volatility-modeling
Volatility estimation, forecasting, and regime classification using GARCH, EWMA, realized volatility, and volatility cones
- Installs
- 499
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/agiprolabs/claude-trading-skills --skill volatility-modeling- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads volatility-modeling automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
Volatility Modeling
Volatility — the magnitude of price fluctuations — is arguably the single most important quantity in trading. It drives position sizing, stop placement, option pricing, and regime detection. This skill covers estimation, forecasting, and practical application of volatility in crypto markets.
Why Volatility Matters
| Use Case | How Volatility Is Used |
|---|---|
| Position sizing | Scale position inversely with vol so each trade risks a consistent dollar amount |
| Stop placement | ATR-based stops widen in high-vol regimes, tighten in low-vol |
| Strategy selection | Mean-reversion works in low vol; momentum works in high vol |
| Risk budgeting | Vol-target portfolios maintain constant portfolio-level risk |
| Regime detection | Vol regime shifts signal changing market dynamics |
| Option pricing | Implied vs realized vol gap creates trading opportunities |
Types of Volatility
Historical (Realized) Volatility
Computed from observed past returns. The most common and directly measurable form. Multiple estimators exist with different statistical efficiency.
Implied Volatility
Derived from option prices via Black-Scholes or similar models. Limited in crypto DeFi where liquid options markets are sparse, but available on Deribit for BTC/ETH.
Forecast Volatility
Predicted future volatility from models like EWMA or GARCH. Used for forward-looking position sizing and risk budgets.
Estimation Methods
1. Close-to-Close (Standard Deviation of Log Returns)
The simplest estimator. Compute the standard deviation of log returns and annualize.
import numpy as np
log_returns = np.log(closes[1:] / closes[:-1])
vol_daily = np.std(log_returns, ddof=1)
vol_annual = vol_daily * np.sqrt(365) # crypto trades 365 days
- Pros: Simple, widely understood.
- Cons: Uses only close prices — ignores intraday range.
2. Parkinson (High-Low Range)
Uses the daily high-low range, which is ~5x more statistically efficient than close-to-close.
hl_ratio = np.log(highs / lows)
vol_parkinson = np.sqrt(np.mean(hl_ratio**2) / (4 * np.log(2))) * np.sqrt(365)
- Pros: More efficient, captures intraday moves.
- Cons: Downward bias with discrete sampling; ignores close-to-close jumps.
3. Garman-Klass (OHLC)
The most efficient single-day OHLC estimator.
hl = np.log(highs / lows)
co = np.log(closes / opens)
gk = np.mean(0.5 * hl**2 - (2 * np.log(2) - 1) * co**2)
vol_gk = np.sqrt(gk) * np.sqrt(365)
- Pros: Best efficiency among OHLC estimators.
- Cons: Assumes no drift; sensitive to opening gaps.
4. Yang-Zhang
Combines overnight (close-to-open) and open-to-close components. Handles gaps properly. Less relevant for 24/7 crypto but useful for tokens with sporadic trading.
5. EWMA (Exponentially Weighted Moving Average)
RiskMetrics approach — no parameters to estimate beyond λ.
lam = 0.94 # RiskMetrics default for daily
ewma_var = np.zeros(len(returns))
ewma_var[0] = returns[0] ** 2
for t in range(1, len(returns)):
ewma_var[t] = lam * ewma_var[t - 1] + (1 - lam) * returns[t - 1] ** 2
vol_ewma = np.sqrt(ewma_var) * np.sqrt(365)
- λ = 0.94 for daily data (RiskMetrics).
- λ = 0.97 for weekly data.
- Higher λ → smoother, slower reaction to new information.
6. GARCH(1,1)
The workhorse autoregressive volatility model. Captures volatility clustering.
σ²_t = ω + α · r²_{t-1} + β · σ²_{t-1}
- ω: long-run variance weight.
- α: reaction to recent shock (typically 0.05–0.15 for crypto).
- β: persistence (typically 0.80–0.90 for crypto).
- α + β < 1: stationarity constraint.
- Long-run variance: ω / (1 − α − β).
Estimated via maximum likelihood. See references/estimators.md for details.
Volatility Cones
Volatility cones show the percentile distribution of realized volatility at different lookback windows, revealing whether current vol is historically high or low.
Construction
- Get 1+ years of daily data.
- For each lookback window (5, 10, 20, 60, 120 days):
- Compute rolling realized volatility.
- Extract percentiles: 5th, 25th, 50th, 75th, 95th.
- Plot percentiles vs window length — the "cone" shape.
- Overlay current realized vol at each window.
Interpretation
- Current vol > 75th percentile: historically elevated — expect mean reversion.
