Return Distribution Models
Study daily, weekly, monthly, and cycle returns. Bitcoin returns are often fat-tailed, meaning extreme moves occur more often than a normal bell curve would imply.
Read detailsQuant Research
Bitcoin quant models and deep market theory.
Advanced frameworks for studying Bitcoin returns, volatility, cycles, liquidity, options, on-chain behavior, miners, risk, and portfolio construction without pretending the model is the market.
Quant models are tools for asking better questions. They can fail when market structure changes, liquidity disappears, regulation shocks the system, data quality is weak, or human behavior breaks historical patterns.
Model Library
Core quantitative fields for Bitcoin.
Each model should be tested against Bitcoin's unique properties: fixed supply, 24/7 trading, global venues, derivatives leverage, on-chain transparency, custody risk, miner economics, and changing institutional access.
Volatility Models
Use realized volatility, rolling volatility, EWMA, GARCH-style thinking, and volatility clustering to understand risk that changes over time.
Read detailsDrawdown and Ruin Models
Measure peak-to-trough losses, recovery time, maximum drawdown, underwater periods, and how much decline a holder can survive.
Read detailsFactor Models
Compare Bitcoin exposure to equity beta, Nasdaq sensitivity, dollar strength, gold, real yields, liquidity, momentum, and volatility factors.
Read detailsRegime-Switching Models
Separate market states such as accumulation, bull trend, leverage mania, distribution, crash, and recovery instead of assuming one stable behavior.
Read detailsMonte Carlo Simulation
Generate many possible future paths using assumptions for return, volatility, drawdown, rebalancing, fees, and tax drag.
Read detailsPower Law and Log Regression
Some Bitcoin models fit price against time on logarithmic scales. Useful for cycle framing, dangerous if treated as destiny.
Read detailsLiquidity and Flow Models
Model ETF flows, exchange balances, stablecoin liquidity, order-book depth, bid-ask spreads, futures funding, and macro liquidity.
Read detailsOptions and Implied Volatility
Use options markets to study expected volatility, skew, hedging demand, tail risk pricing, and dealer positioning.
Read detailsOn-Chain Quant Signals
Use realized cap, MVRV, SOPR, HODL waves, dormancy, exchange flows, miner flows, and fee pressure as Bitcoin-native data.
Read detailsMiner Economics Models
Model hashprice, difficulty, energy cost, machine efficiency, curtailment, treasury policy, and forced selling risk.
Read detailsPortfolio Risk Models
Study allocation size, rebalancing, correlation instability, Sharpe ratio, Sortino ratio, VaR, CVaR, and liquidity needs.
Read detailsNetwork Adoption Models
Estimate value from users, wallets, nodes, developers, liquidity, institutional rails, merchant tools, and settlement demand.
Read detailsMarket Microstructure
Study how 24/7 venues, fragmented liquidity, futures, ETFs, arbitrage, market makers, and exchange outages shape price.
Read detailsModel Risk and Backtest Discipline
Protect readers from curve fitting, overfitting, survivorship bias, look-ahead bias, cherry-picked timeframes, and false precision.
Read detailsScenario Engine
Combine adoption, liquidity, regulation, miner security, lost supply, ETF demand, and macro inflation into transparent scenario ranges.
Read detailsReturn Distributions
Bitcoin does not behave like a calm bell curve.
Classical finance often begins with normally distributed returns. Bitcoin often violates that assumption: returns can be skewed, fat-tailed, clustered, and regime-dependent.
| Concept | Meaning | Bitcoin Use | Model Risk |
|---|---|---|---|
| Mean return | Average return over a period. | Useful for long-run assumptions. | Can be dominated by a few extreme cycles. |
| Median return | Middle observation. | Often more realistic than average for skewed data. | May understate upside tails. |
| Fat tails | Extreme moves happen more often than normal models imply. | Critical for risk sizing and leverage avoidance. | Hard to estimate from short history. |
| Skewness | Returns lean toward extreme upside or downside. | Bitcoin has historically shown asymmetric upside, but not smoothly. | Past skew may not persist. |
| Kurtosis | Tail heaviness. | Helps explain why ordinary VaR can fail. | Unstable across regimes. |
Volatility
Bitcoin risk changes through time.
