Why usFeaturesTemplatesBlogGlossary

Crypto Volatility-Targeted Core

Holds Bitcoin and Ethereum continuously, but scales total exposure to hit a 25% annualised portfolio volatility target — cutting position size when realised volatility spikes and leaving the remainder in cash. Rebalanced weekly. No leverage.

Why this is expected to work

Volatility clusters. Tomorrow's volatility is predicted far better by recent volatility than tomorrow's return is predicted by recent returns, which is why sizing on volatility is a more reliable lever than timing on price. Harvey et al. ("The Impact of Volatility Targeting", 2018) found the approach improved Sharpe ratios across equities, credit and commodities, with the largest gains in the highest-volatility assets — and crypto is the highest-volatility liquid asset class available. The second effect matters more for a real investor than the Sharpe number: because volatility spikes and crashes coincide, targeting volatility mechanically reduces exposure into the worst drawdowns without ever forecasting them. Leverage is capped at 1.0 here, so the overlay only ever de-risks; the book holds cash when 25% is not achievable, and never borrows to reach it. A 25% target is still roughly double the long-run volatility of the Nifty 50 — this is a risk control, not a safe strategy.

cryptovolatility-targetingriskposition-sizing
Universe
Crypto Perpetuals HYPERLIQ
Rebalance
Weekly Inverse volatility

How the pipeline works

Filter

Priced today

Keeps rows where close > 0

close > 0
Size

Risk-scaled sizing

Sizes positions to a 25% volatility target

What this template teaches

  • WeightingRule volatility-target overlay
  • max_leverage=1.0 — de-risk only, cash is the residual
  • Gross exposure as a separate decision from relative weights
  • PassthroughWeighting when the pipeline sizes explicitly
Read the full write-up

Volatility Targeting: Sizing Is a More Reliable Lever Than Timing

Tomorrow's volatility is predicted far better by recent volatility than tomorrow's return is by recent returns. That single asymmetry is why scaling position size works when forecasting direction does not.