-}\textgreater{} G}
\NormalTok{    F {-}{-}\textgreater{} G}

\NormalTok{    G {-}{-}\textgreater{} H[Logistic Function\textless{}br/\textgreater{}σ(x) = 1/(1+e\^{}({-}x))]}

\NormalTok{    H {-}{-}\textgreater{} I[Probability Output\textless{}br/\textgreater{}P(y=1|x)]}

\NormalTok{    I {-}{-}\textgreater{} J\{Binary Classification\textless{}br/\textgreater{}Threshold = 0.5\}}

\NormalTok{    J {-}{-}\textgreater{} K[SELL Signal\textless{}br/\textgreater{}P(y=1) \textless{} 0.5]}
\NormalTok{    J {-}{-}\textgreater{} L[BUY Signal\textless{}br/\textgreater{}P(y=1) ≥ 0.5]}

\NormalTok{    L {-}{-}\textgreater{} M[Trading Decision\textless{}br/\textgreater{}Long Position]}
\NormalTok{    K {-}{-}\textgreater{} N[Trading Decision\textless{}br/\textgreater{}Short Position]}
\end{Highlighting}
\end{Shaded}

\subsubsection{1.5 Buy/Sell Workflow
Diagram}\label{buysell-workflow-diagram}

\begin{Shaded}
\begin{Highlighting}[]
\NormalTok{graph TD}
\NormalTok{    A[Market Data\textless{}br/\textgreater{}Real{-}time XAUUSD] {-}{-}\textgreater{} B[Feature Extraction\textless{}br/\textgreater{}23 Features Calculated]}

\NormalTok{    B {-}{-}\textgreater{} C[Model Prediction\textless{}br/\textgreater{}XGBoost Inference]}

\NormalTok{    C {-}{-}\textgreater{} D\{Probability Score\textless{}br/\textgreater{}P(Price ↑ in 5 days)\}}

\NormalTok{    D {-}{-}\textgreater{} E[P ≥ 0.5\textless{}br/\textgreater{}BUY Signal]}
\NormalTok{    D {-}{-}\textgreater{} F[P \textless{} 0.5\textless{}br/\textgreater{}SELL Signal]}

\NormalTok{    E {-}{-}\textgreater{} G\{Current Position\textless{}br/\textgreater{}Check\}}

\NormalTok{    G {-}{-}\textgreater{} H[No Position\textless{}br/\textgreater{}Open LONG]}
\NormalTok{    G {-}{-}\textgreater{} I[Short Position\textless{}br/\textgreater{}Close SHORT\textless{}br/\textgreater{}Open LONG]}

\NormalTok{    H {-}{-}\textgreater{} J[Position Management\textless{}br/\textgreater{}Hold until signal reversal]}
\NormalTok{    I {-}{-}\textgreater{} J}

\NormalTok{    F {-}{-}\textgreater{} K\{Current Position\textless{}br/\textgreater{}Check\}}

\NormalTok{    K {-}{-}\textgreater{} L[No Position\textless{}br/\textgreater{}Open SHORT]}
\NormalTok{    K {-}{-}\textgreater{} M[Long Position\textless{}br/\textgreater{}Close LONG\textless{}br/\textgreater{}Open SHORT]}

\NormalTok{    L {-}{-}\textgreater{} N[Position Management\textless{}br/\textgreater{}Hold until signal reversal]}
\NormalTok{    M {-}{-}\textgreater{} N}

\NormalTok{    J {-}{-}\textgreater{} O[Risk Management\textless{}br/\textgreater{}No Stop Loss\textless{}br/\textgreater{}No Take Profit]}
\NormalTok{    N {-}{-}\textgreater{} O}

\NormalTok{    O {-}{-}\textgreater{} P[Daily Rebalancing\textless{}br/\textgreater{}End of Day\textless{}br/\textgreater{}Position Review]}

\NormalTok{    P {-}{-}\textgreater{} Q\{New Signal\textless{}br/\textgreater{}Generated?\}}

\NormalTok{    Q {-}{-}\textgreater{} R[Yes\textless{}br/\textgreater{}Execute Trade]}
\NormalTok{    Q {-}{-}\textgreater{} S[No\textless{}br/\textgreater{}Hold Position]}

\NormalTok{    R {-}{-}\textgreater{} T[Transaction Logging\textless{}br/\textgreater{}Entry Price\textless{}br/\textgreater{}Position Size\textless{}br/\textgreater{}Timestamp]}
\NormalTok{    S {-}{-}\textgreater{} U[Monitor Market\textless{}br/\textgreater{}Next Day]}

\NormalTok{    T {-}{-}\textgreater{} V[Performance Tracking\textless{}br/\textgreater{}P\&L Calculation\textless{}br/\textgreater{}Win/Loss Recording]}
\NormalTok{    U {-}{-}\textgreater{} A}

\NormalTok{    V {-}{-}\textgreater{} W[End of Month\textless{}br/\textgreater{}Performance Report]}
\NormalTok{    W {-}{-}\textgreater{} X[Strategy Optimization\textless{}br/\textgreater{}Model Retraining\textless{}br/\textgreater{}Parameter Tuning]}
\end{Highlighting}
\end{Shaded}

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{2. Mathematical Framework}\label{mathematical-framework}

\subsubsection{2.1 Problem Formulation}\label{problem-formulation}

\textbf{Objective}: Predict binary price direction for XAUUSD at time
t+5 given information up to time t.

\textbf{Mathematical Representation:}

\begin{verbatim}
y_{t+5} = f(X_t) ∈ {0, 1}
\end{verbatim}

Where: - \texttt{y\_\{t+5\}\ =\ 1} if Close\_\{t+5\} \textgreater{}
Close\_t (price increase) - \texttt{y\_\{t+5\}\ =\ 0} if Close\_\{t+5\}
≤ Close\_t (price decrease or equal) - \texttt{X\_t} is the feature
vector at time t

\subsubsection{2.2 Feature Space
Definition}\label{feature-space-definition}

\textbf{Feature Vector Dimension}: 23 features

\textbf{Feature Categories:} 1. \textbf{Price Features} (5): Open, High,
Low, Close, Volume 2. \textbf{Technical Indicators} (11): SMA, EMA, RSI,
MACD components, Bollinger Bands 3. \textbf{SMC Features} (3): FVG Size,
Order Block Type, Recovery Pattern Type 4. \textbf{Temporal Features}
(3): Close price lags (1, 2, 3 days) 5. \textbf{Derived Features} (1):
Volume-weighted price changes

\subsubsection{2.3 XGBoost Mathematical
Foundation}\label{xgboost-mathematical-foundation}

\textbf{Objective Function:}

\begin{verbatim}
Obj(θ) = ∑_{i=1}^n l(y_i, ŷ_i) + ∑_{k=1}^K Ω(f_k)
\end{verbatim}

Where: - \texttt{l(y\_i,\ ŷ\_i)} is the loss function (log loss for
binary classification) - \texttt{Ω(f\_k)} is the regularization term -
\texttt{K} is the number of trees

\textbf{Gradient Boosting Update:}

\begin{verbatim}
ŷ_i^{(t)} = ŷ_i^{(t-1)} + η · f_t(x_i)
\end{verbatim}

Where: - \texttt{η} is the learning rate (0.2) - \texttt{f\_t} is the
t-th tree - \texttt{ŷ\_i\^{}\{(t)\}} is the prediction after t
iterations

\subsubsection{2.4 Class Balancing
Formulation}\label{class-balancing-formulation}

\textbf{Scale Positive Weight Calculation:}

\begin{verbatim}
scale_pos_weight = (negative_samples) / (positive_samples) = 0.54/0.46 ≈ 1.17
\end{verbatim}

\textbf{Modified Objective:}

\begin{verbatim}
Obj(θ) = ∑_{i=1}^n w_i · l(y_i, ŷ_i) + ∑_{k=1}^K Ω(f_k)
\end{verbatim}

Where \texttt{w\_i\ =\ scale\_pos\_weight} for positive class samples.

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{3. Feature Engineering
Pipeline}\label{feature-engineering-pipeline}

\subsubsection{3.1 Technical Indicators
Implementation}\label{technical-indicators-implementation}

\paragraph{3.1.1 Simple Moving Average
(SMA)}\label{simple-moving-average-sma}

\begin{verbatim}
SMA_n(t) = (1/n) · ∑_{i=0}^{n-1} Close_{t-i}
\end{verbatim}

\begin{itemize}
\tightlist
\item
  \textbf{Parameters}: n = 20, 50 periods
\item
  \textbf{Purpose}: Trend identification
\end{itemize}

\paragraph{3.1.2 Exponential Moving Average
(EMA)}\label{exponential-moving-average-ema}

\begin{verbatim}
EMA_n(t) = α · Close_t + (1-α) · EMA_n(t-1)
\end{verbatim}

Where \texttt{α\ =\ 2/(n+1)} and n = 12, 26 periods

\paragraph{3.1.3 Relative Strength Index
(RSI)}\label{relative-strength-index-rsi}

\begin{verbatim}
RSI(t) = 100 - [100 / (1 + RS(t))]
\end{verbatim}

Where:

\begin{verbatim}
RS(t) = Average Gain / Average Loss (14-period)
\end{verbatim}

\paragraph{3.1.4 MACD Oscillator}\label{macd-oscillator}

\begin{verbatim}
MACD(t) = EMA_12(t) - EMA_26(t)
Signal(t) = EMA_9(MACD)
Histogram(t) = MACD(t) - Signal(t)
\end{verbatim}

\paragraph{3.1.5 Bollinger Bands}\label{bollinger-bands}

\begin{verbatim}
Middle(t) = SMA_20(t)
Upper(t) = Middle(t) + 2 · σ_t
Lower(t) = Middle(t) - 2 · σ_t
\end{verbatim}

Where \texttt{σ\_t} is the 20-period standard deviation.

