You're researching AI crypto trading and drowning in jargon. Neural networks. Gradient descent. Reinforcement learning. Sentiment analysis. Every article assumes you already know what these terms mean.
You don't. And that's okay.
This ultimate glossary defines 150+ AI trading terms in plain English. No technical background required. Each definition tells you what the term means and why it matters for your trading.
Bookmark this page. You'll reference it often.
Foundational AI terms you'll encounter constantly.
- Definition: Computer systems designed to perform tasks that typically require human intelligence-pattern recognition, decision-making, learning from experience.
In trading: AI analyzes market data, identifies patterns, generates signals, and can execute trades. It processes information faster than humans and identifies patterns too subtle for human detection.
Why it matters: AI gives traders access to analysis capabilities previously available only to institutions with PhD-level quants.
Definition: A set of step-by-step instructions that tell a computer how to complete a task. Like a recipe, but for computers.
-
In trading: Trading algorithms define rules for when to buy, sell, how much, and under what conditions. "AI algorithm" means those rules involve learning from data rather than just following preset instructions.
-
Why it matters: Understanding that trading bots follow algorithms helps you understand their behavior-they do exactly what they're programmed to do, nothing more.
Definition: A subset of AI where computers learn patterns from data rather than being explicitly programmed with rules.
- In trading: Instead of telling a computer "buy when RSI is below 30," ML learns which conditions have historically preceded profitable trades and applies those patterns.
Why it matters: ML enables trading systems that adapt and improve, rather than following static rules.
- Definition: Machine learning using neural networks with many layers ("deep" networks). Good at finding complex patterns in large datasets.
In trading: Deep learning models analyze thousands of market variables simultaneously to predict price movements or identify trading opportunities.
- Why it matters: Deep learning powers the most sophisticated trading AI, but also the most "black box" (hard to understand why it makes decisions).
-
Definition: An AI structure loosely inspired by the human brain. Composed of interconnected nodes that process information in layers.
-
In trading: Neural networks process market data through multiple layers, each extracting increasingly abstract patterns, ultimately outputting predictions or signals.
-
Why it matters: When someone says their AI uses "neural networks," they mean a specific type of machine learning architecture capable of complex pattern recognition.
Definition: AI that understands and processes human language-text and speech.
In trading: NLP analyzes news articles, social media posts, earnings calls, and regulatory announcements to gauge market sentiment or extract trading-relevant information.
Why it matters: NLP enables AI to understand textual information that affects markets-something pure numerical analysis misses.
-
Definition: The historical data used to teach machine learning models. The model learns patterns from this data.
-
In trading: Training data typically includes historical prices, volumes, market indicators, and outcomes. The quality and relevance of training data significantly impacts model performance.
Why it matters: "Garbage in, garbage out." AI is only as good as the data it learned from.
- Definition: The mathematical representation created by machine learning. After training, the model encodes learned patterns and can make predictions on new data.
In trading: A trading model might predict price direction, estimate volatility, or score the likelihood of a signal working out.
- Why it matters: When someone says "our AI model predicts X," they're referring to this trained mathematical representation.
Terms describing how AI learns and operates.
Definition: ML where the model learns from labeled examples. Given input data AND correct answers, it learns to predict answers for new inputs.
-
In trading: Showing the model "these conditions led to 5% rallies" (labeled examples), then having it predict which current conditions will lead to rallies.
-
Example: Training on historical data where each day is labeled "price went up" or "price went down."
Definition: ML where the model finds patterns in data without being given correct answers. It discovers structure on its own.
- In trading: Clustering market regimes, identifying unusual patterns, or finding correlations without predefined categories.
Example: AI discovers there are 4 distinct "types" of market conditions without being told what to look for.
Definition: ML where the model learns by trial and error, receiving rewards for good decisions and penalties for bad ones.
In trading: AI learns trading strategies by simulating thousands of trades, being rewarded for profitable decisions and penalized for losses.
- Why it matters: Reinforcement learning can discover strategies humans wouldn't think of, but requires careful design to avoid pathological behaviors.
- Definition: When a model learns training data too well, including noise and random patterns. Performs great on historical data, poorly on new data.
In trading: A strategy that looks amazing in backtesting but fails live. The model memorized the past rather than learning generalizable patterns.
- Why it matters: This is the #1 killer of AI trading strategies. Always verify performance on out-of-sample data.
- Definition: When a model is too simple to capture the underlying patterns in data. Performs poorly on both training and new data.
In trading: A model so basic it misses obvious patterns-like trying to predict market direction using only the day of the week.
- Why it matters: Less common than overfitting, but a sign your model needs more complexity or better features.
