MakeoverMonday: Risk Exposure Analysis

MakeoverMonday

Data Viz

Python

Finance

Trading

Risk

Visualizing the relationship between Cash Reserve (PAT Balance) and Margin Exposure over time

Author

chokotto

Published

January 27, 2026

Overview

This MakeoverMonday project visualizes the relationship between Cash Reserve (cumulative post-tax profit) and Margin Exposure (open margin positions) from my personal trading accounts.

Key metrics:

PAT Balance: Cumulative Profit After Tax - represents available cash reserve
Exposure: Total margin position value (JPY equivalent)
Exposure Ratio: Exposure / |PAT Balance| - measures leverage relative to cash reserve

This analysis helps understand the risk profile of the trading portfolio over time.

Dataset

Show code

import pandas as pd
import numpy as np
import plotly.express as px
import plotly.graph_objects as go
from plotly.subplots import make_subplots
from pathlib import Path
from datetime import datetime

Show code

# Define base path
base_path = Path.cwd()
while not (base_path / "data").exists() and base_path.parent != base_path:
    base_path = base_path.parent

# Load daily balance data
balance_file = base_path / "data" / "trading_account" / "account_balance" / "daily_balance.parquet"

if balance_file.exists():
    daily_balance = pd.read_parquet(balance_file)
    daily_balance['date'] = pd.to_datetime(daily_balance['date'])
    print(f"Loaded {len(daily_balance)} daily balance records")
else:
    print(f"Data not found at {balance_file}")
    daily_balance = pd.DataFrame()

# Extract month-end data for cleaner visualization
if not daily_balance.empty:
    daily_balance['year_month'] = daily_balance['date'].dt.to_period('M').astype(str)
    
    # Get month-end data for each broker
    month_end = daily_balance.groupby(['year_month', 'broker']).apply(
        lambda x: x.loc[x['date'].idxmax()],
        include_groups=False
    ).reset_index()
    
    print(f"Month-end records: {len(month_end)}")
    print(f"Date range: {month_end['year_month'].min()} to {month_end['year_month'].max()}")

Loaded 780 daily balance records
Month-end records: 26
Date range: 2025-01 to 2026-01

Show code

# Calculate totals across brokers
total_by_month = month_end.groupby('year_month').agg({
    'pat_balance': 'sum',
    'exposure': 'sum'
}).reset_index()

total_by_month['broker'] = 'Total'

# Calculate exposure ratio (using absolute value of PAT for meaningful ratio)
total_by_month['exposure_ratio'] = np.where(
    total_by_month['pat_balance'] != 0,
    total_by_month['exposure'] / np.abs(total_by_month['pat_balance']) * 100,
    np.nan
)

# Add ratio to month_end data
month_end['exposure_ratio'] = np.where(
    month_end['pat_balance'] != 0,
    month_end['exposure'] / np.abs(month_end['pat_balance']) * 100,
    np.nan
)

# Combine broker data with total
all_data = pd.concat([month_end, total_by_month], ignore_index=True)

# Convert year_month to datetime for plotting
all_data['date'] = pd.to_datetime(all_data['year_month'] + '-01')

print(f"\nData summary:")
print(f"Brokers: {all_data['broker'].unique().tolist()}")
print(f"\nLatest Total values:")
latest = total_by_month[total_by_month['year_month'] == total_by_month['year_month'].max()].iloc[0]
print(f"  PAT Balance: ¥{latest['pat_balance']:,.0f}")
print(f"  Exposure: ¥{latest['exposure']:,.0f}")
print(f"  Exposure Ratio: {latest['exposure_ratio']:.1f}%")


Data summary:
Brokers: ['Rakuten', 'SBI', 'Total']

Latest Total values:
  PAT Balance: ¥12,477,234
  Exposure: ¥35,691,977
  Exposure Ratio: 286.1%

Chart 1: PAT Balance & Exposure Time Series (Total)

This chart shows the absolute values of Cash Reserve and Margin Exposure over time.