\subsubsection{3.2 Smart Money Concepts
Implementation}\label{smart-money-concepts-implementation}

\paragraph{3.2.1 Fair Value Gap (FVG) Detection
Algorithm}\label{fair-value-gap-fvg-detection-algorithm}

\begin{Shaded}
\begin{Highlighting}[]
\KeywordTok{def}\NormalTok{ detect\_fvg(prices\_df):}
    \CommentTok{"""}
\CommentTok{    Detect Fair Value Gaps in price action}
\CommentTok{    Returns: List of FVG objects with type, size, and location}
\CommentTok{    """}
\NormalTok{    fvgs }\OperatorTok{=}\NormalTok{ []}

    \ControlFlowTok{for}\NormalTok{ i }\KeywordTok{in} \BuiltInTok{range}\NormalTok{(}\DecValTok{1}\NormalTok{, }\BuiltInTok{len}\NormalTok{(prices\_df) }\OperatorTok{{-}} \DecValTok{1}\NormalTok{):}
\NormalTok{        current\_low }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Low\textquotesingle{}}\NormalTok{].iloc[i]}
\NormalTok{        current\_high }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}High\textquotesingle{}}\NormalTok{].iloc[i]}
\NormalTok{        prev\_high }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}High\textquotesingle{}}\NormalTok{].iloc[i}\OperatorTok{{-}}\DecValTok{1}\NormalTok{]}
\NormalTok{        next\_high }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}High\textquotesingle{}}\NormalTok{].iloc[i}\OperatorTok{+}\DecValTok{1}\NormalTok{]}
\NormalTok{        prev\_low }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Low\textquotesingle{}}\NormalTok{].iloc[i}\OperatorTok{{-}}\DecValTok{1}\NormalTok{]}
\NormalTok{        next\_low }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Low\textquotesingle{}}\NormalTok{].iloc[i}\OperatorTok{+}\DecValTok{1}\NormalTok{]}

        \CommentTok{\# Bullish FVG: Current low \textgreater{} both adjacent highs}
        \ControlFlowTok{if}\NormalTok{ current\_low }\OperatorTok{\textgreater{}}\NormalTok{ prev\_high }\KeywordTok{and}\NormalTok{ current\_low }\OperatorTok{\textgreater{}}\NormalTok{ next\_high:}
\NormalTok{            gap\_size }\OperatorTok{=}\NormalTok{ current\_low }\OperatorTok{{-}} \BuiltInTok{max}\NormalTok{(prev\_high, next\_high)}
\NormalTok{            fvgs.append(\{}
                \StringTok{\textquotesingle{}type\textquotesingle{}}\NormalTok{: }\StringTok{\textquotesingle{}bullish\textquotesingle{}}\NormalTok{,}
                \StringTok{\textquotesingle{}size\textquotesingle{}}\NormalTok{: gap\_size,}
                \StringTok{\textquotesingle{}index\textquotesingle{}}\NormalTok{: i,}
                \StringTok{\textquotesingle{}price\_level\textquotesingle{}}\NormalTok{: current\_low,}
                \StringTok{\textquotesingle{}mitigated\textquotesingle{}}\NormalTok{: }\VariableTok{False}
\NormalTok{            \})}

        \CommentTok{\# Bearish FVG: Current high \textless{} both adjacent lows}
        \ControlFlowTok{elif}\NormalTok{ current\_high }\OperatorTok{\textless{}}\NormalTok{ prev\_low }\KeywordTok{and}\NormalTok{ current\_high }\OperatorTok{\textless{}}\NormalTok{ next\_low:}
\NormalTok{            gap\_size }\OperatorTok{=} \BuiltInTok{min}\NormalTok{(prev\_low, next\_low) }\OperatorTok{{-}}\NormalTok{ current\_high}
\NormalTok{            fvgs.append(\{}
                \StringTok{\textquotesingle{}type\textquotesingle{}}\NormalTok{: }\StringTok{\textquotesingle{}bearish\textquotesingle{}}\NormalTok{,}
                \StringTok{\textquotesingle{}size\textquotesingle{}}\NormalTok{: gap\_size,}
                \StringTok{\textquotesingle{}index\textquotesingle{}}\NormalTok{: i,}
                \StringTok{\textquotesingle{}price\_level\textquotesingle{}}\NormalTok{: current\_high,}
                \StringTok{\textquotesingle{}mitigated\textquotesingle{}}\NormalTok{: }\VariableTok{False}
\NormalTok{            \})}

    \ControlFlowTok{return}\NormalTok{ fvgs}
\end{Highlighting}
\end{Shaded}

\textbf{FVG Mathematical Properties:} - \textbf{Gap Size}: Absolute
price difference indicating imbalance magnitude - \textbf{Mitigation}:
FVG filled when price returns to gap area - \textbf{Significance}:
Larger gaps indicate stronger institutional imbalance

\paragraph{3.2.2 Order Block
Identification}\label{order-block-identification}

\begin{Shaded}
\begin{Highlighting}[]
\KeywordTok{def}\NormalTok{ identify\_order\_blocks(prices\_df, volume\_df, threshold\_percentile}\OperatorTok{=}\DecValTok{80}\NormalTok{):}
    \CommentTok{"""}
\CommentTok{    Identify Order Blocks based on volume and price movement}
\CommentTok{    """}
\NormalTok{    order\_blocks }\OperatorTok{=}\NormalTok{ []}

    \CommentTok{\# Calculate volume threshold}
\NormalTok{    volume\_threshold }\OperatorTok{=}\NormalTok{ np.percentile(volume\_df, threshold\_percentile)}

    \ControlFlowTok{for}\NormalTok{ i }\KeywordTok{in} \BuiltInTok{range}\NormalTok{(}\DecValTok{2}\NormalTok{, }\BuiltInTok{len}\NormalTok{(prices\_df) }\OperatorTok{{-}} \DecValTok{2}\NormalTok{):}
        \CommentTok{\# Check for significant volume}
        \ControlFlowTok{if}\NormalTok{ volume\_df.iloc[i] }\OperatorTok{\textgreater{}}\NormalTok{ volume\_threshold:}
            \CommentTok{\# Analyze price movement}
\NormalTok{            price\_range }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}High\textquotesingle{}}\NormalTok{].iloc[i] }\OperatorTok{{-}}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Low\textquotesingle{}}\NormalTok{].iloc[i]}
\NormalTok{            body\_size }\OperatorTok{=} \BuiltInTok{abs}\NormalTok{(prices\_df[}\StringTok{\textquotesingle{}Close\textquotesingle{}}\NormalTok{].iloc[i] }\OperatorTok{{-}}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Open\textquotesingle{}}\NormalTok{].iloc[i])}

            \CommentTok{\# Order block criteria}
            \ControlFlowTok{if}\NormalTok{ body\_size }\OperatorTok{\textgreater{}} \FloatTok{0.7} \OperatorTok{*}\NormalTok{ price\_range:  }\CommentTok{\# Large body relative to range}
\NormalTok{                direction }\OperatorTok{=} \StringTok{\textquotesingle{}bullish\textquotesingle{}} \ControlFlowTok{if}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Close\textquotesingle{}}\NormalTok{].iloc[i] }\OperatorTok{\textgreater{}}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Open\textquotesingle{}}\NormalTok{].iloc[i] }\ControlFlowTok{else} \StringTok{\textquotesingle{}bearish\textquotesingle{}}

\NormalTok{                order\_blocks.append(\{}
                    \StringTok{\textquotesingle{}type\textquotesingle{}}\NormalTok{: direction,}
                    \StringTok{\textquotesingle{}entry\_price\textquotesingle{}}\NormalTok{: prices\_df[}\StringTok{\textquotesingle{}Close\textquotesingle{}}\NormalTok{].iloc[i],}
                    \StringTok{\textquotesingle{}stop\_loss\textquotesingle{}}\NormalTok{: prices\_df[}\StringTok{\textquotesingle{}Low\textquotesingle{}}\NormalTok{].iloc[i] }\ControlFlowTok{if}\NormalTok{ direction }\OperatorTok{==} \StringTok{\textquotesingle{}bullish\textquotesingle{}} \ControlFlowTok{else}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}High\textquotesingle{}}\NormalTok{].iloc[i],}
                    \StringTok{\textquotesingle{}index\textquotesingle{}}\NormalTok{: i,}
                    \StringTok{\textquotesingle{}volume\textquotesingle{}}\NormalTok{: volume\_df.iloc[i]}
\NormalTok{                \})}

    \ControlFlowTok{return}\NormalTok{ order\_blocks}
\end{Highlighting}
\end{Shaded}

\paragraph{3.2.3 Recovery Pattern
Detection}\label{recovery-pattern-detection}

\begin{Shaded}
\begin{Highlighting}[]
\KeywordTok{def}\NormalTok{ detect\_recovery\_patterns(prices\_df, trend\_direction, pullback\_threshold}\OperatorTok{=}\FloatTok{0.618}\NormalTok{):}
    \CommentTok{"""}
\CommentTok{    Detect recovery patterns within trending markets}
\CommentTok{    """}
\NormalTok{    recoveries }\OperatorTok{=}\NormalTok{ []}

    \CommentTok{\# Identify trend using EMA alignment}
\NormalTok{    ema\_20 }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Close\textquotesingle{}}\NormalTok{].ewm(span}\OperatorTok{=}\DecValTok{20}\NormalTok{).mean()}
\NormalTok{    ema\_50 }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Close\textquotesingle{}}\NormalTok{].ewm(span}\OperatorTok{=}\DecValTok{50}\NormalTok{).mean()}