-
Definition: An input variable used by the model. Each piece of information the model considers when making predictions.
-
In trading: Features might include RSI values, volume, funding rates, sentiment scores, time of day-any data point the model uses.
-
Why it matters: Feature selection (choosing what data to feed the model) often matters more than the algorithm itself.
-
Definition: Settings that control how a model learns, set before training begins. Different from parameters the model learns from data.
-
In trading: Examples include learning rate, number of layers in a neural network, or how much data to consider.
-
Why it matters: Hyperparameter tuning can significantly improve model performance-it's part of the "art" of ML.
- Definition: Testing a trading strategy on historical data to see how it would have performed.
In trading: "My AI model was backtested over 5 years" means they ran the model against 5 years of historical data to measure performance.
Why it matters: Backtesting is essential for validation but can be misleading if not done properly (see overfitting).
-
Definition: Testing a strategy in real-time with simulated money. Real market conditions, fake capital.
-
In trading: Running your AI signals but tracking results without actual trades. Validates that backtest performance holds in live conditions.
-
Why it matters: Bridges the gap between backtesting and live trading. Reveals execution issues backtests miss.
-
Definition: Testing model performance on data it wasn't trained on. The gold standard for validation.
-
In trading: Train on 2020-2022 data, test on 2023 data. If performance holds, the model likely learned real patterns.
-
Why it matters: The only way to know if your model generalizes vs. just memorized the past.
AI terms specifically used in trading contexts.
- Definition: An alert generated by AI indicating a potential trading opportunity. Can include direction, confidence level, and supporting analysis.
Example: "BTC Funding flip detected. Historically bullish 67% of the time. Current bias: moderately bullish."
- Why it matters: Signals are how AI communicates actionable insights to traders.
- Definition: Software that automatically executes trades based on AI-generated signals or rules. Operates without human intervention for each trade.
Types:
-
Signal-following bots: Execute when AI signals trigger
-
ML-based bots: Make decisions using machine learning models
-
Hybrid bots: Combine rules and ML
-
Why it matters: Bots automate execution but require careful oversight.
-
Definition: Trading based on mathematical models and statistical analysis rather than intuition or fundamental analysis.
-
In AI context: Most AI trading is quantitative-using data-driven models to make decisions.
Why it matters: AI trading is a subset of quantitative trading. Understanding quant principles helps you use AI effectively.
- Definition: Using computer programs to execute trades based on predefined instructions. May or may not involve AI.
Distinction from AI trading:
-
Simple algo: "Buy when price crosses moving average"
-
AI algo: "Buy when model predicts >60% upward probability"
-
Why it matters: Not all algo trading is AI trading, but all AI trading is algorithmic.
- Definition: Trading at extremely high speeds, often executing thousands of trades per second, capitalizing on tiny price differences.
In AI context: HFT uses algorithms, sometimes AI, but is defined by speed rather than intelligence.
- Why it matters: Retail traders can't compete with HFT on speed. AI gives you an edge in analysis, not execution speed.
-
Definition: Using AI (typically NLP) to determine the emotional tone of text-positive, negative, or neutral.
-
In trading: Analyzing news, social media, and forum posts to gauge market sentiment. "Bullish sentiment on Twitter" = more positive than negative posts.
-
Why it matters: Sentiment can precede price movements. AI sentiment analysis processes thousands of sources simultaneously.
- Definition: Large, sophisticated investors or institutions assumed to have informational advantages. Often tracked via on-chain analysis or whale watching.
In AI context: AI can identify and track smart money wallets, alerting when they accumulate or distribute.
- Why it matters: Following smart money (with AI assistance) can provide directional insight.
- Definition: An entity that provides liquidity by continuously offering to buy and sell an asset. Profits from the bid-ask spread.
In AI context: AI-powered market making uses ML to optimize pricing and inventory management.
- Why it matters: Understanding market makers helps interpret order book data and spread dynamics.
- Definition: The stream of buy and sell orders entering the market. Reveals real-time supply and demand.
In AI context: AI analyzes order flow patterns to predict short-term price movements or detect large player activity.
- Why it matters: Order flow is leading information-shows intent before price moves.
Terms related to market data and analysis.
-
Definition: In perpetual swaps, a periodic payment between long and short traders to keep the perpetual price close to spot price.
-
Positive funding: Longs pay shorts (bullish crowd)
-
Negative funding: Shorts pay longs (bearish crowd)
-
Why it matters: Funding rate extremes often signal crowded positioning and potential reversals.
- Definition: The total number of outstanding derivative contracts that haven't been settled. Different from volume.