- Current vol < 25th percentile: historically compressed — expect expansion.
- Cone narrowing at longer windows: vol mean-reverts over longer horizons.
See references/volatility_cones.md for full methodology and worked examples.
Crypto Volatility Characteristics
Crypto vol differs from traditional assets in important ways:
| Characteristic | Detail |
|---|---|
| Level | 50–150% annualized is typical; TradFi equities are 15–25% |
| Clustering | Strong — high-vol days cluster together |
| Weekday patterns | Weekend vol often lower but weekend gaps can be large |
| Volume correlation | Vol and volume are positively correlated |
| Regime dependence | Bull market vol ≠ bear market vol; ranges are different |
| Mean reversion | Vol mean-reverts more reliably than price |
| Tail risk | Fat tails — more extreme moves than normal distribution predicts |
Regime Classification by Volatility
| Regime | Annualized Vol Range | Characteristics |
|---|---|---|
| Low vol | < 40% | Range-bound, mean reversion works |
| Normal vol | 40–80% | Trending possible, balanced strategies |
| High vol | 80–120% | Strong trends or sharp reversals |
| Crisis vol | > 120% | Liquidation cascades, reduced position size |
Volatility Forecasting
EWMA Forecast
Simple and effective. The current EWMA variance estimate is the 1-step forecast. Multi-step forecasts are flat (same as 1-step).
GARCH Forecast
GARCH produces a term structure of variance forecasts:
σ²_{t+h} = V_L + (α + β)^h · (σ²_t − V_L)
Where V_L = ω / (1 − α − β) is the long-run variance.
- Short-horizon forecasts reflect current conditions.
- Long-horizon forecasts converge to long-run variance.
- The speed of convergence depends on α + β (persistence).
See scripts/vol_forecast.py for a working implementation.
Practical Applications
Position Sizing with Volatility
# Vol-target position sizing
target_vol = 0.02 # 2% daily portfolio vol target
current_vol = 0.05 # 5% daily asset vol (annualized ~95%)
weight = target_vol / current_vol # = 0.40 → 40% allocation
See the position-sizing skill for complete integration.
ATR-Based Stop Placement
atr_14 = talib.ATR(highs, lows, closes, timeperiod=14)
stop_distance = 2.0 * atr_14[-1] # 2x ATR stop
stop_price = entry_price - stop_distance # for longs
Vol-Regime Strategy Selection
vol_percentile = current_vol_percentile(token, window=30)
if vol_percentile < 25:
strategy = "mean_reversion"
elif vol_percentile > 75:
strategy = "momentum_breakout"
else:
strategy = "balanced"
Files
References
| File | Description |
|---|---|
references/estimators.md | Full derivations and details for all volatility estimators |
references/volatility_cones.md | Cone construction methodology and interpretation guide |
Scripts
| File | Description |
|---|---|
scripts/estimate_volatility.py | Multi-estimator volatility computation with cone analysis |
scripts/vol_forecast.py | EWMA and GARCH forecasting with term structure output |
Related Skills
regime-detection— Classify market regimes using volatility as a key input.position-sizing— Scale positions inversely with volatility.risk-management— Portfolio-level vol targeting and risk budgets.pandas-ta— ATR and Bollinger Bands are volatility-based indicators.custom-indicators— Build crypto-specific volatility indicators.
Dependencies
uv pip install pandas numpy scipy
Optional for live data:
uv pip install httpx
FAQ
What does the volatility-modeling skill do?
Volatility estimation, forecasting, and regime classification using GARCH, EWMA, realized volatility, and volatility cones
How do I install volatility-modeling?
Run npx skills add https://github.com/agiprolabs/claude-trading-skills --skill volatility-modeling in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/volatility-modeling/ (personal) or .claude/skills/volatility-modeling/ (project).
Which AI agents can use volatility-modeling?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is volatility-modeling safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains volatility-modeling?
It is published by agiprolabs in agiprolabs/claude-trading-skills. It has 499 recorded installs.