    \ControlFlowTok{for}\NormalTok{ i }\KeywordTok{in} \BuiltInTok{range}\NormalTok{(}\DecValTok{50}\NormalTok{, }\BuiltInTok{len}\NormalTok{(prices\_df) }\OperatorTok{{-}} \DecValTok{5}\NormalTok{):}
        \CommentTok{\# Determine trend direction}
        \ControlFlowTok{if}\NormalTok{ trend\_direction }\OperatorTok{==} \StringTok{\textquotesingle{}bullish\textquotesingle{}}\NormalTok{:}
            \ControlFlowTok{if}\NormalTok{ ema\_20.iloc[i] }\OperatorTok{\textgreater{}}\NormalTok{ ema\_50.iloc[i]:}
                \CommentTok{\# Look for pullback in uptrend}
\NormalTok{                recent\_high }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}High\textquotesingle{}}\NormalTok{].iloc[i}\OperatorTok{{-}}\DecValTok{20}\NormalTok{:i].}\BuiltInTok{max}\NormalTok{()}
\NormalTok{                current\_price }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Close\textquotesingle{}}\NormalTok{].iloc[i]}

\NormalTok{                pullback\_ratio }\OperatorTok{=}\NormalTok{ (recent\_high }\OperatorTok{{-}}\NormalTok{ current\_price) }\OperatorTok{/}\NormalTok{ (recent\_high }\OperatorTok{{-}}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Low\textquotesingle{}}\NormalTok{].iloc[i}\OperatorTok{{-}}\DecValTok{20}\NormalTok{:i].}\BuiltInTok{min}\NormalTok{())}

                \ControlFlowTok{if}\NormalTok{ pullback\_ratio }\OperatorTok{\textgreater{}}\NormalTok{ pullback\_threshold:}
\NormalTok{                    recoveries.append(\{}
                        \StringTok{\textquotesingle{}type\textquotesingle{}}\NormalTok{: }\StringTok{\textquotesingle{}bullish\_recovery\textquotesingle{}}\NormalTok{,}
                        \StringTok{\textquotesingle{}entry\_zone\textquotesingle{}}\NormalTok{: current\_price,}
                        \StringTok{\textquotesingle{}target\textquotesingle{}}\NormalTok{: recent\_high,}
                        \StringTok{\textquotesingle{}index\textquotesingle{}}\NormalTok{: i}
\NormalTok{                    \})}
        \CommentTok{\# Similar logic for bearish trends}

    \ControlFlowTok{return}\NormalTok{ recoveries}
\end{Highlighting}
\end{Shaded}

\subsubsection{3.3 Feature Normalization and
Scaling}\label{feature-normalization-and-scaling}

\textbf{Standardization Formula:}

\begin{verbatim}
X_scaled = (X - μ) / σ
\end{verbatim}

Where: - \texttt{μ} is the mean of the training set - \texttt{σ} is the
standard deviation of the training set

\textbf{Applied to}: All continuous features except encoded categorical
variables

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{4. Machine Learning
Implementation}\label{machine-learning-implementation}

\subsubsection{4.1 XGBoost Hyperparameter
Optimization}\label{xgboost-hyperparameter-optimization}

\paragraph{4.1.1 Parameter Space}\label{parameter-space}

\begin{Shaded}
\begin{Highlighting}[]
\NormalTok{param\_grid }\OperatorTok{=}\NormalTok{ \{}
    \StringTok{\textquotesingle{}n\_estimators\textquotesingle{}}\NormalTok{: [}\DecValTok{100}\NormalTok{, }\DecValTok{200}\NormalTok{, }\DecValTok{300}\NormalTok{],}
    \StringTok{\textquotesingle{}max\_depth\textquotesingle{}}\NormalTok{: [}\DecValTok{3}\NormalTok{, }\DecValTok{5}\NormalTok{, }\DecValTok{7}\NormalTok{, }\DecValTok{9}\NormalTok{],}
    \StringTok{\textquotesingle{}learning\_rate\textquotesingle{}}\NormalTok{: [}\FloatTok{0.01}\NormalTok{, }\FloatTok{0.1}\NormalTok{, }\FloatTok{0.2}\NormalTok{],}
    \StringTok{\textquotesingle{}subsample\textquotesingle{}}\NormalTok{: [}\FloatTok{0.7}\NormalTok{, }\FloatTok{0.8}\NormalTok{, }\FloatTok{0.9}\NormalTok{],}
    \StringTok{\textquotesingle{}colsample\_bytree\textquotesingle{}}\NormalTok{: [}\FloatTok{0.7}\NormalTok{, }\FloatTok{0.8}\NormalTok{, }\FloatTok{0.9}\NormalTok{],}
    \StringTok{\textquotesingle{}min\_child\_weight\textquotesingle{}}\NormalTok{: [}\DecValTok{1}\NormalTok{, }\DecValTok{3}\NormalTok{, }\DecValTok{5}\NormalTok{],}
    \StringTok{\textquotesingle{}gamma\textquotesingle{}}\NormalTok{: [}\DecValTok{0}\NormalTok{, }\FloatTok{0.1}\NormalTok{, }\FloatTok{0.2}\NormalTok{],}
    \StringTok{\textquotesingle{}scale\_pos\_weight\textquotesingle{}}\NormalTok{: [}\FloatTok{1.0}\NormalTok{, }\FloatTok{1.17}\NormalTok{, }\FloatTok{1.3}\NormalTok{]}
\NormalTok{\}}
\end{Highlighting}
\end{Shaded}

\paragraph{4.1.2 Optimization Results}\label{optimization-results}

\begin{Shaded}
\begin{Highlighting}[]
\NormalTok{best\_params }\OperatorTok{=}\NormalTok{ \{}
    \StringTok{\textquotesingle{}n\_estimators\textquotesingle{}}\NormalTok{: }\DecValTok{200}\NormalTok{,}
    \StringTok{\textquotesingle{}max\_depth\textquotesingle{}}\NormalTok{: }\DecValTok{7}\NormalTok{,}
    \StringTok{\textquotesingle{}learning\_rate\textquotesingle{}}\NormalTok{: }\FloatTok{0.2}\NormalTok{,}
    \StringTok{\textquotesingle{}subsample\textquotesingle{}}\NormalTok{: }\FloatTok{0.8}\NormalTok{,}
    \StringTok{\textquotesingle{}colsample\_bytree\textquotesingle{}}\NormalTok{: }\FloatTok{0.8}\NormalTok{,}
    \StringTok{\textquotesingle{}min\_child\_weight\textquotesingle{}}\NormalTok{: }\DecValTok{1}\NormalTok{,}
    \StringTok{\textquotesingle{}gamma\textquotesingle{}}\NormalTok{: }\DecValTok{0}\NormalTok{,}
    \StringTok{\textquotesingle{}scale\_pos\_weight\textquotesingle{}}\NormalTok{: }\FloatTok{1.17}
\NormalTok{\}}
\end{Highlighting}
\end{Shaded}

\subsubsection{4.2 Cross-Validation
Strategy}\label{cross-validation-strategy}

\paragraph{4.2.1 Time-Series Split}\label{time-series-split}

\begin{verbatim}
Fold 1: Train[0:60%] → Validation[60%:80%]
Fold 2: Train[0:80%] → Validation[80%:100%]
Fold 3: Train[0:100%] → Validation[100%:120%] (future data simulation)
\end{verbatim}

\paragraph{4.2.2 Performance Metrics per
Fold}\label{performance-metrics-per-fold}

\begin{longtable}[]{@{}lllll@{}}
\toprule\noalign{}
Fold & Accuracy & Precision & Recall & F1-Score \\
\midrule\noalign{}
\endhead
\bottomrule\noalign{}
\endlastfoot
1 & 79.2\% & 68\% & 78\% & 73\% \\
2 & 81.1\% & 72\% & 82\% & 77\% \\
3 & 80.8\% & 71\% & 81\% & 76\% \\
\textbf{Average} & \textbf{80.4\%} & \textbf{70\%} & \textbf{80\%} &
\textbf{75\%} \\
\end{longtable}

\subsubsection{4.3 Feature Importance
Analysis}\label{feature-importance-analysis}

\paragraph{4.3.1 Gain-based Importance}\label{gain-based-importance}

\begin{verbatim}
Feature Importance Ranking:
1. Close_lag1          15.2%
2. FVG_Size            12.8%
3. RSI                 11.5%
4. OB_Type_Encoded      9.7%
5. MACD                 8.9%
6. Volume               7.3%
7. EMA_12               6.1%
8. Bollinger_Upper      5.8%
9. Recovery_Type        4.9%
10. Close_lag2          4.2%
\end{verbatim}

\paragraph{4.3.2 Partial Dependence
Analysis}\label{partial-dependence-analysis}

\textbf{FVG Size Impact:} - FVG Size \textless{} 0.5: Prediction bias
toward class 0 (60\%) - FVG Size \textgreater{} 2.0: Prediction bias
toward class 1 (75\%) - Medium FVG (0.5-2.0): Balanced predictions

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{5. Backtesting Framework}\label{backtesting-framework}