Show code

# Filter for Total only
total_data = all_data[all_data['broker'] == 'Total'].copy()

fig1 = make_subplots(specs=[[{"secondary_y": False}]])

# PAT Balance as area
fig1.add_trace(
    go.Scatter(
        x=total_data['date'],
        y=total_data['pat_balance'],
        name='PAT Balance (Cash Reserve)',
        fill='tozeroy',
        fillcolor='rgba(34, 197, 94, 0.3)',
        line=dict(color='#22c55e', width=2),
        hovertemplate='%{x|%Y-%m}<br>PAT Balance: ¥%{y:,.0f}<extra></extra>'
    )
)

# Exposure as line
fig1.add_trace(
    go.Scatter(
        x=total_data['date'],
        y=total_data['exposure'],
        name='Exposure (Margin Position)',
        line=dict(color='#ef4444', width=3),
        hovertemplate='%{x|%Y-%m}<br>Exposure: ¥%{y:,.0f}<extra></extra>'
    )
)

# Add zero line
fig1.add_hline(y=0, line_dash="dash", line_color="gray", opacity=0.5)

fig1.update_layout(
    title=dict(
        text='<b>Cash Reserve vs Margin Exposure</b><br><sup>Monthly time series (Total across all brokers)</sup>',
        x=0.5,
        xanchor='center'
    ),
    xaxis_title='Month',
    yaxis_title='Amount (JPY)',
    yaxis=dict(tickformat=',.0f', tickprefix='¥'),
    legend=dict(
        orientation='h',
        yanchor='bottom',
        y=1.02,
        xanchor='center',
        x=0.5
    ),
    template='plotly_white',
    height=500,
    hovermode='x unified'
)

fig1.show()

Chart 2: Exposure Ratio Time Series

This chart shows how the leverage ratio (Exposure / |PAT Balance|) changes over time.

Show code

# Create ratio chart
fig2 = go.Figure()

# Add ratio line
fig2.add_trace(
    go.Scatter(
        x=total_data['date'],
        y=total_data['exposure_ratio'],
        name='Exposure Ratio',
        fill='tozeroy',
        fillcolor='rgba(239, 68, 68, 0.2)',
        line=dict(color='#ef4444', width=2),
        hovertemplate='%{x|%Y-%m}<br>Ratio: %{y:.1f}%<extra></extra>'
    )
)

# Add reference lines
fig2.add_hline(y=100, line_dash="dash", line_color="#f59e0b", 
               annotation_text="100% (Exposure = |PAT|)", 
               annotation_position="right")
fig2.add_hline(y=200, line_dash="dot", line_color="#dc2626",
               annotation_text="200% (High Risk)",
               annotation_position="right")

fig2.update_layout(
    title=dict(
        text='<b>Exposure Ratio Over Time</b><br><sup>Exposure / |PAT Balance| × 100 (Total)</sup>',
        x=0.5,
        xanchor='center'
    ),
    xaxis_title='Month',
    yaxis_title='Exposure Ratio (%)',
    yaxis=dict(ticksuffix='%'),
    template='plotly_white',
    height=400,
    showlegend=False
)

fig2.show()

Chart 3: Broker Breakdown - Stacked Area

This chart shows how each broker contributes to the total exposure and cash reserve.

Show code

# Filter for brokers only (not Total)
broker_data = all_data[all_data['broker'] != 'Total'].copy()

# Pivot for stacked area
exposure_pivot = broker_data.pivot(index='date', columns='broker', values='exposure').fillna(0)

fig3 = go.Figure()

colors = {'Rakuten': '#3b82f6', 'SBI': '#8b5cf6'}

for broker in exposure_pivot.columns:
    fig3.add_trace(
        go.Scatter(
            x=exposure_pivot.index,
            y=exposure_pivot[broker],
            name=broker,
            stackgroup='one',
            fillcolor=colors.get(broker, '#64748b'),
            line=dict(width=0.5, color=colors.get(broker, '#64748b')),
            hovertemplate=f'{broker}<br>' + '%{x|%Y-%m}<br>Exposure: ¥%{y:,.0f}<extra></extra>'
        )
    )

fig3.update_layout(
    title=dict(
        text='<b>Margin Exposure by Broker</b><br><sup>Stacked area chart showing broker contribution</sup>',
        x=0.5,
        xanchor='center'
    ),
    xaxis_title='Month',
    yaxis_title='Exposure (JPY)',
    yaxis=dict(tickformat=',.0f', tickprefix='¥'),
    legend=dict(
        orientation='h',
        yanchor='bottom',
        y=1.02,
        xanchor='center',
        x=0.5
    ),
    template='plotly_white',
    height=400,
    hovermode='x unified'
)

fig3.show()

