Maximizing Profitability
Leveraging sophisticated machine learning models and technical indicators to identify high-probability trading opportunities, aiming to enhance profitability through data-driven insights.
x
L
O
A
D
I
N
G
Built to reduce trading risk while maximizing return potential over time
Objectives
Fully Automated Trading
From pulling market data and calculating indicators to executing trades and generating reports—eliminating the need for manual intervention. This hands-free approach ensures consistent operation and minimizes human error.
Risk-Adjusted Trading
Each trade is executed in line with predefined risk tolerance levels, ensuring that trading decisions are aligned with the risk profile of each account. This built-in risk management mitigates the chances of substantial losses.
Scalable Architecture
Designed to accommodate growth, XLR8 can efficiently scale to handle an increasing number of trading accounts, market data sources, and asset classes without sacrificing performance or accuracy.
Features
Market Data
Integration
Connects directly with real-time market data providers to ensure accurate data retrieval. The data is processed and stored using a scalable infrastructure.
Advanced Tech
Indicators
Uses raw market data to publish technical indicators. These indicators, crafted from a deep understanding of market behavior, go beyond standard commercial and traditional indicators to recommend profitable trading decisions.
ML-Powered
Decisions
Leverages a powerful machine learning ensemble model that analyzes the system technical indicators. This model generates trade signals and adapts dynamically to changing market conditions.
Automated
Trading
Automates trade execution by integrating with a brokerage platform (Interactive Brokers). The system allocates trades based on each account’s available funds and risk profile, ensuring consistency in execution without manual intervention.
Multiple Accounts
Mgmt.
Designed to manage multiple trading accounts, each with distinct risk profiles. The system tracks and reports performance metrics on an account-by-account basis, offering a clear view of how each account is performing.
4.7Million +
Installation
Performance Metrics
| Capital
%Win R Multiple
%Win R Multiple = $ Profit of Winning Trade / $Risk (R)
If we invest $100K in 2011, we observe the following capital metrics (FAQ.10 shows how XLR8 optimizes revenue)
Table.1: XLR8 revenue simulated for the period between 4/2011 and 9/2024
- The years highlighted in yellow represent the training period, while those in gray indicate the testing period.
- The $ risk per trade increases over time as trading capital grows, driven by the accumulation of profits.
- Taxes are submitted in April, considering 35% deductions from the revenue generated in the previous year.
System Performance
Performance Metrics | Profitability
After-Tax Annualized Return of Investment (ATAROI):
- Measures the compounded rate of return achieved by the system over a year.
- Standardizes returns to an annual period
- Makes it easier to compare performance across different strategies and trading systems.
Profit Factor (PF):
- Helps assess the system's overall profitability
- Compares the amount gained from winning trades to the amount lost from losing trades
If we invest $100K in 2011, we observe the following profitability metrics
Table.2: XLR8 ROI simulated for the period between 4/2011 and 9/2024:
- The After-Tax Annualized Return of Investment (ATAROI) for XLR8 is projected to be 36.5%.
- Estimated taxes are calculated based on the annual revenue for the current period, which differs from the tax reporting in Table.1, where taxes were based on the previous year's figures.
System Performance
Performance Metrics | Risk
Drawdown (DD) & Max Drawdown (MDD):
Ulcer Index (UI)
- Volatility-based risk metric that measures the severity and duration of drawdowns in the equity curve.
- Focuses exclusively on downward risk
- Lower Ulcer Index indicates smoother equity growth with fewer and less severe drawdowns
- The drawdown in 3/2020 (-36.4%) was more severe than the drawdown in February 2016 (-29.9%). However, the Ulcer Index in 3/2020 (19.22) was lower than in February 2016 (27.03), indicating that despite the deeper drawdown in 2020, XLR8 recovered more efficiently compared to 2016.
System Performance
Benchmark Metrics | Profitability
As shown in Table.3, the XLR8 system’s After-Tax Annualized Return on Investment (ATAROI) of 36.49% over a 14-year period demonstrates its strong outperformance compared to traditional indices such as the SPY (10.06%), QQQ (16.6%), and BRK (11.72%) over the same period.
With its significantly higher return on investment, XLR8 effectively capitalizes on market opportunities, especially in volatile environments. Specifically, XLR8’s ROI is projected to be 27.1 times higher than SPY, 10.1 times higher than QQQ, and 20.7 times higher than BRK.
Appendix.4 offers a more conservative analysis by adjusting the return rates from the training period to align with the performance observed during testing. These results project XLR8’s ROI to be 4.95 times higher than SPY, 1.84 times higher than QQQ, and 3.35 times higher than BRK over the 14-year period.
Table.3: XLR8 ROI benchmarks for the period between 4/2011 and 9/2024:
System Performance
Benchmark Metrics | Risk
High-reward trading strategies often come with elevated risk levels. In the case of XLR8, despite its significantly higher returns compared to SPY, QQQ, and BRK, its MDD remains competitive during the 2011–2024 period.