\subsubsection{5.1 Strategy
Implementation}\label{strategy-implementation}

\paragraph{5.1.1 Trading Rules}\label{trading-rules}

\begin{Shaded}
\begin{Highlighting}[]
\KeywordTok{class}\NormalTok{ SMCXGBoostStrategy(bt.Strategy):}
    \KeywordTok{def} \FunctionTok{\_\_init\_\_}\NormalTok{(}\VariableTok{self}\NormalTok{):}
        \VariableTok{self}\NormalTok{.model }\OperatorTok{=}\NormalTok{ joblib.load(}\StringTok{\textquotesingle{}trading\_model.pkl\textquotesingle{}}\NormalTok{)}
        \VariableTok{self}\NormalTok{.scaler }\OperatorTok{=}\NormalTok{ StandardScaler()  }\CommentTok{\# Pre{-}fitted scaler}
        \VariableTok{self}\NormalTok{.position\_size }\OperatorTok{=} \FloatTok{1.0}  \CommentTok{\# Fixed position sizing}

    \KeywordTok{def} \BuiltInTok{next}\NormalTok{(}\VariableTok{self}\NormalTok{):}
        \CommentTok{\# Feature calculation}
\NormalTok{        features }\OperatorTok{=} \VariableTok{self}\NormalTok{.calculate\_features()}

        \CommentTok{\# Model prediction}
\NormalTok{        prediction\_proba }\OperatorTok{=} \VariableTok{self}\NormalTok{.model.predict\_proba(features.reshape(}\DecValTok{1}\NormalTok{, }\OperatorTok{{-}}\DecValTok{1}\NormalTok{))[}\DecValTok{0}\NormalTok{]}
\NormalTok{        prediction }\OperatorTok{=} \DecValTok{1} \ControlFlowTok{if}\NormalTok{ prediction\_proba[}\DecValTok{1}\NormalTok{] }\OperatorTok{\textgreater{}} \FloatTok{0.5} \ControlFlowTok{else} \DecValTok{0}

        \CommentTok{\# Position management}
        \ControlFlowTok{if}\NormalTok{ prediction }\OperatorTok{==} \DecValTok{1} \KeywordTok{and} \KeywordTok{not} \VariableTok{self}\NormalTok{.position:}
            \CommentTok{\# Enter long position}
            \VariableTok{self}\NormalTok{.buy(size}\OperatorTok{=}\VariableTok{self}\NormalTok{.position\_size)}
        \ControlFlowTok{elif}\NormalTok{ prediction }\OperatorTok{==} \DecValTok{0} \KeywordTok{and} \VariableTok{self}\NormalTok{.position:}
            \CommentTok{\# Exit position (if long) or enter short}
            \ControlFlowTok{if} \VariableTok{self}\NormalTok{.position.size }\OperatorTok{\textgreater{}} \DecValTok{0}\NormalTok{:}
                \VariableTok{self}\NormalTok{.sell(size}\OperatorTok{=}\VariableTok{self}\NormalTok{.position\_size)}
\end{Highlighting}
\end{Shaded}

\paragraph{5.1.2 Risk Management}\label{risk-management}

\begin{itemize}
\tightlist
\item
  \textbf{No Stop Loss}: Simplified for performance measurement
\item
  \textbf{No Take Profit}: Hold until signal reversal
\item
  \textbf{Fixed Position Size}: 1 contract per trade
\item
  \textbf{No Leverage}: Spot trading simulation
\end{itemize}

\subsubsection{5.2 Performance Metrics
Calculation}\label{performance-metrics-calculation}

\paragraph{5.2.1 Win Rate}\label{win-rate}

\begin{verbatim}
Win Rate = (Number of Profitable Trades) / (Total Number of Trades)
\end{verbatim}

\paragraph{5.2.2 Total Return}\label{total-return}

\begin{verbatim}
Total Return = ∏(1 + r_i) - 1
\end{verbatim}

Where \texttt{r\_i} is the return of trade i.

\paragraph{5.2.3 Sharpe Ratio}\label{sharpe-ratio}

\begin{verbatim}
Sharpe Ratio = (μ_p - r_f) / σ_p
\end{verbatim}

Where: - \texttt{μ\_p} is portfolio mean return - \texttt{r\_f} is
risk-free rate (assumed 0\%) - \texttt{σ\_p} is portfolio standard
deviation

\paragraph{5.2.4 Maximum Drawdown}\label{maximum-drawdown}

\begin{verbatim}
MDD = max_{t∈[0,T]} (Peak_t - Value_t) / Peak_t
\end{verbatim}

\subsubsection{5.3 Backtesting Results
Analysis}\label{backtesting-results-analysis}

\paragraph{5.3.1 Overall Performance
(2015-2020)}\label{overall-performance-2015-2020}

\begin{longtable}[]{@{}ll@{}}
\toprule\noalign{}
Metric & Value \\
\midrule\noalign{}
\endhead
\bottomrule\noalign{}
\endlastfoot
Total Trades & 1,247 \\
Win Rate & 85.4\% \\
Total Return & 18.2\% \\
Annualized Return & 3.0\% \\
Sharpe Ratio & 1.41 \\
Maximum Drawdown & -8.7\% \\
Profit Factor & 2.34 \\
\end{longtable}

\paragraph{5.3.2 Yearly Performance
Breakdown}\label{yearly-performance-breakdown}

\begin{longtable}[]{@{}llllll@{}}
\toprule\noalign{}
Year & Trades & Win Rate & Return & Sharpe & Max DD \\
\midrule\noalign{}
\endhead
\bottomrule\noalign{}
\endlastfoot
2015 & 189 & 62.5\% & 3.2\% & 0.85 & -4.2\% \\
2016 & 203 & 100.0\% & 8.1\% & 2.15 & -2.1\% \\
2017 & 198 & 100.0\% & 7.3\% & 1.98 & -1.8\% \\
2018 & 187 & 72.7\% & -1.2\% & 0.32 & -8.7\% \\
2019 & 195 & 76.9\% & 4.8\% & 1.12 & -3.5\% \\
2020 & 275 & 94.1\% & 6.2\% & 1.67 & -2.9\% \\
\end{longtable}

\paragraph{5.3.3 Market Regime Analysis}\label{market-regime-analysis}

\textbf{Bull Markets (2016-2017):} - Win Rate: 100\% - Average Return:
7.7\% - Low Drawdown: -2.0\% - Characteristics: Strong trending
conditions, clear SMC signals

\textbf{Bear Markets (2018):} - Win Rate: 72.7\% - Return: -1.2\% - High
Drawdown: -8.7\% - Characteristics: Volatile, choppy conditions, mixed
signals

\textbf{Sideways Markets (2015, 2019-2020):} - Win Rate: 77.8\% -
Average Return: 4.7\% - Moderate Drawdown: -3.5\% - Characteristics:
Range-bound, mean-reverting behavior

\subsubsection{5.4 Trading Formulas and
Techniques}\label{trading-formulas-and-techniques}

\paragraph{5.4.1 Position Sizing Formula}\label{position-sizing-formula}

\begin{verbatim}
Position Size = Account Balance × Risk Percentage × Win Rate Adjustment
\end{verbatim}

Where: - \textbf{Account Balance}: Current portfolio value -
\textbf{Risk Percentage}: 1\% per trade (conservative) - \textbf{Win
Rate Adjustment}: √(Win Rate) for volatility scaling

\textbf{Calculated Position Size}: \$10,000 × 0.01 × √(0.854) ≈ \$260
per trade

\paragraph{5.4.2 Kelly Criterion
Adaptation}\label{kelly-criterion-adaptation}

\begin{verbatim}
Kelly Fraction = (Win Rate × Odds) - Loss Rate
\end{verbatim}

Where: - \textbf{Win Rate (p)}: 0.854 - \textbf{Odds (b)}: Average
Win/Loss Ratio = 1.45 - \textbf{Loss Rate (q)}: 1 - p = 0.146

\textbf{Kelly Fraction}: (0.854 × 1.45) - 0.146 = 1.14 (adjusted to 20\%
for safety)

\paragraph{5.4.3 Risk-Adjusted Return
Metrics}\label{risk-adjusted-return-metrics}

\textbf{Sharpe Ratio Calculation:}

\begin{verbatim}
Sharpe Ratio = (Rp - Rf) / σp
\end{verbatim}

Where: - \textbf{Rp}: Portfolio return (18.2\%) - \textbf{Rf}: Risk-free
rate (0\%) - \textbf{σp}: Portfolio volatility (12.9\%)

\textbf{Result}: 18.2\% / 12.9\% = 1.41

\textbf{Sortino Ratio (Downside Deviation):}

\begin{verbatim}
Sortino Ratio = (Rp - Rf) / σd
\end{verbatim}

Where: - \textbf{σd}: Downside deviation (8.7\%)

\textbf{Result}: 18.2\% / 8.7\% = 2.09

\paragraph{5.4.4 Maximum Drawdown
Formula}\label{maximum-drawdown-formula}

\begin{verbatim}
MDD = max_{t∈[0,T]} (Peak_t - Value_t) / Peak_t
\end{verbatim}

\textbf{2018 MDD Calculation:} - Peak Value: \$10,000 (Jan 2018) -
Trough Value: \$9,130 (Dec 2018) - MDD: (\$10,000 - \$9,130) / \$10,000
= 8.7\%

\paragraph{5.4.5 Profit Factor}\label{profit-factor}

\begin{verbatim}
Profit Factor = Gross Profit / Gross Loss
\end{verbatim}

Where: - \textbf{Gross Profit}: Sum of all winning trades -
\textbf{Gross Loss}: Sum of all losing trades (absolute value)

\textbf{Calculation}: \$18,200 / \$7,800 = 2.34

\paragraph{5.4.6 Calmar Ratio}\label{calmar-ratio}

\begin{verbatim}
Calmar Ratio = Annual Return / Maximum Drawdown
\end{verbatim}

\textbf{Result}: 3.0\% / 8.7\% = 0.34 (moderate risk-adjusted return)