What actually happened.
Calculated from historical returns. Useful for measuring current market turbulence, but backward-looking.
Recent data matters more.
Exponentially weighted models react faster to new volatility but can still miss sudden shocks.
Volatility clusters.
Large moves often follow large moves. Calm periods can persist until a catalyst breaks them.
Quiet can be deceptive.
Low volatility may signal mature liquidity or a coiled market before a large move.
Drawdown Models
The real question is what a holder can survive.
Largest peak-to-trough decline. Bitcoin history includes deep drawdowns, so sizing must assume emotional stress.
How long an investor waits to recover a prior high. This matters for confidence and liquidity planning.
The chance that leverage, forced selling, or personal cash needs destroy the position before recovery.
Bad early returns can damage a plan even if long-run returns are positive.
Factor Models
What is Bitcoin really exposed to?
| Factor | Question | Bitcoin Interpretation |
|---|---|---|
| Equity beta | Does BTC move with stocks? | Correlation can rise during liquidity stress and fall during Bitcoin-specific catalysts. |
| Nasdaq sensitivity | Does BTC trade like high-growth tech? | Often during risk-on/risk-off periods, less so during custody or ETF-specific events. |
| Gold factor | Does BTC behave like monetary insurance? | Sometimes, especially when fiat credibility or banking stress is central. |
| Dollar factor | Does USD strength pressure BTC? | A stronger dollar can tighten global liquidity and pressure risk assets. |
| Momentum | Does trend persistence matter? | Bitcoin has strong momentum phases, but reversals can be violent. |
| Liquidity | Does money availability drive price? | Global liquidity, ETF flows, stablecoins, and leverage can all matter. |
Regime Switching
One Bitcoin model rarely fits all market states.
Low attention, patient buyers.
Volatility may compress, long-term holders build, and bad news has less effect over time.
Trend confirms.
Momentum, ETF flows, media attention, and liquidity can reinforce each other.
Fragile upside.
Funding rates, open interest, and options skew can signal crowded positioning.
Forced selling dominates.
Liquidations, exchange stress, miner selling, and macro shocks can overwhelm fundamentals.
Structure rebuilds.
Volatility remains high, but long-term holders, builders, and infrastructure continue.
Simulation
Monte Carlo shows ranges, not destiny.
Assumptions matter.
Return, volatility, correlation, fees, taxes, rebalancing, drawdown limits, and time horizon all shape outputs.
Many futures, not one forecast.
Simulations show possible distributions of outcomes instead of one headline target.
Include bad worlds.
Model deep drawdowns, long flat periods, liquidity shocks, exchange failures, and policy restrictions.
Use percentiles.
Median, 10th percentile, 90th percentile, worst path, and recovery time are more useful than one average.
Power Laws and Valuation
Long-run curves need humility.
Power law, logarithmic regression, and adoption curves can make Bitcoin look orderly over long periods. The danger is treating a fitted line as a law of nature.
Fits Bitcoin price to time on log scales. Useful for framing maturity, risky as a forecast.
Smooths massive price ranges into channels. Helpful visually, vulnerable to curve fitting.
Assumes adoption starts slow, accelerates, then matures. Useful if Bitcoin behaves like a network technology.
Compares Bitcoin to gold, cash, bonds, real estate, offshore wealth, or reserves. Useful but highly assumption-driven.
Liquidity and Flows
Price often follows marginal buyers and sellers.
| Flow | Why It Matters | What To Watch |
|---|---|---|
| ETF flows | Brokerage demand can absorb or release large BTC exposure. | Daily inflows, outflows, issuer concentration. |
| Exchange balances | Coins moving to exchanges may signal sell pressure; withdrawals may signal custody demand. | Net flows and exchange-specific context. |
| Stablecoin liquidity | Stablecoins can act as crypto market cash. | Supply growth, exchange deposits, redemption stress. |
| Order-book depth | Thin books move more on the same order size. | Bid depth, ask depth, spreads, slippage. |
| Futures funding | Shows leverage imbalance. | Persistently high funding can signal crowded longs. |
| Macro liquidity | Global money conditions affect risk appetite. | Rates, dollar strength, central bank liquidity, credit stress. |
Options Theory
Options price uncertainty.