Rising OI + rising price: New longs entering
Rising OI + falling price: New shorts entering
Falling OI + rising price: Shorts closing
Falling OI + falling price: Longs closing
Why it matters: OI shows conviction and leverage in the market.
-
Definition: Forced closure of a leveraged position when the trader can't meet margin requirements. Creates mechanical buying or selling.
-
Why it matters: Liquidation cascades accelerate price moves. AI detects liquidations to identify potential reversals or momentum.
-
Definition: Information derived directly from blockchain transactions-immutable, transparent, and comprehensive.
-
Examples: Wallet balances, transaction volumes, exchange flows, holder distribution.
Why it matters: On-chain data can't be faked and provides unique insights unavailable in traditional markets.
- Definition: An entity holding a large amount of cryptocurrency-large enough to significantly impact price with their trades.
In AI context: AI tracks whale wallet movements to identify smart money activity.
- Why it matters: Whale movements often provide early signals of accumulation or distribution.
- Definition: The total value of assets deposited in a DeFi protocol. A measure of protocol usage and confidence.
Why it matters: TVL trends indicate DeFi protocol health and can influence token prices.
-
Definition: Ratio comparing current market cap to "realized" cap (value based on last on-chain movement price).
-
High MVRV: Market is above average cost basis-profit-taking likely
-
Low MVRV: Market is below average cost basis-accumulation zone
Why it matters: MVRV helps identify market cycle positioning.
Terms AI uses from traditional technical analysis.
- Definition: The average price over a specified period, plotted as a line on charts. Smooths price data to identify trends.
Common types: SMA (simple), EMA (exponential-weights recent prices more heavily).
Why it matters: AI often uses moving averages as features and for trend identification.
- Definition: Momentum indicator measuring the speed and magnitude of price changes. Ranges from 0-100.
Interpretation:
-
Above 70: Overbought (potential reversal down)
-
Below 30: Oversold (potential reversal up)
-
Why it matters: Common AI input feature for mean reversion strategies.
- Definition: Trend-following momentum indicator showing relationship between two moving averages.
Why it matters: AI uses MACD crosses and divergences as trading signals.
- Definition: Volatility bands placed above and below a moving average. Bands widen when volatility increases, narrow when it decreases.
Why it matters: AI uses Bollinger Band positioning for volatility assessment and mean reversion signals.
- Definition: Price levels where buying (support) or selling (resistance) historically concentrated. Often act as floors and ceilings.
Why it matters: AI identifies S/R levels and signals when price approaches them.
-
Definition: When price moves in one direction while an indicator moves in the opposite direction. Often signals potential reversals.
-
Example: Price makes new high, but RSI makes lower high = bearish divergence.
Why it matters: AI detects divergences that humans might miss across multiple indicators.
-
Definition: Horizontal lines indicating potential support/resistance based on Fibonacci ratios (23.6%, 38.2%, 61.8%, etc.).
-
Why it matters: Many traders use Fibonacci levels, making them self-fulfilling to some degree.
-
Definition: Analysis showing trading volume at different price levels, revealing where most activity occurred.
-
Why it matters: High-volume nodes act as support/resistance. AI uses volume profile for level identification.
Terms related to protecting capital.
-
Definition: The peak-to-trough decline in account value. Maximum drawdown is the largest decline from a peak.
-
Example: Account peaks at $10,000, drops to $8,500, then recovers. Drawdown was 15%.
-
Why it matters: Maximum drawdown indicates worst-case scenario. Essential for risk planning.
-
Definition: Determining how much capital to allocate to each trade. Balances potential gain against potential loss.
-
Common method: Risk 1-2% of account per trade.
-
Why it matters: Position sizing is the primary tool for controlling risk.
Definition: A predetermined price at which you exit a losing trade to limit losses.
- Why it matters: Stops prevent small losses from becoming catastrophic. Essential for risk management.
-
Definition: The ratio of potential loss to potential gain. 1:3 means risking $1 to potentially make $3.
-
Why it matters: With 1:3 R:R, you only need 25%+ win rate to be profitable.
- Definition: Measure of risk-adjusted return. Higher is better. Compares return to volatility.
Formula: (Return - Risk-Free Rate) / Standard Deviation
-
Interpretation: Above 1.0 is acceptable, above 2.0 is good, above 3.0 is excellent (or suspicious).
-
Why it matters: Sharpe Ratio lets you compare strategies accounting for risk taken.
-
Definition: Percentage of trades that are profitable.
-
Important: Win rate alone doesn't determine profitability. A 30% win rate with 4:1 R:R is profitable.
-
Why it matters: Win rate combined with R:R determines expectancy.
- Definition: Gross profits divided by gross losses. Above 1.0 means profitable.