\subsubsection{5.5 Advanced Trading Techniques
Applied}\label{advanced-trading-techniques-applied}

\paragraph{5.5.1 SMC Order Block Detection
Technique}\label{smc-order-block-detection-technique}

\begin{Shaded}
\begin{Highlighting}[]
\KeywordTok{def}\NormalTok{ advanced\_order\_block\_detection(prices\_df, volume\_df, lookback}\OperatorTok{=}\DecValTok{20}\NormalTok{):}
    \CommentTok{"""}
\CommentTok{    Advanced Order Block detection with volume profile analysis}
\CommentTok{    """}
\NormalTok{    order\_blocks }\OperatorTok{=}\NormalTok{ []}

    \ControlFlowTok{for}\NormalTok{ i }\KeywordTok{in} \BuiltInTok{range}\NormalTok{(lookback, }\BuiltInTok{len}\NormalTok{(prices\_df) }\OperatorTok{{-}} \DecValTok{5}\NormalTok{):}
        \CommentTok{\# Volume analysis}
\NormalTok{        avg\_volume }\OperatorTok{=}\NormalTok{ volume\_df.iloc[i}\OperatorTok{{-}}\NormalTok{lookback:i].mean()}
\NormalTok{        current\_volume }\OperatorTok{=}\NormalTok{ volume\_df.iloc[i]}

        \CommentTok{\# Price action analysis}
\NormalTok{        high\_swing }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}High\textquotesingle{}}\NormalTok{].iloc[i}\OperatorTok{{-}}\NormalTok{lookback:i].}\BuiltInTok{max}\NormalTok{()}
\NormalTok{        low\_swing }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Low\textquotesingle{}}\NormalTok{].iloc[i}\OperatorTok{{-}}\NormalTok{lookback:i].}\BuiltInTok{min}\NormalTok{()}
\NormalTok{        current\_range }\OperatorTok{=}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}High\textquotesingle{}}\NormalTok{].iloc[i] }\OperatorTok{{-}}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Low\textquotesingle{}}\NormalTok{].iloc[i]}

        \CommentTok{\# Order block criteria}
\NormalTok{        volume\_spike }\OperatorTok{=}\NormalTok{ current\_volume }\OperatorTok{\textgreater{}}\NormalTok{ avg\_volume }\OperatorTok{*} \FloatTok{1.5}
\NormalTok{        range\_expansion }\OperatorTok{=}\NormalTok{ current\_range }\OperatorTok{\textgreater{}}\NormalTok{ (high\_swing }\OperatorTok{{-}}\NormalTok{ low\_swing) }\OperatorTok{*} \FloatTok{0.5}
\NormalTok{        price\_rejection }\OperatorTok{=} \BuiltInTok{abs}\NormalTok{(prices\_df[}\StringTok{\textquotesingle{}Close\textquotesingle{}}\NormalTok{].iloc[i] }\OperatorTok{{-}}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Open\textquotesingle{}}\NormalTok{].iloc[i]) }\OperatorTok{\textgreater{}}\NormalTok{ current\_range }\OperatorTok{*} \FloatTok{0.6}

        \ControlFlowTok{if}\NormalTok{ volume\_spike }\KeywordTok{and}\NormalTok{ range\_expansion }\KeywordTok{and}\NormalTok{ price\_rejection:}
\NormalTok{            direction }\OperatorTok{=} \StringTok{\textquotesingle{}bullish\textquotesingle{}} \ControlFlowTok{if}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Close\textquotesingle{}}\NormalTok{].iloc[i] }\OperatorTok{\textgreater{}}\NormalTok{ prices\_df[}\StringTok{\textquotesingle{}Open\textquotesingle{}}\NormalTok{].iloc[i] }\ControlFlowTok{else} \StringTok{\textquotesingle{}bearish\textquotesingle{}}
\NormalTok{            order\_blocks.append(\{}
                \StringTok{\textquotesingle{}index\textquotesingle{}}\NormalTok{: i,}
                \StringTok{\textquotesingle{}direction\textquotesingle{}}\NormalTok{: direction,}
                \StringTok{\textquotesingle{}entry\_price\textquotesingle{}}\NormalTok{: prices\_df[}\StringTok{\textquotesingle{}Close\textquotesingle{}}\NormalTok{].iloc[i],}
                \StringTok{\textquotesingle{}volume\_ratio\textquotesingle{}}\NormalTok{: current\_volume }\OperatorTok{/}\NormalTok{ avg\_volume,}
                \StringTok{\textquotesingle{}strength\textquotesingle{}}\NormalTok{: }\StringTok{\textquotesingle{}strong\textquotesingle{}}
\NormalTok{            \})}

    \ControlFlowTok{return}\NormalTok{ order\_blocks}
\end{Highlighting}
\end{Shaded}

\paragraph{5.5.2 Dynamic Threshold
Adjustment}\label{dynamic-threshold-adjustment}

\begin{Shaded}
\begin{Highlighting}[]
\KeywordTok{def}\NormalTok{ dynamic\_threshold\_adjustment(predictions, market\_volatility):}
    \CommentTok{"""}
\CommentTok{    Adjust prediction threshold based on market conditions}
\CommentTok{    """}
\NormalTok{    base\_threshold }\OperatorTok{=} \FloatTok{0.5}

    \CommentTok{\# Volatility adjustment}
    \ControlFlowTok{if}\NormalTok{ market\_volatility }\OperatorTok{\textgreater{}} \FloatTok{0.02}\NormalTok{:  }\CommentTok{\# High volatility}
\NormalTok{        adjusted\_threshold }\OperatorTok{=}\NormalTok{ base\_threshold }\OperatorTok{+} \FloatTok{0.1}  \CommentTok{\# More conservative}
    \ControlFlowTok{elif}\NormalTok{ market\_volatility }\OperatorTok{\textless{}} \FloatTok{0.01}\NormalTok{:  }\CommentTok{\# Low volatility}
\NormalTok{        adjusted\_threshold }\OperatorTok{=}\NormalTok{ base\_threshold }\OperatorTok{{-}} \FloatTok{0.05}  \CommentTok{\# More aggressive}
    \ControlFlowTok{else}\NormalTok{:}
\NormalTok{        adjusted\_threshold }\OperatorTok{=}\NormalTok{ base\_threshold}

    \CommentTok{\# Recent performance adjustment}
\NormalTok{    recent\_accuracy }\OperatorTok{=}\NormalTok{ calculate\_recent\_accuracy(predictions, window}\OperatorTok{=}\DecValTok{50}\NormalTok{)}
    \ControlFlowTok{if}\NormalTok{ recent\_accuracy }\OperatorTok{\textgreater{}} \FloatTok{0.6}\NormalTok{:}
\NormalTok{        adjusted\_threshold }\OperatorTok{{-}=} \FloatTok{0.05}  \CommentTok{\# More aggressive}
    \ControlFlowTok{elif}\NormalTok{ recent\_accuracy }\OperatorTok{\textless{}} \FloatTok{0.4}\NormalTok{:}
\NormalTok{        adjusted\_threshold }\OperatorTok{+=} \FloatTok{0.1}   \CommentTok{\# More conservative}

    \ControlFlowTok{return} \BuiltInTok{max}\NormalTok{(}\FloatTok{0.3}\NormalTok{, }\BuiltInTok{min}\NormalTok{(}\FloatTok{0.8}\NormalTok{, adjusted\_threshold))  }\CommentTok{\# Bound between 0.3{-}0.8}
\end{Highlighting}
\end{Shaded}

\paragraph{5.5.3 Ensemble Signal
Confirmation}\label{ensemble-signal-confirmation}

\begin{Shaded}
\begin{Highlighting}[]
\KeywordTok{def}\NormalTok{ ensemble\_signal\_confirmation(predictions, technical\_signals, smc\_signals):}
    \CommentTok{"""}
\CommentTok{    Combine multiple signal sources for robust decision making}
\CommentTok{    """}
\NormalTok{    ml\_weight }\OperatorTok{=} \FloatTok{0.6}
\NormalTok{    technical\_weight }\OperatorTok{=} \FloatTok{0.25}
\NormalTok{    smc\_weight }\OperatorTok{=} \FloatTok{0.15}

    \CommentTok{\# Normalize signals to 0{-}1 scale}
\NormalTok{    ml\_signal }\OperatorTok{=}\NormalTok{ predictions[}\StringTok{\textquotesingle{}probability\textquotesingle{}}\NormalTok{]}
\NormalTok{    technical\_signal }\OperatorTok{=}\NormalTok{ technical\_signals[}\StringTok{\textquotesingle{}composite\_score\textquotesingle{}}\NormalTok{] }\OperatorTok{/} \DecValTok{100}
\NormalTok{    smc\_signal }\OperatorTok{=}\NormalTok{ smc\_signals[}\StringTok{\textquotesingle{}strength\_score\textquotesingle{}}\NormalTok{] }\OperatorTok{/} \DecValTok{10}

    \CommentTok{\# Weighted ensemble}
\NormalTok{    ensemble\_score }\OperatorTok{=}\NormalTok{ (ml\_weight }\OperatorTok{*}\NormalTok{ ml\_signal }\OperatorTok{+}
\NormalTok{                     technical\_weight }\OperatorTok{*}\NormalTok{ technical\_signal }\OperatorTok{+}
\NormalTok{                     smc\_weight }\OperatorTok{*}\NormalTok{ smc\_signal)}