Show code

# PAT Balance by broker
pat_pivot = broker_data.pivot(index='date', columns='broker', values='pat_balance').fillna(0)

fig3b = go.Figure()

for broker in pat_pivot.columns:
    fig3b.add_trace(
        go.Scatter(
            x=pat_pivot.index,
            y=pat_pivot[broker],
            name=broker,
            stackgroup='one',
            fillcolor=colors.get(broker, '#64748b'),
            line=dict(width=0.5, color=colors.get(broker, '#64748b')),
            hovertemplate=f'{broker}<br>' + '%{x|%Y-%m}<br>PAT Balance: ¥%{y:,.0f}<extra></extra>'
        )
    )

fig3b.add_hline(y=0, line_dash="dash", line_color="gray", opacity=0.5)

fig3b.update_layout(
    title=dict(
        text='<b>PAT Balance (Cash Reserve) by Broker</b><br><sup>Cumulative post-tax profit by broker</sup>',
        x=0.5,
        xanchor='center'
    ),
    xaxis_title='Month',
    yaxis_title='PAT Balance (JPY)',
    yaxis=dict(tickformat=',.0f', tickprefix='¥'),
    legend=dict(
        orientation='h',
        yanchor='bottom',
        y=1.02,
        xanchor='center',
        x=0.5
    ),
    template='plotly_white',
    height=400,
    hovermode='x unified'
)

fig3b.show()

Chart 4: Risk-Balance Matrix (Scatter Plot)

This scatter plot shows the relationship between Cash Reserve and Exposure. Each point represents a month, with color indicating time progression.

Show code

fig4 = go.Figure()

# Add scatter points with color gradient by time
fig4.add_trace(
    go.Scatter(
        x=total_data['pat_balance'],
        y=total_data['exposure'],
        mode='markers+text',
        marker=dict(
            size=12,
            color=list(range(len(total_data))),
            colorscale='Viridis',
            showscale=True,
            colorbar=dict(title='Time<br>(older→newer)')
        ),
        text=total_data['year_month'].str[2:7],  # Show YY-MM
        textposition='top center',
        textfont=dict(size=9),
        hovertemplate='%{text}<br>PAT: ¥%{x:,.0f}<br>Exposure: ¥%{y:,.0f}<extra></extra>'
    )
)

# Add diagonal lines for reference
max_val = max(total_data['exposure'].max(), abs(total_data['pat_balance']).max())
min_val = min(total_data['pat_balance'].min(), 0)

# 100% line (Exposure = |PAT|)
fig4.add_trace(
    go.Scatter(
        x=[min_val, max_val],
        y=[abs(min_val), max_val],
        mode='lines',
        line=dict(dash='dash', color='#f59e0b', width=1),
        name='100% (Exposure = |PAT|)',
        hoverinfo='skip'
    )
)

# Zero PAT line
fig4.add_vline(x=0, line_dash="dot", line_color="gray", opacity=0.5)

# Add risk zones annotation
fig4.add_annotation(
    x=total_data['pat_balance'].max() * 0.8,
    y=total_data['exposure'].max() * 0.3,
    text="Lower Risk Zone<br>(High PAT, Low Exposure)",
    showarrow=False,
    font=dict(color='#22c55e', size=11),
    bgcolor='rgba(255,255,255,0.8)'
)

fig4.add_annotation(
    x=total_data['pat_balance'].min() * 0.5 if total_data['pat_balance'].min() < 0 else 0,
    y=total_data['exposure'].max() * 0.8,
    text="Higher Risk Zone<br>(Low/Negative PAT, High Exposure)",
    showarrow=False,
    font=dict(color='#ef4444', size=11),
    bgcolor='rgba(255,255,255,0.8)'
)

fig4.update_layout(
    title=dict(
        text='<b>Risk-Balance Matrix</b><br><sup>Each point = one month (color: time progression)</sup>',
        x=0.5,
        xanchor='center'
    ),
    xaxis_title='PAT Balance (Cash Reserve) - JPY',
    yaxis_title='Margin Exposure - JPY',
    xaxis=dict(tickformat=',.0f', tickprefix='¥', zeroline=True),
    yaxis=dict(tickformat=',.0f', tickprefix='¥'),
    template='plotly_white',
    height=600,
    legend=dict(
        orientation='h',
        yanchor='bottom',
        y=-0.15,
        xanchor='center',
        x=0.5
    )
)

fig4.show()