More from agiprolabs/claude-trading-skills
walk-forward-validation
agiprolabs/claude-trading-skills
Walk-forward validation framework for trading strategies and ML models with time-series-aware splits, overfit detection, and regime-aware validation
risk-management
agiprolabs/claude-trading-skills
Portfolio-level risk controls, drawdown management, exposure limits, and circuit breakers for crypto trading
pandas-ta
agiprolabs/claude-trading-skills
Technical analysis with 130+ indicators using pandas-ta for crypto market data
position-sizing
agiprolabs/claude-trading-skills
Trade sizing methods including fixed fractional, volatility-adjusted, Kelly criterion, and liquidity-constrained sizing
regime-detection
agiprolabs/claude-trading-skills
Market regime identification using volatility clustering, trend detection, and statistical methods for adaptive trading
kelly-criterion
agiprolabs/claude-trading-skills
Kelly criterion optimal sizing with fractional variants, edge estimation, and practical application for crypto trading
Popular business & finance skills
sales-enablement
coreyhaines31/marketingskills
When the user wants to create sales collateral, pitch decks, one-pagers, objection handling docs, or demo scripts. Also use when the user mentions 'sales deck,' 'pitch deck,' 'one-pager,' 'leave-behind,' 'objection handling,' 'deal-specific ROI analysis,' 'demo script,' 'talk track,' 'sales playbook,' 'proposal template,' 'buyer persona card,' 'battle card,' 'battlecard,' 'competitive one-pager,' 'win-loss analysis,' 'why are we losing deals,' 'loss reasons,' 'how do I respond to this objection,' 'prospect just said,' 'help my sales team,' 'sales materials,' or 'what should I give my sales reps,' 'cold call script,' or 'voicemail script.' Use this for any document or asset that helps a sales team close deals. For public competitor comparison pages, see competitors. For running buyer interviews or churn research, see customer-research. For marketing website copy, see copywriting. For cold outreach emails, see cold-email. For the offer being sold (bonuses, guarantees, pricing structure),
revops
coreyhaines31/marketingskills
When the user wants help with revenue operations, lead lifecycle management, or marketing-to-sales handoff processes. Also use when the user mentions 'RevOps,' 'revenue operations,' 'lead scoring,' 'outbound pipeline stages,' 'lead routing,' 'MQL,' 'SQL,' 'pipeline stages,' 'deal desk,' 'CRM automation,' 'marketing-to-sales handoff,' 'data hygiene,' 'leads aren't getting to sales,' 'pipeline management,' 'lead qualification,' or 'when should marketing hand off to sales.' Use this for anything involving the systems and processes that connect marketing to revenue. For cold outreach emails, see cold-email. For email drip campaigns, see emails. For pricing decisions, see pricing.
sales-funnel-blueprint
autonnel/autonnel-skills
Turn an offer into a concrete multi-step sales funnel spec - page-by-page structure, price ladder, copy outline and the metrics each step must hit. Use when asked to build a sales funnel, marketing funnel, landing page flow, lead magnet funnel, webinar funnel, tripwire or VSL funnel, when planning a product launch page flow, or when someone asks "what pages do I need" for selling something online.
pricing
coreyhaines31/marketingskills
When the user wants help with pricing decisions, packaging, or monetization strategy. Also use when the user mentions 'pricing,' 'pricing tiers,' 'freemium,' 'free trial,' 'packaging,' 'price increase,' 'value metric,' 'Van Westendorp,' 'willingness to pay,' 'monetization,' 'how much should I charge,' 'my pricing is wrong,' 'pricing page,' 'annual vs monthly,' 'per seat pricing,' 'should I offer a free plan,' 'pricing page teardown,' 'pricing page audit,' 'is my pricing page AI-readable,' or 'can AI read my pricing.' Use this whenever someone is figuring out what to charge, how to structure their plans, or wants to audit a pricing page (for humans and for the AI agents that shortlist tools). For in-app upgrade screens, see paywalls. For offer construction (bonuses, guarantees, value framing, naming) on services/courses/coaching/high-ticket B2B, see offers.
cross-border-ecommerce
nexscope-ai/ecommerce-skills
Cross-border e-commerce expansion advisor. Scores target markets on 8 weighted dimensions (market size, ecommerce penetration, competition, regulatory complexity, logistics infrastructure, payment ecosystem, cultural distance, IP protection), compares 5 fulfillment models with cost and transit data, provides country-by-country tax/duty compliance guides (EU VAT/IOSS, UK VAT, US sales tax, CA GST, AU GST, JP consumption tax), maps local payment preferences by market, and builds a phased expansion roadmap. No API key required.
pricing-strategy
coreyhaines31/marketingskills
When the user wants help with pricing decisions, packaging, or monetization strategy. Also use when the user mentions 'pricing,' 'pricing tiers,' 'freemium,' 'free trial,' 'packaging,' 'price increase,' 'value metric,' 'Van Westendorp,' 'willingness to pay,' 'monetization,' 'how much should I charge,' 'my pricing is wrong,' 'pricing page,' 'annual vs monthly,' 'per seat pricing,' or 'should I offer a free plan.' Use this whenever someone is figuring out what to charge or how to structure their plans. For in-app upgrade screens, see paywall-upgrade-cro.