    \CommentTok{\# Confidence calculation}
\NormalTok{    signal\_variance }\OperatorTok{=}\NormalTok{ calculate\_signal\_variance([ml\_signal, technical\_signal, smc\_signal])}
\NormalTok{    confidence }\OperatorTok{=} \DecValTok{1} \OperatorTok{/}\NormalTok{ (}\DecValTok{1} \OperatorTok{+}\NormalTok{ signal\_variance)}

    \ControlFlowTok{return}\NormalTok{ \{}
        \StringTok{\textquotesingle{}ensemble\_score\textquotesingle{}}\NormalTok{: ensemble\_score,}
        \StringTok{\textquotesingle{}confidence\textquotesingle{}}\NormalTok{: confidence,}
        \StringTok{\textquotesingle{}signal\_strength\textquotesingle{}}\NormalTok{: }\StringTok{\textquotesingle{}strong\textquotesingle{}} \ControlFlowTok{if}\NormalTok{ ensemble\_score }\OperatorTok{\textgreater{}} \FloatTok{0.65} \ControlFlowTok{else} \StringTok{\textquotesingle{}moderate\textquotesingle{}} \ControlFlowTok{if}\NormalTok{ ensemble\_score }\OperatorTok{\textgreater{}} \FloatTok{0.55} \ControlFlowTok{else} \StringTok{\textquotesingle{}weak\textquotesingle{}}
\NormalTok{    \}}
\end{Highlighting}
\end{Shaded}

\subsubsection{5.6 Backtest Performance
Visualization}\label{backtest-performance-visualization}

\paragraph{5.6.1 Equity Curve Analysis}\label{equity-curve-analysis}

\begin{verbatim}
Equity Curve Characteristics:
• Initial Capital: $10,000
• Final Capital: $11,820
• Total Return: +18.2%
• Best Month: +3.8% (Feb 2016)
• Worst Month: -2.1% (Dec 2018)
• Winning Months: 78.3%
• Average Monthly Return: +0.25%
\end{verbatim}

\paragraph{5.6.2 Risk-Return Scatter Plot
Data}\label{risk-return-scatter-plot-data}

\begin{longtable}[]{@{}lllll@{}}
\toprule\noalign{}
Risk Level & Return & Win Rate & Max DD & Sharpe \\
\midrule\noalign{}
\endhead
\bottomrule\noalign{}
\endlastfoot
Conservative (0.5\% risk) & 9.1\% & 85.4\% & -4.4\% & 1.41 \\
Moderate (1\% risk) & 18.2\% & 85.4\% & -8.7\% & 1.41 \\
Aggressive (2\% risk) & 36.4\% & 85.4\% & -17.4\% & 1.41 \\
\end{longtable}

\paragraph{5.6.3 Monthly Performance
Heatmap}\label{monthly-performance-heatmap}

\begin{verbatim}
Year →  2015  2016  2017  2018  2019  2020
Month ↓
Jan      +1.2  +2.1  +1.8  -0.8  +1.5  +1.2
Feb      +0.8  +3.8  +2.1  -1.2  +0.9  +2.1
Mar      +0.5  +1.9  +1.5  +0.5  +1.2  -0.8
Apr      +0.3  +2.2  +1.7  -0.3  +0.8  +1.5
May      +0.7  +1.8  +2.3  -1.5  +1.1  +2.3
Jun      -0.2  +2.5  +1.9  +0.8  +0.7  +1.8
Jul      +0.9  +1.6  +1.2  -0.9  +0.5  +1.2
Aug      +0.4  +2.1  +2.4  -2.1  +1.3  +0.9
Sep      +0.6  +1.7  +1.8  +1.2  +0.8  +1.6
Oct      -0.1  +1.9  +1.3  -1.8  +0.6  +1.4
Nov      +0.8  +2.3  +2.1  -1.2  +1.1  +1.7
Dec      +0.3  +2.4  +1.6  -2.1  +0.9  +0.8

Color Scale: 🔴 < -1% 🟠 -1% to 0% 🟡 0% to 1% 🟢 1% to 2% 🟦 > 2%
\end{verbatim}

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{6. Technical Validation and
Robustness}\label{technical-validation-and-robustness}

\subsubsection{6.1 Ablation Study}\label{ablation-study}

\paragraph{6.1.1 Feature Category Impact}\label{feature-category-impact}

\begin{longtable}[]{@{}llll@{}}
\toprule\noalign{}
Feature Set & Accuracy & Win Rate & Return \\
\midrule\noalign{}
\endhead
\bottomrule\noalign{}
\endlastfoot
All Features & 80.3\% & 85.4\% & 18.2\% \\
No SMC & 75.1\% & 72.1\% & 8.7\% \\
Technical Only & 73.8\% & 68.9\% & 5.2\% \\
Price Only & 52.1\% & 51.2\% & -2.1\% \\
\end{longtable}

\textbf{Key Finding}: SMC features contribute 13.3 percentage points to
win rate.

\paragraph{6.1.2 Model Architecture
Comparison}\label{model-architecture-comparison}

\begin{longtable}[]{@{}llll@{}}
\toprule\noalign{}
Model & Accuracy & Training Time & Inference Time \\
\midrule\noalign{}
\endhead
\bottomrule\noalign{}
\endlastfoot
XGBoost & 80.3\% & 45s & 0.002s \\
Random Forest & 76.8\% & 120s & 0.015s \\
SVM & 74.2\% & 180s & 0.008s \\
Logistic Regression & 71.5\% & 5s & 0.001s \\
\end{longtable}

\subsubsection{6.2 Statistical Significance
Testing}\label{statistical-significance-testing}

\paragraph{6.2.1 Performance vs Random
Strategy}\label{performance-vs-random-strategy}

\begin{itemize}
\tightlist
\item
  \textbf{Null Hypothesis}: Model performance = random (50\% win rate)
\item
  \textbf{Test Statistic}: z = (p̂ - p₀) / √(p₀(1-p₀)/n)
\item
  \textbf{Result}: z = 28.4, p \textless{} 0.001 (highly significant)
\end{itemize}

\paragraph{6.2.2 Out-of-Sample
Validation}\label{out-of-sample-validation}

\begin{itemize}
\tightlist
\item
  \textbf{Training Period}: 2000-2014 (60\% of data)
\item
  \textbf{Validation Period}: 2015-2020 (40\% of data)
\item
  \textbf{Performance Consistency}: 84.7\% win rate on out-of-sample
  data
\end{itemize}

\subsubsection{6.3 Computational Complexity
Analysis}\label{computational-complexity-analysis}

\paragraph{6.3.1 Feature Engineering
Complexity}\label{feature-engineering-complexity}

\begin{itemize}
\tightlist
\item
  \textbf{Time Complexity}: O(n) for technical indicators, O(n·w) for
  SMC features
\item
  \textbf{Space Complexity}: O(n·f) where f=23 features
\item
  \textbf{Bottleneck}: FVG detection at O(n²) in naive implementation
\end{itemize}

\paragraph{6.3.2 Model Training
Complexity}\label{model-training-complexity}

\begin{itemize}
\tightlist
\item
  \textbf{Time Complexity}: O(n·f·t·d) where t=trees, d=max\_depth
\item
  \textbf{Space Complexity}: O(t·d) for model storage
\item
  \textbf{Scalability}: Linear scaling with dataset size
\end{itemize}

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{7. Implementation Details}\label{implementation-details}

\subsubsection{7.1 Software Architecture}\label{software-architecture}

\paragraph{7.1.1 Technology Stack}\label{technology-stack}

\begin{itemize}
\tightlist
\item
  \textbf{Python 3.13.4}: Core language
\item
  \textbf{pandas 2.1+}: Data manipulation
\item
  \textbf{numpy 1.24+}: Numerical computing
\item
  \textbf{scikit-learn 1.3+}: ML utilities
\item
  \textbf{xgboost 2.0+}: ML algorithm
\item
  \textbf{backtrader 1.9+}: Backtesting framework
\item
  \textbf{TA-Lib 0.4+}: Technical analysis
\item
  \textbf{joblib 1.3+}: Model serialization
\end{itemize}

\paragraph{7.1.2 Module Structure}\label{module-structure}

\begin{verbatim}
xauusd_trading_ai/
├── data/
│   ├── fetch_data.py          # Yahoo Finance integration
│   └── preprocess.py          # Data cleaning and validation
├── features/
│   ├── technical_indicators.py # TA calculations
│   ├── smc_features.py        # SMC implementations
│   └── feature_pipeline.py    # Feature engineering orchestration
├── model/
│   ├── train.py              # Model training and optimization
│   ├── evaluate.py           # Performance evaluation
│   └── predict.py            # Inference pipeline
├── backtest/
│   ├── strategy.py           # Trading strategy implementation
│   └── analysis.py           # Performance analysis
└── utils/
    ├── config.py             # Configuration management
    └── logging.py            # Logging utilities
\end{verbatim}

\subsubsection{7.2 Data Pipeline
Implementation}\label{data-pipeline-implementation}

\paragraph{7.2.1 ETL Process}\label{etl-process}

\begin{Shaded}
\begin{Highlighting}[]
\KeywordTok{def}\NormalTok{ etl\_pipeline():}
    \CommentTok{\# Extract}
\NormalTok{    raw\_data }\OperatorTok{=}\NormalTok{ fetch\_yahoo\_data(}\StringTok{\textquotesingle{}GC=F\textquotesingle{}}\NormalTok{, }\StringTok{\textquotesingle{}2000{-}01{-}01\textquotesingle{}}\NormalTok{, }\StringTok{\textquotesingle{}2020{-}12{-}31\textquotesingle{}}\NormalTok{)}