Summary Statistics

Show code

# Calculate summary statistics
summary = total_by_month.describe()

print("=== Risk Exposure Summary ===\n")
print(f"Analysis Period: {total_by_month['year_month'].min()} to {total_by_month['year_month'].max()}")
print(f"Total Months: {len(total_by_month)}")
print()

print("PAT Balance (Cash Reserve):")
print(f"  Current: ¥{total_by_month['pat_balance'].iloc[-1]:,.0f}")
print(f"  Maximum: ¥{total_by_month['pat_balance'].max():,.0f}")
print(f"  Minimum: ¥{total_by_month['pat_balance'].min():,.0f}")
print()

print("Exposure (Margin Position):")
print(f"  Current: ¥{total_by_month['exposure'].iloc[-1]:,.0f}")
print(f"  Maximum: ¥{total_by_month['exposure'].max():,.0f}")
print(f"  Minimum: ¥{total_by_month['exposure'].min():,.0f}")
print()

print("Exposure Ratio:")
valid_ratios = total_by_month['exposure_ratio'].dropna()
if len(valid_ratios) > 0:
    print(f"  Current: {total_by_month['exposure_ratio'].iloc[-1]:.1f}%")
    print(f"  Maximum: {valid_ratios.max():.1f}%")
    print(f"  Average: {valid_ratios.mean():.1f}%")

=== Risk Exposure Summary ===

Analysis Period: 2025-01 to 2026-01
Total Months: 13

PAT Balance (Cash Reserve):
  Current: ¥12,477,234
  Maximum: ¥12,477,234
  Minimum: ¥-3,342,595

Exposure (Margin Position):
  Current: ¥35,691,977
  Maximum: ¥35,691,977
  Minimum: ¥0

Exposure Ratio:
  Current: 286.1%
  Maximum: 117450.7%
  Average: 20090.3%

Key Insights

Leverage Dynamics: The exposure ratio shows how aggressively margin is used relative to available cash reserve.
Risk Periods: Months where exposure significantly exceeds PAT balance represent higher risk periods.
Broker Distribution: The stacked area charts reveal which broker carries more exposure at any given time.
Trajectory: The scatter plot’s color gradient shows the portfolio’s risk trajectory over time.

Design Decisions

Dual Metrics: Showing both absolute values (JPY) and ratio (%) provides complete picture of risk.
Interactive Plotly: Hover details allow precise value inspection at any point.
Color Coding: Green for cash/profit, Red for exposure/risk - intuitive color associations.
Reference Lines: 100% and 200% ratio lines provide quick risk assessment benchmarks.
Time-Colored Scatter: Viridis colorscale shows risk evolution trajectory clearly.

This post is part of the MakeoverMonday weekly data visualization project.

Disclaimer

This analysis is for educational and practice purposes only. The figures shown are from personal trading records and should not be considered as investment advice. Margin trading involves significant risk of loss.

