19 min
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.
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.
Why it matters: ML enables trading systems that adapt and improve, rather than following static rules.
In trading: Deep learning models analyze thousands of market variables simultaneously to predict price movements or identify trading opportunities.
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.
In trading: A trading model might predict price direction, estimate volatility, or score the likelihood of a signal working out.
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.
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.
In trading: A strategy that looks amazing in backtesting but fails live. The model memorized the past rather than learning generalizable patterns.
In trading: A model so basic it misses obvious patterns-like trying to predict market direction using only the day of the week.
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.
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.
Example: "BTC Funding flip detected. Historically bullish 67% of the time. Current bias: moderately bullish."
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.
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.
In AI context: HFT uses algorithms, sometimes AI, but is defined by speed rather than intelligence.
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.
In AI context: AI can identify and track smart money wallets, alerting when they accumulate or distribute.
In AI context: AI-powered market making uses ML to optimize pricing and inventory management.
In AI context: AI analyzes order flow patterns to predict short-term price movements or detect large player activity.
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.
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: Cryptocurrency moving to or from exchanges. Inflows suggest selling preparation; outflows suggest accumulation.
Why it matters: Large flow movements often precede significant price action.
In AI context: AI tracks whale wallet movements to identify smart money activity.
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.
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.
Interpretation:
Above 70: Overbought (potential reversal down)
Below 30: Oversold (potential reversal up)
Why it matters: Common AI input feature for mean reversion strategies.
Why it matters: AI uses MACD crosses and divergences as trading signals.
Why it matters: AI uses Bollinger Band positioning for volatility assessment and mean reversion signals.
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.
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.
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.
Interpretation: 1.5 is decent, 2.0 is good, 3.0+ is excellent.
Formula: (Win% × Avg Win) - (Loss% × Avg Loss)
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 difference between expected trade price and actual execution price.
Why it matters: Slippage eats into profits. More pronounced in low-liquidity conditions.
Definition: The difference between the best buy (bid) and best sell (ask) prices.
Why it matters: You pay the spread on every trade. Narrower spreads = lower trading costs.
Definition: The list of all pending buy and sell orders at various prices.
Why it matters: Order book analysis reveals support/resistance and potential large moves.
Definition: The volume of orders at different price levels. Deep markets can absorb large orders without price impact.
Why it matters: Depth indicates how easy it is to execute large positions.
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: Testing a strategy on historical data to evaluate performance.
Why it matters: Backtesting validates strategies but can be misleading if overfit.
Definition: A notification that trading conditions of interest have occurred.
Definition: A visual interface displaying multiple metrics and information sources in one view.
Definition: A record of all trades with entry/exit details, reasoning, and outcomes.
Terms for measuring trading success.
Formula: (Gain - Cost) / Cost × 100
Definition: Yearly percentage gain or loss on capital.
Why it matters: Allows comparison across different time periods.
Definition: Average annual return accounting for compounding over multiple years.
Why it matters: More accurate than simple average for multi-year performance.
Definition: The largest peak-to-trough decline in account value.
Why it matters: Indicates worst-case scenario you might experience.
Definition: Annual return divided by maximum drawdown. Measures return vs. worst-case risk.
Why it matters: Higher is better. Shows how much return you're getting per unit of drawdown risk.
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
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
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.
AI platforms, bots, and systematic signal workflows.
Perp funding, OI, and liquidation context
Discover the top AI-powered tools for crypto trading in 2026. From signal generators to portfolio managers, find the tools that fit your strategy.