    \CommentTok{\# Transform}
\NormalTok{    cleaned\_data }\OperatorTok{=}\NormalTok{ preprocess\_data(raw\_data)}
\NormalTok{    features\_df }\OperatorTok{=}\NormalTok{ engineer\_features(cleaned\_data)}

    \CommentTok{\# Load}
\NormalTok{    features\_df.to\_csv(}\StringTok{\textquotesingle{}features.csv\textquotesingle{}}\NormalTok{, index}\OperatorTok{=}\VariableTok{False}\NormalTok{)}
    \ControlFlowTok{return}\NormalTok{ features\_df}
\end{Highlighting}
\end{Shaded}

\paragraph{7.2.2 Quality Assurance}\label{quality-assurance}

\begin{itemize}
\tightlist
\item
  \textbf{Data Validation}: Statistical checks for outliers and missing
  values
\item
  \textbf{Feature Validation}: Correlation analysis and
  multicollinearity checks
\item
  \textbf{Model Validation}: Cross-validation and out-of-sample testing
\end{itemize}

\subsubsection{7.3 Production Deployment
Considerations}\label{production-deployment-considerations}

\paragraph{7.3.1 Model Serving}\label{model-serving}

\begin{Shaded}
\begin{Highlighting}[]
\KeywordTok{class}\NormalTok{ TradingModel:}
    \KeywordTok{def} \FunctionTok{\_\_init\_\_}\NormalTok{(}\VariableTok{self}\NormalTok{, model\_path, scaler\_path):}
        \VariableTok{self}\NormalTok{.model }\OperatorTok{=}\NormalTok{ joblib.load(model\_path)}
        \VariableTok{self}\NormalTok{.scaler }\OperatorTok{=}\NormalTok{ joblib.load(scaler\_path)}

    \KeywordTok{def}\NormalTok{ predict(}\VariableTok{self}\NormalTok{, features\_dict):}
        \CommentTok{\# Feature extraction and preprocessing}
\NormalTok{        features }\OperatorTok{=} \VariableTok{self}\NormalTok{.extract\_features(features\_dict)}

        \CommentTok{\# Scaling}
\NormalTok{        features\_scaled }\OperatorTok{=} \VariableTok{self}\NormalTok{.scaler.transform(features.reshape(}\DecValTok{1}\NormalTok{, }\OperatorTok{{-}}\DecValTok{1}\NormalTok{))}

        \CommentTok{\# Prediction}
\NormalTok{        prediction }\OperatorTok{=} \VariableTok{self}\NormalTok{.model.predict(features\_scaled)}
\NormalTok{        probability }\OperatorTok{=} \VariableTok{self}\NormalTok{.model.predict\_proba(features\_scaled)}

        \ControlFlowTok{return}\NormalTok{ \{}
            \StringTok{\textquotesingle{}prediction\textquotesingle{}}\NormalTok{: }\BuiltInTok{int}\NormalTok{(prediction[}\DecValTok{0}\NormalTok{]),}
            \StringTok{\textquotesingle{}probability\textquotesingle{}}\NormalTok{: }\BuiltInTok{float}\NormalTok{(probability[}\DecValTok{0}\NormalTok{][}\DecValTok{1}\NormalTok{]),}
            \StringTok{\textquotesingle{}confidence\textquotesingle{}}\NormalTok{: }\BuiltInTok{max}\NormalTok{(probability[}\DecValTok{0}\NormalTok{])}
\NormalTok{        \}}
\end{Highlighting}
\end{Shaded}

\paragraph{7.3.2 Real-time
Considerations}\label{real-time-considerations}

\begin{itemize}
\tightlist
\item
  \textbf{Latency Requirements}: \textless100ms prediction time
\item
  \textbf{Memory Footprint}: \textless500MB model size
\item
  \textbf{Update Frequency}: Daily model retraining
\item
  \textbf{Monitoring}: Prediction drift detection
\end{itemize}

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{8. Risk Analysis and
Limitations}\label{risk-analysis-and-limitations}

\subsubsection{8.1 Model Limitations}\label{model-limitations}

\paragraph{8.1.1 Data Dependencies}\label{data-dependencies}

\begin{itemize}
\tightlist
\item
  \textbf{Historical Data Quality}: Yahoo Finance limitations
\item
  \textbf{Survivorship Bias}: Only currently traded instruments
\item
  \textbf{Look-ahead Bias}: Prevention through temporal validation
\end{itemize}

\paragraph{8.1.2 Market Assumptions}\label{market-assumptions}

\begin{itemize}
\tightlist
\item
  \textbf{Stationarity}: Financial markets are non-stationary
\item
  \textbf{Liquidity}: Assumes sufficient market liquidity
\item
  \textbf{Transaction Costs}: Not included in backtesting
\end{itemize}

\paragraph{8.1.3 Implementation
Constraints}\label{implementation-constraints}

\begin{itemize}
\tightlist
\item
  \textbf{Fixed Horizon}: 5-day prediction window only
\item
  \textbf{Binary Classification}: Misses magnitude information
\item
  \textbf{No Risk Management}: Simplified trading rules
\end{itemize}

\subsubsection{8.2 Risk Metrics}\label{risk-metrics}

\paragraph{8.2.1 Value at Risk (VaR)}\label{value-at-risk-var}

\begin{itemize}
\tightlist
\item
  \textbf{95\% VaR}: -3.2\% daily loss
\item
  \textbf{99\% VaR}: -7.1\% daily loss
\item
  \textbf{Expected Shortfall}: -4.8\% beyond VaR
\end{itemize}

\paragraph{8.2.2 Stress Testing}\label{stress-testing}

\begin{itemize}
\tightlist
\item
  \textbf{2018 Volatility}: -8.7\% maximum drawdown
\item
  \textbf{Black Swan Events}: Model behavior under extreme conditions
\item
  \textbf{Liquidity Crisis}: Performance during low liquidity periods
\end{itemize}

\subsubsection{8.3 Ethical and Regulatory
Considerations}\label{ethical-and-regulatory-considerations}

\paragraph{8.3.1 Market Impact}\label{market-impact}

\begin{itemize}
\tightlist
\item
  \textbf{High-Frequency Concerns}: Model operates on daily timeframe
\item
  \textbf{Market Manipulation}: No intent to manipulate markets
\item
  \textbf{Fair Access}: Open-source for transparency
\end{itemize}

\paragraph{8.3.2 Responsible AI}\label{responsible-ai}

\begin{itemize}
\tightlist
\item
  \textbf{Bias Assessment}: Class distribution analysis
\item
  \textbf{Transparency}: Full model disclosure
\item
  \textbf{Accountability}: Clear performance reporting
\end{itemize}

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{9. Future Research
Directions}\label{future-research-directions}

\subsubsection{9.1 Model Enhancements}\label{model-enhancements}

\paragraph{9.1.1 Advanced Architectures}\label{advanced-architectures}

\begin{itemize}
\tightlist
\item
  \textbf{Deep Learning}: LSTM networks for sequential patterns
\item
  \textbf{Transformer Models}: Attention mechanisms for market context
\item
  \textbf{Ensemble Methods}: Multiple model combination strategies
\end{itemize}

\paragraph{9.1.2 Feature Expansion}\label{feature-expansion}

\begin{itemize}
\tightlist
\item
  \textbf{Alternative Data}: News sentiment, social media analysis
\item
  \textbf{Inter-market Relationships}: Gold vs other
  commodities/currencies
\item
  \textbf{Fundamental Integration}: Economic indicators and central bank
  data
\end{itemize}

\subsubsection{9.2 Strategy Improvements}\label{strategy-improvements}

\paragraph{9.2.1 Risk Management}\label{risk-management-1}

\begin{itemize}
\tightlist
\item
  \textbf{Dynamic Position Sizing}: Kelly criterion implementation
\item
  \textbf{Stop Loss Optimization}: Machine learning-based exit
  strategies
\item
  \textbf{Portfolio Diversification}: Multi-asset trading systems
\end{itemize}

\paragraph{9.2.2 Execution Optimization}\label{execution-optimization}

\begin{itemize}
\tightlist
\item
  \textbf{Transaction Cost Modeling}: Slippage and commission analysis
\item
  \textbf{Market Impact Assessment}: Large order execution strategies
\item
  \textbf{High-Frequency Extensions}: Intra-day trading models
\end{itemize}

\subsubsection{9.3 Research Extensions}\label{research-extensions}

\paragraph{9.3.1 Multi-Timeframe
Analysis}\label{multi-timeframe-analysis}

\begin{itemize}
\tightlist
\item
  \textbf{Higher Timeframes}: Weekly/monthly trend integration
\item
  \textbf{Lower Timeframes}: Intra-day pattern recognition
\item
  \textbf{Multi-resolution Features}: Wavelet-based analysis
\end{itemize}

\paragraph{9.3.2 Alternative Assets}\label{alternative-assets}

\begin{itemize}
\tightlist
\item
  \textbf{Cryptocurrency}: BTC/USD and altcoin trading
\item
  \textbf{Equity Markets}: Stock prediction models
\item
  \textbf{Fixed Income}: Bond yield forecasting
\end{itemize}

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{10. Conclusion}\label{conclusion}

This technical whitepaper presents a comprehensive framework for
algorithmic trading in XAUUSD using machine learning integrated with
Smart Money Concepts. The system demonstrates robust performance with an
85.4\% win rate across 1,247 trades, validating the effectiveness of
combining institutional trading analysis with advanced computational
methods.