--- title: "MakeoverMonday: Risk Exposure Analysis" description: "Visualizing the relationship between Cash Reserve (PAT Balance) and Margin Exposure over time" date: "2026-01-27" author: "chokotto" categories: [MakeoverMonday, Data Viz, Python, Finance, Trading, Risk] image: "thumbnail.svg" code-fold: true code-tools: true code-summary: "Show code" twitter-card: card-type: summary_large_image image: "thumbnail.svg" title: "MakeoverMonday: Risk Exposure Analysis" description: "Cash Reserve vs Margin Exposure time series visualization" --- ## Overview This MakeoverMonday project visualizes the relationship between **Cash Reserve** (cumulative post-tax profit) and **Margin Exposure** (open margin positions) from my personal trading accounts. Key metrics: - **PAT Balance**: Cumulative Profit After Tax - represents available cash reserve - **Exposure**: Total margin position value (JPY equivalent) - **Exposure Ratio**: Exposure / |PAT Balance| - measures leverage relative to cash reserve This analysis helps understand the risk profile of the trading portfolio over time. ## Dataset ```{python} #| label: load-packages #| message: false import pandas as pd import numpy as np import plotly.express as px import plotly.graph_objects as go from plotly.subplots import make_subplots from pathlib import Path from datetime import datetime ``` ```{python} #| label: load-data #| message: false # Define base path base_path = Path.cwd() while not (base_path / "data").exists() and base_path.parent != base_path: base_path = base_path.parent # Load daily balance data balance_file = base_path / "data" / "trading_account" / "account_balance" / "daily_balance.parquet" if balance_file.exists(): daily_balance = pd.read_parquet(balance_file) daily_balance['date'] = pd.to_datetime(daily_balance['date']) print(f"Loaded {len(daily_balance)} daily balance records") else: print(f"Data not found at {balance_file}") daily_balance = pd.DataFrame() # Extract month-end data for cleaner visualization if not daily_balance.empty: daily_balance['year_month'] = daily_balance['date'].dt.to_period('M').astype(str) # Get month-end data for each broker month_end = daily_balance.groupby(['year_month', 'broker']).apply( lambda x: x.loc[x['date'].idxmax()], include_groups=False ).reset_index() print(f"Month-end records: {len(month_end)}") print(f"Date range: {month_end['year_month'].min()} to {month_end['year_month'].max()}") ``` ```{python} #| label: prepare-data #| message: false # Calculate totals across brokers total_by_month = month_end.groupby('year_month').agg({ 'pat_balance': 'sum', 'exposure': 'sum' }).reset_index() total_by_month['broker'] = 'Total' # Calculate exposure ratio (using absolute value of PAT for meaningful ratio) total_by_month['exposure_ratio'] = np.where( total_by_month['pat_balance'] != 0, total_by_month['exposure'] / np.abs(total_by_month['pat_balance']) * 100, np.nan ) # Add ratio to month_end data month_end['exposure_ratio'] = np.where( month_end['pat_balance'] != 0, month_end['exposure'] / np.abs(month_end['pat_balance']) * 100, np.nan ) # Combine broker data with total all_data = pd.concat([month_end, total_by_month], ignore_index=True) # Convert year_month to datetime for plotting all_data['date'] = pd.to_datetime(all_data['year_month'] + '-01') print(f"\nData summary:") print(f"Brokers: {all_data['broker'].unique().tolist()}") print(f"\nLatest Total values:") latest = total_by_month[total_by_month['year_month'] == total_by_month['year_month'].max()].iloc[0] print(f" PAT Balance: ¥{latest['pat_balance']:,.0f}") print(f" Exposure: ¥{latest['exposure']:,.0f}") print(f" Exposure Ratio: {latest['exposure_ratio']:.1f}%") ``` ## Chart 1: PAT Balance & Exposure Time Series (Total) This chart shows the absolute values of Cash Reserve and Margin Exposure over time. ```{python} #| label: chart-1-amounts #| fig-width: 12 #| fig-height: 6 # Filter for Total only total_data = all_data[all_data['broker'] == 'Total'].copy() fig1 = make_subplots(specs=[[{"secondary_y": False}]]) # PAT Balance as area fig1.add_trace( go.Scatter( x=total_data['date'], y=total_data['pat_balance'], name='PAT Balance (Cash Reserve)', fill='tozeroy', fillcolor='rgba(34, 197, 94, 0.3)', line=dict(color='#22c55e', width=2), hovertemplate='%{x|%Y-%m} PAT Balance: ¥%{y:,.0f}<extra></extra>' ) ) # Exposure as line