Interpretation: 1.5 is decent, 2.0 is good, 3.0+ is excellent.
- Why it matters: Simple profitability indicator across all trades.
- Definition: Average expected profit per trade over time.
Formula: (Win% × Avg Win) - (Loss% × Avg Loss)
- Why it matters: Positive expectancy = profitable system over time.
-
Definition: Trading with borrowed capital. 10x leverage means your position is 10x your actual capital.
-
Why it matters: Leverage amplifies gains AND losses. Excessive leverage is the #1 account killer.
Terms describing how markets function.
-
Definition: How easily an asset can be bought or sold without significantly affecting price. High liquidity = easy, low liquidity = difficult.
-
Why it matters: Low liquidity causes slippage and execution problems.
-
Definition: The degree of price variation over time. High volatility = large price swings.
-
Why it matters: Volatility affects position sizing, stop distances, and strategy selection.
-
Definition: The current state of market conditions-trending, ranging, volatile, etc.
-
Why it matters: Different strategies work in different regimes. AI can detect regime changes.
Terms describing AI trading platforms and tools.
Definition: A way for software programs to communicate with each other. Exchange APIs let trading software access market data and execute trades.
Why it matters: API access enables automated trading and data retrieval.
-
Definition: Simulated trading with fake money. Real market conditions, no real risk.
-
Why it matters: Essential for testing strategies before risking capital.
Definition: A notification that trading conditions of interest have occurred.
- Why it matters: Alerts enable timely action without constant monitoring.
Definition: A visual interface displaying multiple metrics and information sources in one view.
- Why it matters: Good dashboards improve decision-making by centralizing information.
Definition: A record of all trades with entry/exit details, reasoning, and outcomes.
- Why it matters: Journaling is essential for improvement. AI journals add automated analysis.
Terms for measuring trading success.
- Definition: Profit as a percentage of investment.
Formula: (Gain - Cost) / Cost × 100
- Why it matters: Standard measure of investment performance.
-
Definition: Like Sharpe Ratio, but only penalizes downside volatility, not upside.
-
Why it matters: More relevant for traders who don't mind upside volatility.
Start with: AI, Machine Learning, Signal, Backtest, Overfitting, Funding Rate, Open Interest, Position Sizing, Stop Loss, Win Rate. Master these before expanding vocabulary.
No. You need to understand what they mean for your trading, not how they work mathematically. "Funding rate indicates positioning" is sufficient-you don't need the formula.
Be skeptical when you hear: "guaranteed returns," "100% win rate," "no risk," "passive income." These terms indicate marketing hype, not legitimate AI trading.
You don't need to memorize. Bookmark this glossary and reference it when you encounter unfamiliar terms. Understanding builds through use, not memorization.
Thrive uses standard industry terminology. Signal, Coaching, Journal, and Dashboard in Thrive mean the same as the definitions here. No proprietary jargon to learn.
"AI" itself. Many products labeled "AI" use simple if-then rules. True AI involves learning from data. Ask what kind of AI when evaluating products.
This glossary covered 150+ essential AI trading terms across:
Core AI Concepts: AI, Algorithm, Machine Learning, Deep Learning, Neural Network, NLP, Training Data, Model
- Machine Learning Terms: Supervised/Unsupervised Learning, Overfitting, Features, Backtesting, Out-of-Sample Testing
Trading-Specific AI Terms: AI Signal, Trading Bot, Quant Trading, Sentiment Analysis, Smart Money, Order Flow
-
Data Terms: Funding Rate, Open Interest, Liquidation, On-Chain Data, Exchange Flow, Whale, TVL, MVRV
-
Technical Analysis Terms: Moving Average, RSI, MACD, Bollinger Bands, Support/Resistance, Divergence
Risk Management Terms: Drawdown, Position Sizing, Stop Loss, R:R, Sharpe Ratio, Win Rate, Profit Factor
- Market Structure Terms: Liquidity, Slippage, Spread, Order Book, Volatility, Market Regime
Performance Metrics: ROI, CAGR, Max Drawdown, Calmar Ratio, Sortino Ratio
Bookmark this page. Return whenever you encounter unfamiliar terms.
Now that you understand the terminology, experience AI trading in action:
✅ Real AI Signals - See funding rate alerts, liquidation cascades, and whale movements in action
✅ Clear Terminology - No confusing jargon-everything explained as you learn
✅ Trade Journaling - Apply position sizing, track win rate, measure drawdown
✅ Performance Metrics - See Sharpe Ratio, Profit Factor, and Expectancy for YOUR trades
Understanding the terms is step one. Using them is step two.
→ Start AI Trading with Thrive
