This project showcases the end-to-end development of a quantitative investment strategy, starting from academic research and ending with a live performance dashboard. The strategy is based on the long-short, zero-investment methodology referenced in “LIU, Y., TSYVINSKI, A. and WU, X. (2022), Common Risk Factors in Cryptocurrency. The Journal of Finance, 77: 1133-1177”.
The original study found that cryptocurrency returns can be explained by three core factors: the overall Market, asset Size (Market Cap), and Momentum (past performance).
My process involved replicating this research and then building upon it to create a robust, tradable model. The workflow is presented in the five dashboards below.
Project Phase: Foundational Research & Discovery
This is the very first step. The goal here was to replicate the academic research and see if the factors mentioned in the paper—like an asset's size (Market Cap) or its recent performance (Momentum)—actually had any predictive power on their own.
Page 1: Growth & Drawdown: This page is a visual "horse race." It charts the historical performance of portfolios built using each single factor against the benchmark (Bitcoin). This allowed me to immediately see which factors grew wealth and which ones failed. It also shows the "drawdown," or the largest historical losses, for each one.
Page 2: Returns & Risk: This is the statistical "scorecard" for the race. It shows tables of key metrics like volatility, annual returns, and, most importantly, risk-adjusted ratios. This page helped me quantify which factor was the most efficient, delivering the best performance for the amount of risk taken.
Page 3: Statistical Analysis: This page validates the quality of the winning factor's returns. It shows a distribution chart (a histogram) to see the typical weekly wins and losses and confirms if the positive performance was statistically real or just random luck.
Project Phase: Strategy Development & Model "Bake-Off"
Once I identified promising factors, the next step was to find the best way to combine them into a single, robust strategy. This dashboard is a "bake-off" where I tested different portfolio-building theories (like Global Minimum Variance, Risk Parity, etc.) against each other.
Page 1: Growth & Drawdown: Another visual "horse race," but this time it compares the different portfolio models to see which one creates the best trade-off between growth and stability.
Page 2: Returns & Risk: The scorecard for this bake-off. It provides the hard metrics (risk, return, and ratios) for each of the combined portfolio models, allowing me to pick the most efficient and robust "winner."
Page 3: Statistical Analysis: A statistical deep-dive into each model, showing their return distributions. This helped me confirm the quality and consistency of the model I ultimately chose.
Project Phase: Final Backtest & Validation
After selecting the winning model from the bake-off, I created this dashboard. It shows the final, polished historical backtest of the strategy, which I named "Alt Beta." This is the final "product" of the research phase.
Page 1: Growth & Drawdown: This is the final historical equity curve, showing how "Alt Beta" would have performed against the benchmark (Bitcoin). It includes a shaded "out-of-sample" period, which is a more honest test using data the model wasn't designed on.
Page 2: Returns & Risk: The final report card. It directly compares "Alt Beta" to Bitcoin on all key risk and return metrics. The goal was to prove its value, especially its ability to dramatically reduce risk and large losses.
Page 3: Statistical Analysis: The final statistical validation for "Alt Beta." It confirms that its "edge" (positive performance) is statistically significant and not just a fluke.
Project Phase: Live Operations & Execution
This dashboard marks the critical shift from historical research to live trading. This is the strategy's "control panel." It doesn't look at the past; it shows what's happening right now.
This is a single-page dashboard that runs the "Alt Beta" algorithm on the most recent market data. Its main job is to generate the exact trading signals for the day, showing a clear list of which assets to "BUY" and which to "SELL" (short). It also tracks all my open positions, their current status, and the portfolio's live profit or loss.
Project Phase: Live Monitoring & The Verdict
This is the "moment of truth." This dashboard tracks the actual, real-world performance of the "Alt Beta" strategy after it was launched. This is all new, "out-of-sample" data that the model has never seen before.
Page 1: Return Measures: Tracks the live profit and loss. It shows a growth chart of my strategy's live funds versus the benchmark and a table of its actual return metrics.
Page 2: Risk Measures: Tracks the strategy's live risk. It monitors its real-world volatility and drawdowns, answering the question: "Is the strategy behaving as calmly as the backtest predicted?"
Page 3: Correlation Analysis: This is a crucial page. It tracks how closely the strategy's live returns follow the benchmark (Bitcoin). The goal is to see a very low correlation, which proves the strategy is a true diversification tool.
Page 4: Strategy Vs BTC: A simple, four-quadrant summary chart that directly compares my "Alt Beta" strategy against Bitcoin on the four key metrics: Growth, Volatility, Drawdown, and Correlation.
Page 5: Factors Vs BTC: This is a diagnostic page. It uses the same four-quadrant layout to break down the performance of the individual factors that make up my strategy. This helps me understand why the overall strategy is performing the way it is.