For instance, in 2020, during the COVID-19 market shock, XLR8 experienced a MDD of -36.4%, which, while slightly higher than SPY (-34.1%), QQQ (-28.6%), and BRK (-29.6%), was accompanied by a significantly larger annual return for XLR8 (142.64%) compared to SPY (16.16%), QQQ (47.57%), and BRK (2.37%).
Moreover, as shown in Figure.2, the UI illustrates XLR8’s ability to recover from drawdown losses more swiftly than the benchmarks. Despite having the highest MDD in 2020, XLR8 recorded the lowest UI at 5.9, compared to 7.5 for SPY, 5.9 for QQQ, and 11.3 for BRK. This suggests that, while XLR8 faced deeper drawdowns, it rebounded more efficiently, resulting in smoother long-term equity growth. This trend is also evident in the years between 2018 and 2021.
Figure.2: Annual Avg. Ulcer Index and Max Drawdown Benchmark Metrics
- At the foundation of XLR8 is the ability to pull real-time and historical market data from external providers, such as Polygon.
- Data is continuously fetched using Python-based integrations with the API. The real-time data is ingested and processed in an AWS RDS Oracle database for further analysis.
- The system uses scalable Extract Transform Load (ETL) processes to ensure data is cleaned and structured to feed downstream data transformation and technical indicators.
- Once market data is ingested, XLR8 applies a series of transformations to develop customized and advanced technical indicators that form the basis of the system's trading strategies. These indicators are split into four categories:
- Exponential Moving Averages (EMA): These are used to smooth out price data and identify trends over different time horizons.
- Trends: to track price movements to determine uptrends, downtrends, or periods of consolidation.
- Support and Resistance: to detect price levels to detect reversal points or breakout zones.
- Statistical Indicators: These include volatility measures and statistical metrics to gauge market momentum and strength.
- The indicators are developed using Python libraries and custom-built algorithms to extract insights from raw market data. These metrics are then normalized and passed on as features for the machine learning models.
- At the core of the trading decision-making process is a two-layer machine learning ensemble model to generate trade signals that predict the likelihood of profitable trades.
- The technical indicators and historical market data serve as input features for the machine learning model.
- The ensemble model is structured in two layers:
- Layer 1: set of 36 individual classifiers that make predictions based on 90 different market features.
- Layer 2: consolidates the predictions from Layer 1 to generate trading signals.
- Once trade signals are generated, XLR8 moves to the trade execution phase.
- The system is fully integrated with Interactive Brokers (IB) via their Python API, allowing for seamless, real-time trade execution.
- Trades are executed in alignment with each account’s risk profile, ensuring that exposure is kept within predefined limits. The system calculates the trade size based on available funds and the overall risk tolerance of the account.
- Automated monitoring systems are in place to detect any issues during trade execution, including connectivity issues or order rejection.
- The reporting layer of XLR8 aggregates data and provides real-time analytics on both account performance and individual trades. This includes capital allocation across trades and accounts, ensuring that users can track how funds are being utilized.
- XLR8 tracks the performance of individual trades, strategies, and overall portfolio performance in comparison to market indices like SPY and QQQ.
- Reports also include risk exposure, maximum drawdown, and other risk-adjusted metrics, giving traders clear insights into how risk is being managed.
- All reports are published to AWS S3, making them accessible to users via cloud-based dashboards, including Google Sheets integrations. FAQ.14 adds more details on the XLR8’s reporting tool.
- This google sheet report includes XLR8’s daily and monthly dashboard. Appendix.5 shows snapshots for the daily and monthly reporting dashboard.
About Me
Hi, I’m Nihad. I’ve spent the past 20 years working in analytics, where I’ve built systems and solutions that truly make an impact. I’ve always been fascinated by how data can drive better decisions, and throughout my career, I’ve refined my ability to turn complex data into meaningful insights that fuel innovation and strategy.
In 2014, I decided to take my love for data in a new direction—algorithmic trading. What started as a passion project quickly became an exciting challenge that brought together my interests in data science, finance, and technology. Teaching myself everything from scratch, I created XLR8, an advanced trading system that uses machine learning and technical analysis to generate reliable, data-driven signals.
I’m constantly looking for ways to learn and grow. To deepen my knowledge of financial systems, I pursued a Financial Engineering Master’s program and have connected with incredible thought leaders in the trading space. My mission now is to keep refining XLR8, explore new opportunities in trading, and share my journey with others who are curious about data and trading. Beyond the numbers and code, I’m passionate about helping others grow. Whether it’s mentoring junior analysts, sharing my experiences, or building tools that make data-driven decision-making more accessible, I want to inspire others to see what’s possible.
If you’re interested in the world of algorithmic trading or data science, let’s connect. I’d love to share ideas, learn from your perspective, and explore what’s possible together.