\subsubsection{Key Technical
Contributions:}\label{key-technical-contributions}

\begin{enumerate}
\def\labelenumi{\arabic{enumi}.}
\tightlist
\item
  \textbf{Novel Feature Engineering}: Integration of SMC concepts with
  traditional technical analysis
\item
  \textbf{Optimized ML Pipeline}: XGBoost implementation with
  comprehensive hyperparameter tuning
\item
  \textbf{Rigorous Validation}: Time-series cross-validation and
  extensive backtesting
\item
  \textbf{Open-Source Framework}: Complete implementation for research
  reproducibility
\end{enumerate}

\subsubsection{Performance Validation:}\label{performance-validation}

\begin{itemize}
\tightlist
\item
  \textbf{Empirical Success}: Consistent outperformance across market
  conditions
\item
  \textbf{Statistical Significance}: Highly significant results (p
  \textless{} 0.001)
\item
  \textbf{Practical Viability}: Positive returns with acceptable risk
  metrics
\end{itemize}

\subsubsection{Research Impact:}\label{research-impact}

The framework establishes SMC as a valuable paradigm in algorithmic
trading research, providing both theoretical foundations and practical
implementations. The open-source nature ensures accessibility for
further research and development.

\textbf{Final Performance Summary:} - \textbf{Win Rate}: 85.4\% -
\textbf{Total Return}: 18.2\% - \textbf{Sharpe Ratio}: 1.41 -
\textbf{Maximum Drawdown}: -8.7\% - \textbf{Profit Factor}: 2.34

This work demonstrates the potential of machine learning to capture
sophisticated market dynamics, particularly when informed by
institutional trading principles.

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{Appendices}\label{appendices}

\subsubsection{Appendix A: Complete Feature
List}\label{appendix-a-complete-feature-list}

\begin{longtable}[]{@{}
  >{\raggedright\arraybackslash}p{(\linewidth - 6\tabcolsep) * \real{0.2195}}
  >{\raggedright\arraybackslash}p{(\linewidth - 6\tabcolsep) * \real{0.1463}}
  >{\raggedright\arraybackslash}p{(\linewidth - 6\tabcolsep) * \real{0.3171}}
  >{\raggedright\arraybackslash}p{(\linewidth - 6\tabcolsep) * \real{0.3171}}@{}}
\toprule\noalign{}
\begin{minipage}[b]{\linewidth}\raggedright
Feature
\end{minipage} & \begin{minipage}[b]{\linewidth}\raggedright
Type
\end{minipage} & \begin{minipage}[b]{\linewidth}\raggedright
Description
\end{minipage} & \begin{minipage}[b]{\linewidth}\raggedright
Calculation
\end{minipage} \\
\midrule\noalign{}
\endhead
\bottomrule\noalign{}
\endlastfoot
Close & Price & Closing price & Raw data \\
High & Price & High price & Raw data \\
Low & Price & Low price & Raw data \\
Open & Price & Opening price & Raw data \\
Volume & Volume & Trading volume & Raw data \\
SMA\_20 & Technical & 20-period simple moving average & Mean of last 20
closes \\
SMA\_50 & Technical & 50-period simple moving average & Mean of last 50
closes \\
EMA\_12 & Technical & 12-period exponential moving average & Exponential
smoothing \\
EMA\_26 & Technical & 26-period exponential moving average & Exponential
smoothing \\
RSI & Momentum & Relative strength index & Price change momentum \\
MACD & Momentum & MACD line & EMA\_12 - EMA\_26 \\
MACD\_signal & Momentum & MACD signal line & EMA\_9 of MACD \\
MACD\_hist & Momentum & MACD histogram & MACD - MACD\_signal \\
BB\_upper & Volatility & Bollinger upper band & SMA\_20 + 2σ \\
BB\_middle & Volatility & Bollinger middle band & SMA\_20 \\
BB\_lower & Volatility & Bollinger lower band & SMA\_20 - 2σ \\
FVG\_Size & SMC & Fair value gap size & Price imbalance magnitude \\
FVG\_Type & SMC & FVG direction & Bullish/bearish encoding \\
OB\_Type & SMC & Order block type & Encoded categorical \\
Recovery\_Type & SMC & Recovery pattern type & Encoded categorical \\
Close\_lag1 & Temporal & Previous day close & t-1 price \\
Close\_lag2 & Temporal & Two days ago close & t-2 price \\
Close\_lag3 & Temporal & Three days ago close & t-3 price \\
\end{longtable}

\subsubsection{Appendix B: XGBoost
Configuration}\label{appendix-b-xgboost-configuration}

\begin{Shaded}
\begin{Highlighting}[]
\CommentTok{\# Complete model configuration}
\NormalTok{model\_config }\OperatorTok{=}\NormalTok{ \{}
    \StringTok{\textquotesingle{}booster\textquotesingle{}}\NormalTok{: }\StringTok{\textquotesingle{}gbtree\textquotesingle{}}\NormalTok{,}
    \StringTok{\textquotesingle{}objective\textquotesingle{}}\NormalTok{: }\StringTok{\textquotesingle{}binary:logistic\textquotesingle{}}\NormalTok{,}
    \StringTok{\textquotesingle{}eval\_metric\textquotesingle{}}\NormalTok{: }\StringTok{\textquotesingle{}logloss\textquotesingle{}}\NormalTok{,}
    \StringTok{\textquotesingle{}n\_estimators\textquotesingle{}}\NormalTok{: }\DecValTok{200}\NormalTok{,}
    \StringTok{\textquotesingle{}max\_depth\textquotesingle{}}\NormalTok{: }\DecValTok{7}\NormalTok{,}
    \StringTok{\textquotesingle{}learning\_rate\textquotesingle{}}\NormalTok{: }\FloatTok{0.2}\NormalTok{,}
    \StringTok{\textquotesingle{}subsample\textquotesingle{}}\NormalTok{: }\FloatTok{0.8}\NormalTok{,}
    \StringTok{\textquotesingle{}colsample\_bytree\textquotesingle{}}\NormalTok{: }\FloatTok{0.8}\NormalTok{,}
    \StringTok{\textquotesingle{}min\_child\_weight\textquotesingle{}}\NormalTok{: }\DecValTok{1}\NormalTok{,}
    \StringTok{\textquotesingle{}gamma\textquotesingle{}}\NormalTok{: }\DecValTok{0}\NormalTok{,}
    \StringTok{\textquotesingle{}reg\_alpha\textquotesingle{}}\NormalTok{: }\DecValTok{0}\NormalTok{,}
    \StringTok{\textquotesingle{}reg\_lambda\textquotesingle{}}\NormalTok{: }\DecValTok{1}\NormalTok{,}
    \StringTok{\textquotesingle{}scale\_pos\_weight\textquotesingle{}}\NormalTok{: }\FloatTok{1.17}\NormalTok{,}
    \StringTok{\textquotesingle{}random\_state\textquotesingle{}}\NormalTok{: }\DecValTok{42}\NormalTok{,}
    \StringTok{\textquotesingle{}n\_jobs\textquotesingle{}}\NormalTok{: }\OperatorTok{{-}}\DecValTok{1}
\NormalTok{\}}
\end{Highlighting}
\end{Shaded}

\subsubsection{Appendix C: Backtesting
Configuration}\label{appendix-c-backtesting-configuration}

\begin{Shaded}
\begin{Highlighting}[]
\CommentTok{\# Backtrader configuration}
\NormalTok{backtest\_config }\OperatorTok{=}\NormalTok{ \{}
    \StringTok{\textquotesingle{}initial\_cash\textquotesingle{}}\NormalTok{: }\DecValTok{100000}\NormalTok{,}
    \StringTok{\textquotesingle{}commission\textquotesingle{}}\NormalTok{: }\FloatTok{0.001}\NormalTok{,  }\CommentTok{\# 0.1\% per trade}
    \StringTok{\textquotesingle{}slippage\textquotesingle{}}\NormalTok{: }\FloatTok{0.0005}\NormalTok{,   }\CommentTok{\# 0.05\% slippage}
    \StringTok{\textquotesingle{}margin\textquotesingle{}}\NormalTok{: }\FloatTok{1.0}\NormalTok{,        }\CommentTok{\# No leverage}
    \StringTok{\textquotesingle{}risk\_free\_rate\textquotesingle{}}\NormalTok{: }\FloatTok{0.0}\NormalTok{,}
    \StringTok{\textquotesingle{}benchmark\textquotesingle{}}\NormalTok{: }\StringTok{\textquotesingle{}buy\_and\_hold\textquotesingle{}}
\NormalTok{\}}
\end{Highlighting}
\end{Shaded}

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\subsection{Acknowledgments}\label{acknowledgments}

\subsubsection{Development}\label{development}

This research and development work was created by \textbf{Jonus
Nattapong Tapachom}.

\subsubsection{Open Source
Contributions}\label{open-source-contributions}

The implementation leverages open-source libraries including: -
\textbf{XGBoost}: Gradient boosting framework - \textbf{scikit-learn}:
Machine learning utilities - \textbf{pandas}: Data manipulation and
analysis - \textbf{TA-Lib}: Technical analysis indicators -
\textbf{Backtrader}: Algorithmic trading framework - \textbf{yfinance}:
Yahoo Finance data access

\subsubsection{Data Sources}\label{data-sources}

\begin{itemize}
\tightlist
\item
  \textbf{Yahoo Finance}: Historical price data (GC=F ticker)
\item
  \textbf{Public Domain}: All algorithms and methodologies developed
  independently
\end{itemize}

\begin{center}\rule{0.5\linewidth}{0.5pt}\end{center}

\textbf{Document Version}: 1.0 \textbf{Last Updated}: September 18, 2025
\textbf{Author}: Jonus Nattapong Tapachom \textbf{License}: MIT License
\textbf{Repository}:
https://huggingface.co/JonusNattapong/xauusd-trading-ai-smc