Mean Reversion
Market-Neutral
Survivorship-Bias Free
Single-Factor
Top 30 Market Cap
Tier 2
Mean Reversion is a basket-based signal diversified across different windows to capture reversal dynamics and statistical arbitrage opportunities in crypto assets.
The asset universe consists of the most liquid and actively traded assets, identified on rolling basis - various techniques employed to keep it both stable and relevant, as well as survivorship-bias free.
To balance each asset's risk contribution, positions are scaled according to the inverse of their rolling volatility.
The factor is available (point-in-time) with hourly updates, 5 minutes past the hour (UTC).
| Benchmark (BTC) | Portfolio | |
|---|---|---|
0.92 | 1.21 | |
45.2% | 41.6% | |
0.00 | 0.38 | |
1.00 | 0.04 | |
-76.6% | -33.8% | |
60.7% | 33.3% | |
0% | 31.4% | |
100% | -1.3% | |
100% | 194.3% |
Sharpe Ratio
Rolling 12 MonthsBeta
Rolling 12 Months - Benchmark: BTCNet Exposure of Portfolio
Portfolio Variants
Current Weights
4/21/2026, 11:03:09 PM (UTC)
The portfolio is designed to be market-neutral on a volatilty-adjusted basis, therefore it may have net directional exposure on any given day. The weighting mechanism ensures that volatility-weighted directional exposure is negligible on the portfolio level.
| Ticker | Name | Arrival Price | Weight | Date |
|---|---|---|---|---|
| BTC | Bitcoin | 75628.352 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| ETH | Ethereum | 2311.375 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| SOL | Solana | 85.309 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| XRP | Ripple | 1.421 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| DOGE | Dogecoin | 0.09443573933214607 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| SUI | Sui | 0.938 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| BNB | Binance Coin | 628.815 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| ADA | Cardano | 0.249 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| LTC | Litecoin | 55.122 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| LINK | Chainlink | 9.301 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| ONDO | ONDO | 0.255 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| AVAX | Avalanche | 9.296 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| DOT | Polkadot | 1.27 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| BCH | Bitcoin Cash | 445.659 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| PEPE | PEPE | 0.000003769455037216 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| HBAR | Hedera | 0.08990358525224852 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| TRX | TRON | 0.334 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| AAVE | Aave | 90.928 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| TAO | Bittensor | 242.918 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| NEAR | NEAR Protocol | 1.37 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| UNI | Uniswap | 3.241 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| TON | Toncoin | 1.351 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| ETC | Ethereum Classic | 8.361 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| XLM | Stellar | 0.177 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| SHIB | SHIB | 0.000006036812002688 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| HYPE | HYPE | 39.497 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| ICP | Internet Computer | 2.439 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| XMR | Monero | 384.934 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| WLFI | WLFI | 0.0784247480108933 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| ASTER | ASTER | 0.676 USD | 0.00% | 2026-04-21T23:03:09.361753+00:00 |
| 0.00% | ||||
Properties
Cross-Sectional | |
Derivative Exchanges | |
400+ Tickers | |
None |
2024-01-01 | |
Daily | |
Hourly, 5 minutes past the hour (UTC) | |
Yes |
Frequently Asked Questions
What is the construction methodology of the model portfolio on display? How do I replicate it?
The portfolio is constructed by applying cross-sectional binning - quantiles. This ensures that the weights are calculated in proportion to the strength of the factor. The assets are then weighted according to the inverse of their rolling volatility, to mitigate the impact of widely different volatilities of digital assets. We provide a replication notebook in the "Resources" section that will produce a very similar portfolio than what's shown on the site.
Is the raw factor data available?
Yes, we serve the raw factor data via the "portfolio/factors" endpoint, please see the API Docs for more information.
What is smoothing? Some factors have by default setting (eg. 10 Day Moving Average) applied. Why is that?
Some of the cross-sectional factors have very high turnover, and despite being highly predictive, the alpha does not survive transaction costs. Smoothing (applying a simple moving average) is a way to reduce the turnover and the impact of transaction costs.
What is the universe of assets? How is it determined? Is it survivorship-bias free?
The universe of assets (there are multiple variants, for example, top 20, 30, 40 market cap digital assets) is the most liquid and actively traded assets, identified on rolling basis - various techniques (volume, open interest, volatility filters) are employed to keep it both stable and relevant. The universe is survivorship-bias free.
What is the rebalancing frequency?
The portfolio is rebalanced daily, with hourly updates, 5 minutes past the hour (UTC), and available point-in-time.
What are the transaction cost / slippage assumptions?
There are 0.05% transaction costs applied on each position adjustment. This is a parameter that can be adjusted in the backtest settings. Slippage, spread is not considered in the simplistic backtest that's on our site. We encourage using an independent backtesting infrastructure to validate performance.
When have been the factors developed? Is there an out-of-sample period?
Our cutoff date for our research process is 2024-01-01, we consider any data onwards as out-of-sample. Many factors have been traded live in some form from 2025-01-01.
Some of the factors look great! Should I pick an individual factor and start trading it?
We recommend using a portfolio of factors to maximize risk-adjusted returns. Alpha from individual factors is simply not consistent enough, and even running a simple arithmetic average of 3-5 relatively orthogonal factors will produce superior results.
Replication & API
Live Weights
Get live weights with the code provided below.
from unravel_client import get_live_weights
UNRAVEL_API_KEY = "null"
portfolio = "mean_reversion.30"
live_weights = get_live_weights(
id=portfolio, api_key=UNRAVEL_API_KEY, smoothing=None, exchange=None
)
print(live_weights)