Alpha Framework
The alpha framework is the set of conceptual approaches TRADEOS.tech uses to generate profitable trade signals. This section describes the categories and principles involved — not the specific implementations, which are proprietary.
What Is Alpha?
In quantitative trading, alpha refers to a systematic, repeatable edge — a pattern in market data that predicts future price movement with better-than-random accuracy. Finding and exploiting alpha is the core challenge of systematic trading.
Alpha sources degrade over time as markets adapt. A good systematic trading system:
- Uses multiple uncorrelated alpha sources (so degradation of one doesn't kill performance)
- Monitors signal quality continuously and discards degraded signals
- Adapts to changing market regimes rather than assuming static conditions
TRADEOS.tech is built around all three of these principles.
Alpha Categories
TRADEOS.tech generates signals across five broad alpha categories:
1. Momentum / Trend Following
Trend-following captures persistent directional price movement across multiple timeframes. Crypto markets exhibit strong trending behavior, particularly during macro-driven bull/bear cycles. Momentum signals are most effective in trending regimes and are suppressed during ranging or choppy conditions.
2. Mean Reversion
Mean reversion exploits the tendency of prices to revert to equilibrium after short-term dislocations. Most effective in ranging regimes and during microstructure-driven price spikes. TRADEOS.tech applies mean-reversion signals selectively — they are regime-gated to avoid fighting genuine trends.
3. Order Flow / Microstructure
Order flow signals read the footprint of large participants in the market. Hawkes process modeling of trade arrival rates, order book imbalance, and volume-at-price analysis can predict short-term price direction. These signals have short decay windows and are used for timing entries within a broader directional thesis.
4. On-Chain Signals
Crypto markets have a unique data layer: the blockchain. On-chain signals include exchange inflow/outflow patterns, large wallet movements, stablecoin flows, and miner behavior. These tend to be medium-to-long term signals with lower frequency but high signal-to-noise ratios when properly filtered.
5. Carry and Funding
Perpetual futures markets pay or charge funding rates that reflect the cost of leverage in the market. When funding rates reach extremes, they create both a carry opportunity and a contrarian signal (excessive leverage tends to precede reversals). TRADEOS.tech tracks funding rates across exchanges and incorporates them as both a carry signal and a crowding indicator.
Signal Quality Management
TRADEOS.tech does not assume that signals which worked historically will continue to work. The system continuously monitors:
- IC (Information Coefficient): rolling correlation between signal prediction and actual outcome
- Signal-to-noise ratio: are signal scores distinguishing between profitable and unprofitable setups?
- Regime fit: is the current market regime one where this signal type has historically worked?
Signals that fail quality thresholds are gated automatically until they recover. This is one of the most important — and most often overlooked — aspects of running a live systematic trading system.
What Is Not Disclosed
The specific algorithms, model architectures, parameter values, ensemble methods, and calibration procedures used in the TRADEOS.tech alpha framework are proprietary and are not disclosed in this documentation. The categories and principles above describe the conceptual approach. The actual implementation contains significant additional depth and engineering not described here.