Files
ai/gateway/knowledge/trading/strategies/volatility/volatility-carry.md
Tim Olson 47471b7700 Expand model tag support: add GLM-5.1, simplify Anthropic IDs, scan tags anywhere in message
- Flink update_bars debouncing
- update_bars subscription idempotency bugfix
- Price decimal correction bugfix of previous commit
- Add GLM-5.1 model tag alongside renamed GLM-5
- Use short Anthropic model IDs (sonnet/haiku/opus) instead of full version strings
- Allow @tags anywhere in message content, not just at start
- Return hasOtherContent flag instead of trimmed rest string
- Only trigger greeting stream when tag has no other content
- Update workspace knowledge base references to platform/workspace and platform/shapes
- Hierarchical knowledge base catalog
- 151 Trading Strategies knowledge base articles
- Shapes knowledge base article
- MutateShapes tool instead of workspace patch
2026-04-28 15:05:15 -04:00

3.8 KiB
Raw Permalink Blame History

description, tags
description tags
Volatility carry with two ETNs shorts VXX (short-maturity VIX futures ETN) and buys VXZ (medium-maturity VIX futures ETN) to harvest the contango roll loss differential, with the hedge ratio determined by serial regression.
volatility
carry
vxx
vxz
etn
contango
roll

Volatility Carry with Two ETNs

Section: 7.3 | Asset Class: Volatility | Type: Carry

Overview

VXX and VXZ are exchange-traded notes (ETNs) tracking VIX futures. VXX tracks short-maturity (months 1-2) futures and suffers greater roll/contango losses than VXZ (months 4-7), because the VIX futures curve is steepest at the short end in contango. The strategy shorts VXX (captures the larger roll loss as profit) and buys VXZ as a hedge (offsets some exposure to VIX spikes), earning the contango roll differential.

Construction / Mechanics

VXX: tracks a constant-maturity position in months 1-2 VIX futures. Each day, a fraction of the front-month futures is sold and replaced with the next-month futures. In contango, the next-month is more expensive → daily roll loss → VXX decays over time.

VXZ: tracks months 4-7 VIX futures. Same roll mechanism but in a less steep part of the contango curve → lower roll loss than VXX.

Basic strategy: short VXX, long VXZ.

Hedge ratio h (number of VXZ units per VXX shorted):

h = β = ρ · σ_X / σ_Z

where:

  • ρ: historical pairwise correlation between VXX and VXZ returns
  • σ_X: historical volatility of VXX
  • σ_Z: historical volatility of VXZ
  • β: slope of serial regression of VXX returns on VXZ returns (with intercept)

The position: short 1 unit of VXX, long h units of VXZ.

Payoff / Return Profile

  • Earns the differential roll loss between VXX and VXZ: the strategy benefits because VXX decays faster than VXZ in contango.
  • The VXZ long position partially hedges against VIX spikes (which cause VXX to spike more sharply than VXZ).
  • Profitable in normal, low-volatility, contango environments.
  • Experiences sharp drawdowns during sudden VIX spikes (equity market selloffs), as VXX spikes more violently than VXZ in the short term.

Key Parameters / Signals

  • Contango in VIX futures curve: the necessary condition for the strategy to profit
  • h = ρ · σ_X / σ_Z: hedge ratio (number of VXZ units per 1 unit of VXX shorted)
  • Roll loss differential between VXX and VXZ: the carry being harvested
  • VIX level and slope of futures curve: risk indicators

Variations

7.3.1 Hedging Short VXX with VIX Futures

Instead of using VXZ to hedge, use a basket of N medium-maturity VIX futures (e.g., months 4-7) directly. The optimal weights w_i for the N futures:

w_i = σ_X Σ_{j=1}^N C_{ij}^{-1} σ_j ρ_j                           (432)

where:

  • ρ_j: historical correlation between futures j and VXX returns
  • C_{ij}: N×N sample covariance matrix of the N futures returns (C_{ii} = σ_i²)
  • σ_X: historical volatility of VXX

Dollar-neutral constraint (optional): Σ_i w_i = 1 (Eq. 433). Some w_i may be negative; can impose w_i ≥ 0 if short futures is undesirable.

Portfolio can be rebalanced monthly or more frequently. This variation allows finer control over the hedge than using VXZ alone.

Notes

  • VXX spikes (which occur during equity market selloffs) can be large and sudden, causing substantial short-term P&L drawdowns even if the strategy is profitable overall.
  • The hedge ratio h should be recalibrated periodically using updated historical data.
  • The corresponding VXZ spikes are typically smaller, providing only partial protection during stress.
  • Transaction costs (bid-ask spread on VXX/VXZ) and ETN management fees must be accounted for in return estimates.
  • In sustained backwardation periods, both VXX and VXZ can rise; the strategy may lose money if VXX rises faster than VXZ.