Expected movement.
Options markets imply how much volatility traders are willing to pay for.
Upside versus downside demand.
Put skew may show crash protection demand; call skew may show upside chase.
Dealer hedging pressure.
Options positioning can amplify or dampen price moves near large strike concentrations.
Short-term versus long-term fear.
Near-term volatility spikes around catalysts; long-dated vol reflects broader uncertainty.
On-Chain Quant
Bitcoin has native public data.
| Metric | Meaning | Use | Weakness |
|---|---|---|---|
| MVRV | Market value versus realized value. | Cycle heat and valuation stress. | Can stay high or low for long periods. |
| SOPR | Spent output profit ratio. | Shows whether spent coins move at profit or loss. | Exchange movement can distort meaning. |
| HODL waves | Age distribution of coins. | Tracks long-term holder behavior. | Lost coins are hard to classify. |
| Dormancy | Old coin movement. | Old coins moving can signal distribution or custody reshuffling. | Intent is not visible on-chain. |
| Fee pressure | Demand for blockspace. | Important for long-term miner security. | Short-term spikes may be temporary. |
Miner Models
Mining links Bitcoin price to energy and hardware.
Revenue per unit of hash rate. It depends on BTC price, fees, subsidy, and difficulty.
Adjusts roughly every two weeks to keep blocks near ten minutes. Rising difficulty can squeeze inefficient miners.
Power price is often the largest operating variable. Curtailment and demand response can change economics.
Newer ASICs produce more hash per watt, changing the cost curve.
Miners that hold BTC have upside exposure but may need to sell during stress.
As subsidy declines, transaction fees become more important to miner revenue and security budget.
Portfolio Risk
A Bitcoin allocation is a sizing problem.
Return per unit volatility.
Can look attractive historically, but volatility is not the only risk.
Downside-focused risk.
More useful when upside volatility is not considered harmful.
Tail loss estimates.
Useful for stress framing, but dangerous if tails are underestimated.
Discipline versus taxes.
Rebalancing can control risk but may create taxable events and emotional friction.
Network Models
Value can come from credible adoption.
More holders can deepen liquidity.
But address count is imperfect because one user can control many addresses.
Verification matters.
Node count and geographic distribution can signal resilience, but node quality also matters.
Maintenance is invisible until needed.
Bitcoin depends on careful protocol stewardship and review culture.
Rails change access.
ETFs, custody, banking access, and accounting standards can expand reachable demand.
Market Microstructure
Where price forms matters.
Bitcoin trades when traditional markets are closed, which can create weekend gaps in liquidity and attention.
Spot price forms across many exchanges, ETFs, OTC desks, and derivatives venues.
Price differences close when capital, settlement, and exchange access work smoothly.
ETFs connect brokerage capital to BTC exposure but trade on traditional market hours.
Leveraged positions can create cascade moves unrelated to long-term value.
Exchange downtime or withdrawal limits can distort price and user behavior.
Model Risk
The model can become the trap.
| Failure | Meaning | Protection |
|---|---|---|
| Overfitting | The model fits history too perfectly and fails live. | Use out-of-sample testing and simple assumptions. |
| Look-ahead bias | The model uses information that was not available at the time. | Time-stamp data and test honestly. |
| Survivorship bias | Failed assets, exchanges, or strategies disappear from the sample. | Include failures and delistings when relevant. |
| Cherry-picking | Choosing start and end dates that flatter the result. | Test multiple windows and cycle phases. |
| False precision | Detailed numbers create fake confidence. | Use ranges, confidence bands, and scenario language. |
Scenario Engine
Combine models into a decision framework.
A serious Bitcoin scenario engine should not output one magic price. It should show assumptions, ranges, stress cases, and what would make the model wrong.
Retail, ETF, corporate, sovereign, and payment adoption should be separate assumptions.
Include interest rates, dollar strength, stablecoins, ETF flows, and leverage.
Include issuance, lost coins, exchange balances, long-term holder supply, and miner selling.
Include tax rules, custody rules, mining policy, exchange access, and capital controls.
Include fee market, miner incentives, node participation, custody failures, and protocol risk.
Every model should state what evidence would make the thesis weaker.