Abstract
We consider the problem of constructing a portfolio that combines traditional financial assets with crypto assets. We show that despite the documented attributes of crypto assets, such as high volatility, heavy tails, excess kurtosis, and skewness, a simple extension of traditional risk allocation provides robust solutions for integrating these emerging assets into broader investment strategies. Examination of the risk allocation holdings suggests an even simpler method, analogous to the traditional 60/40 stocks/bonds allocation, involving a fixed allocation to crypto and traditional assets, dynamically diluted with cash to achieve a target risk level.
Introduction
In quantitative finance, crypto assets are frequently characterized as unpredictable and high-risk, often perceived as requiring specialized, complex machine learning models for effective management. However, research by Kasper Johansson and Stephen Boyd suggests that applying classic risk-management principles can integrate crypto into traditional portfolios using surprisingly simple methods that outperform more complex alternatives.
The Statistical Challenge of Crypto
Institutional and retail investors often hesitate to engage with crypto due to its challenging statistical profile. Crypto assets exhibit "heavy tails" (frequent extreme price moves), high kurtosis (extreme outliers), and massive drawdowns. For instance, Ethereum has historically experienced a 20x drop in value. These "stylized facts" often lead researchers to believe that non-linear machine learning or deep reinforcement learning is necessary to manage risk.
As shown in the Q-Q plots above, crypto returns deviate significantly from a Gaussian distribution, particularly at the extremes. This confirms the presence of heavy tails that traditional models must account for.
Datasets and Asset Behavior
The study utilizes daily close prices from September 2017 to September 2024, covering Bitcoin (BTC), Ethereum (ETH), and four daily-traded industry portfolios (Consumer, Manufacturing, Tech, and Health). The S&P 500 serves as a benchmark for comparison.
Methodology: From Optimization to Simplicity
The authors compare three primary approaches:
- Constrained Risk Allocation (CRA): An extension of risk parity where each asset contributes equally to the total portfolio risk, subject to a 10% maximum weight limit on crypto.
- Industries Only Baseline: A portfolio of traditional industry sectors and cash.
- Dynamically Diluted 90/10 (DD90/10): A heuristic approach maintaining a fixed 90% traditional and 10% crypto split, using cash as a "volume knob" to dilute risk based on current volatility estimates (e.g., EWMA).
Key Results
The backtests demonstrate that adding a small crypto allocation significantly improves risk-adjusted returns. The simplest method, DD90/10, achieved the highest Sharpe Ratio.
| Portfolio | Annualized Return | Volatility | Sharpe Ratio | Max Drawdown |
|---|---|---|---|---|
| Industries Only | 6.0% | 8.2% | 0.73 | 12.5% |
| CRA Combined | 8.2% | 8.2% | 1.00 | 19.6% |
| DD90/10 (EWMA) | 10.4% | 9.8% | 1.06 | 19.9% |
Conclusion
Crypto assets may be "louder" than traditional assets, but they do not break the fundamental logic of risk allocation. By accounting for higher volatility through cash dilution and maintaining a small, fixed crypto tilt, investors can capture significant upside without succumbing to extreme downside risks. Simple, robust frameworks often prove more effective than complex models in the face of crypto's unique statistical challenges.
Summary updated