fig1.add_trace( go.Scatter( x=total_data['date'], y=total_data['exposure'], name='Exposure (Margin Position)', line=dict(color='#ef4444', width=3), hovertemplate='%{x|%Y-%m} Exposure: ¥%{y:,.0f}<extra></extra>' ) ) # Add zero line fig1.add_hline(y=0, line_dash="dash", line_color="gray", opacity=0.5) fig1.update_layout( title=dict( text='Cash Reserve vs Margin Exposure Monthly time series (Total across all brokers)', x=0.5, xanchor='center' ), xaxis_title='Month', yaxis_title='Amount (JPY)', yaxis=dict(tickformat=',.0f', tickprefix='¥'), legend=dict( orientation='h', yanchor='bottom', y=1.02, xanchor='center', x=0.5 ), template='plotly_white', height=500, hovermode='x unified' ) fig1.show() ``` ## Chart 2: Exposure Ratio Time Series This chart shows how the leverage ratio (Exposure / |PAT Balance|) changes over time. ```{python} #| label: chart-2-ratio #| fig-width: 12 #| fig-height: 5 # Create ratio chart fig2 = go.Figure() # Add ratio line fig2.add_trace( go.Scatter( x=total_data['date'], y=total_data['exposure_ratio'], name='Exposure Ratio', fill='tozeroy', fillcolor='rgba(239, 68, 68, 0.2)', line=dict(color='#ef4444', width=2), hovertemplate='%{x|%Y-%m} Ratio: %{y:.1f}%<extra></extra>' ) ) # Add reference lines fig2.add_hline(y=100, line_dash="dash", line_color="#f59e0b", annotation_text="100% (Exposure = |PAT|)", annotation_position="right") fig2.add_hline(y=200, line_dash="dot", line_color="#dc2626", annotation_text="200% (High Risk)", annotation_position="right") fig2.update_layout( title=dict( text='Exposure Ratio Over Time Exposure / |PAT Balance| × 100 (Total)', x=0.5, xanchor='center' ), xaxis_title='Month', yaxis_title='Exposure Ratio (%)', yaxis=dict(ticksuffix='%'), template='plotly_white', height=400, showlegend=False ) fig2.show() ``` ## Chart 3: Broker Breakdown - Stacked Area This chart shows how each broker contributes to the total exposure and cash reserve. ```{python} #| label: chart-3-broker-exposure #| fig-width: 12 #| fig-height: 5 # Filter for brokers only (not Total) broker_data = all_data[all_data['broker'] != 'Total'].copy() # Pivot for stacked area exposure_pivot = broker_data.pivot(index='date', columns='broker', values='exposure').fillna(0) fig3 = go.Figure() colors = {'Rakuten': '#3b82f6', 'SBI': '#8b5cf6'} for broker in exposure_pivot.columns: fig3.add_trace( go.Scatter( x=exposure_pivot.index, y=exposure_pivot[broker], name=broker, stackgroup='one', fillcolor=colors.get(broker, '#64748b'), line=dict(width=0.5, color=colors.get(broker, '#64748b')), hovertemplate=f'{broker} ' + '%{x|%Y-%m} Exposure: ¥%{y:,.0f}<extra></extra>' ) ) fig3.update_layout( title=dict( text='Margin Exposure by Broker Stacked area chart showing broker contribution', x=0.5, xanchor='center' ), xaxis_title='Month', yaxis_title='Exposure (JPY)', yaxis=dict(tickformat=',.0f', tickprefix='¥'), legend=dict( orientation='h', yanchor='bottom', y=1.02, xanchor='center', x=0.5 ), template='plotly_white', height=400, hovermode='x unified' ) fig3.show() ``` ```{python} #| label: chart-3b-broker-pat #| fig-width: 12 #| fig-height: 5 # PAT Balance by broker pat_pivot = broker_data.pivot(index='date', columns='broker', values='pat_balance').fillna(0) fig3b = go.Figure() for broker in pat_pivot.columns: fig3b.add_trace( go.Scatter( x=pat_pivot.index, y=pat_pivot[broker], name=broker, stackgroup='one', fillcolor=colors.get(broker, '#64748b'), line=dict(width=0.5, color=colors.get(broker, '#64748b')), hovertemplate=f'{broker} ' + '%{x|%Y-%m} PAT Balance: ¥%{y:,.0f}<extra></extra>' ) ) fig3b.add_hline(y=0, line_dash="dash", line_color="gray", opacity=0.5) fig3b.update_layout( title=dict( text='PAT Balance (Cash Reserve) by Broker Cumulative post-tax profit by broker', x=0.5, xanchor='center' ), xaxis_title='Month', yaxis_title='PAT Balance (JPY)', yaxis=dict(tickformat=',.0f', tickprefix='¥'), legend=dict( orientation='h', yanchor='bottom', y=1.02, xanchor='center', x=0.5 ), template='plotly_white', height=400, hovermode='x unified' ) fig3b.show() ``` ## Chart 4: Risk-Balance Matrix (Scatter Plot) This scatter plot shows the relationship between Cash Reserve and Exposure. Each point represents a month, with color indicating time progression. ```{python} #| label: chart-4-scatter #| fig-width: 10 #| fig-height: 8 fig4 = go.Figure() # Add scatter points with color gradient by time fig4.add_trace( go.Scatter( x=total_data['pat_balance'], y=total_data['exposure'], mode='markers+text', marker=dict( size=12, color=list(range(len(total_data))), colorscale='Viridis', showscale=True, colorbar=dict(title='Time (older→newer)') ), text=total_data['year_month'].str[2:7], # Show YY-MM textposition='top center', textfont=dict(size=9), hovertemplate='%{text} PAT: ¥%{x:,.0f} Exposure: ¥%{y:,.0f}<extra></extra>' ) ) # Add diagonal lines for reference max_val = max(total_data['exposure'].max(), abs(total_data['pat_balance']).max()) min_val = min(total_data['pat_balance'].min(), 0) # 100% line (Exposure = |PAT|) fig4.add_trace( go.Scatter( x=[min_val, max_val], y=[abs(min_val), max_val], mode='lines', line=dict(dash='dash', color='#f59e0b', width=1), name='100% (Exposure = |PAT|)', hoverinfo='skip' ) ) # Zero PAT line fig4.add_vline(x=0, line_dash="dot", line_color="gray", opacity=0.5) # Add risk zones annotation fig4.add_annotation( x=total_data['pat_balance'].max() * 0.8, y=total_data['exposure'].max() * 0.3, text="Lower Risk Zone (High PAT, Low Exposure)", showarrow=False, font=dict(color='#22c55e', size=11), bgcolor='rgba(255,255,255,0.8)' ) fig4.add_annotation( x=total_data['pat_balance'].min() * 0.5 if total_data['pat_balance'].min() < 0 else 0, y=total_data['exposure'].max() * 0.8, text="Higher Risk Zone (Low/Negative PAT, High Exposure)", showarrow=False, font=dict(color='#ef4444', size=11), bgcolor='rgba(255,255,255,0.8)' ) fig4.update_layout( title=dict( text='Risk-Balance Matrix Each point = one month (color: time progression)', x=0.5, xanchor='center' ), xaxis_title='PAT Balance (Cash Reserve) - JPY', yaxis_title='Margin Exposure - JPY', xaxis=dict(tickformat=',.0f', tickprefix='¥', zeroline=True), yaxis=dict(tickformat=',.0f', tickprefix='¥'), template='plotly_white', height=600, legend=dict( orientation='h', yanchor='bottom', y=-0.15, xanchor='center', x=0.5 ) ) fig4.show() ``` ## Summary Statistics ```{python} #| label: summary-stats # Calculate summary statistics summary = total_by_month.describe() print("=== Risk Exposure Summary ===\n") print(f"Analysis Period: {total_by_month['year_month'].min()} to {total_by_month['year_month'].max()}") print(f"Total Months: {len(total_by_month)}") print() print("PAT Balance (Cash Reserve):") print(f" Current: ¥{total_by_month['pat_balance'].iloc[-1]:,.0f}") print(f" Maximum: ¥{total_by_month['pat_balance'].max():,.0f}") print(f" Minimum: ¥{total_by_month['pat_balance'].min():,.0f}") print() print("Exposure (Margin Position):") print(f" Current: ¥{total_by_month['exposure'].iloc[-1]:,.0f}") print(f" Maximum: ¥{total_by_month['exposure'].max():,.0f}") print(f" Minimum: ¥{total_by_month['exposure'].min():,.0f}") print() print("Exposure Ratio:") valid_ratios = total_by_month['exposure_ratio'].dropna() if len(valid_ratios) > 0: print(f" Current: {total_by_month['exposure_ratio'].iloc[-1]:.1f}%") print(f" Maximum: {valid_ratios.max():.1f}%") print(f" Average: {valid_ratios.mean():.1f}%") ``` ## Key Insights 1. **Leverage Dynamics**: The exposure ratio shows how aggressively margin is used relative to available cash reserve. 2. **Risk Periods**: Months where exposure significantly exceeds PAT balance represent higher risk periods. 3. **Broker Distribution**: The stacked area charts reveal which broker carries more exposure at any given time. 4. **Trajectory**: The scatter plot's color gradient shows the portfolio's risk trajectory over time. ## Design Decisions 1. **Dual Metrics**: Showing both absolute values (JPY) and ratio (%) provides complete picture of risk. 2. **Interactive Plotly**: Hover details allow precise value inspection at any point. 3. **Color Coding**: Green for cash/profit, Red for exposure/risk - intuitive color associations. 4. **Reference Lines**: 100% and 200% ratio lines provide quick risk assessment benchmarks. 5. **Time-Colored Scatter**: Viridis colorscale shows risk evolution trajectory clearly. --- *This post is part of the [MakeoverMonday](https://www.makeovermonday.co.uk/) weekly data visualization project.* :::{.callout-caution collapse="false" appearance="minimal" icon="false"} ## Disclaimer ::: {style="font-size: 0.85em; color: #64748b; line-height: 1.6;"} This analysis is for educational and practice purposes only. The figures shown are from personal trading records and should not be considered as investment advice. Margin trading involves significant risk of loss. ::: :::