Tracking VIX with VIX Futures: Portfolio Construction and Performance
Pith reviewed 2026-05-25 12:20 UTC · model grok-4.3
The pith
Dynamic daily-rebalanced VIX futures portfolios track the spot VIX more closely than static futures mixes or the VXX ETN.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
Static portfolios of VIX futures fail to track the VIX closely. In a discrete-time model calibrated to historical data, a dynamic trading strategy that adjusts daily to optimally track VIX achieves superior performance, outperforming the volatility ETN VXX in both empirical and simulation evaluations.
What carries the argument
Dynamic daily-rebalanced portfolio of VIX futures derived via optimization in a discrete-time model.
If this is right
- Static portfolios of different VIX futures contracts cannot track the index closely.
- Daily dynamic adjustment in the discrete-time model reduces tracking error relative to static holdings.
- The calibrated dynamic strategy exhibits lower tracking error than the VXX ETN.
- Simulation studies reveal the statistical properties of the daily-rebalanced positions.
Where Pith is reading between the lines
- Similar daily optimization could be tested on other volatility futures such as those on the VSTOXX.
- Real-time implementation would require live recalibration of the discrete-time parameters each day.
- The performance edge may shrink during periods of extreme contango or backwardation not fully represented in the calibration sample.
Load-bearing premise
The discrete-time model assumptions and its calibration to historical data will hold for future VIX futures price movements.
What would settle it
An out-of-sample period in which the dynamic strategy's mean squared tracking error exceeds that of VXX would falsify the superiority result.
read the original abstract
We study a series of static and dynamic portfolios of VIX futures and their effectiveness to track the VIX index. We derive each portfolio using optimization methods, and evaluate its tracking performance from both empirical and theoretical perspectives. Among our results, we show that static portfolios of different VIX futures fail to track VIX closely. VIX futures simply do not react quickly enough to movements in the spot VIX. In a discrete-time model, we design and implement a dynamic trading strategy that adjusts daily to optimally track VIX. The model is calibrated to historical data and a simulation study is performed to understand the properties exhibited by the strategy. In addition, comparing to the volatility ETN, VXX, we find that our dynamic strategy has a superior tracking performance.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies static and dynamic portfolios of VIX futures for tracking the VIX index. Static portfolios are shown to track poorly because futures do not react quickly enough to spot VIX moves. A discrete-time model is used to derive an optimal daily-rebalancing dynamic strategy; the model is calibrated to historical VIX futures data, a simulation study is conducted to examine its properties, and the strategy is reported to exhibit superior tracking performance relative to the VXX ETN.
Significance. If the reported outperformance is robust, the work would supply a concrete, optimizable dynamic futures strategy for VIX tracking with both theoretical derivation and simulation evidence, which could be useful for volatility-product design and risk-management applications. The combination of optimization-based portfolio construction and explicit comparison to an existing traded product is a positive feature.
major comments (2)
- [simulation study and VXX comparison] The central empirical claim (superior tracking vs. VXX) rests on a simulation study whose parameters are calibrated to the same historical VIX futures sample used to evaluate performance. No held-out test period, walk-forward validation, or truly out-of-sample period is described, so the reported superiority may be an in-sample artifact rather than evidence of genuine tracking ability (see abstract and the simulation/comparison sections).
- [discrete-time model and calibration] The discrete-time model assumptions (including the form of the futures price dynamics and the daily rebalancing rule) are calibrated to historical data; the paper does not provide a sensitivity analysis or stress test showing that the tracking advantage survives plausible perturbations to those calibrated parameters or to different volatility regimes.
minor comments (2)
- [abstract] The abstract states both empirical and theoretical results but supplies no equations, data summaries, error metrics, or methodology details; the main text should make these explicit early on.
- [model and optimization sections] Notation for the optimization objective, state variables, and rebalancing rule should be introduced consistently and with clear definitions before the simulation results are presented.
Simulated Author's Rebuttal
We thank the referee for the constructive feedback. We address each major comment below and outline revisions to improve the robustness of the empirical results.
read point-by-point responses
-
Referee: [simulation study and VXX comparison] The central empirical claim (superior tracking vs. VXX) rests on a simulation study whose parameters are calibrated to the same historical VIX futures sample used to evaluate performance. No held-out test period, walk-forward validation, or truly out-of-sample period is described, so the reported superiority may be an in-sample artifact rather than evidence of genuine tracking ability (see abstract and the simulation/comparison sections).
Authors: We acknowledge that the model is calibrated and the simulation study is conducted on the full historical sample, and that the VXX comparison uses the same period. This setup means the reported outperformance is not based on a held-out test set. To mitigate concerns about in-sample artifacts, the revised manuscript will include a walk-forward validation exercise that uses rolling calibration windows on earlier data to evaluate the dynamic strategy on subsequent periods, with results reported relative to VXX. revision: yes
-
Referee: [discrete-time model and calibration] The discrete-time model assumptions (including the form of the futures price dynamics and the daily rebalancing rule) are calibrated to historical data; the paper does not provide a sensitivity analysis or stress test showing that the tracking advantage survives plausible perturbations to those calibrated parameters or to different volatility regimes.
Authors: We agree that the absence of sensitivity checks limits the strength of the robustness claim. The revised version will add a dedicated analysis that perturbs the calibrated parameters (e.g., volatility-of-volatility and mean-reversion speeds) within empirically plausible ranges and evaluates tracking performance across distinct volatility sub-periods identified in the historical record. revision: yes
Circularity Check
Dynamic strategy outperformance vs VXX derived from simulation of model calibrated to same historical data
specific steps
-
fitted input called prediction
[Abstract]
"In a discrete-time model, we design and implement a dynamic trading strategy that adjusts daily to optimally track VIX. The model is calibrated to historical data and a simulation study is performed to understand the properties exhibited by the strategy. In addition, comparing to the volatility ETN, VXX, we find that our dynamic strategy has a superior tracking performance."
The superior tracking performance is exhibited only in the simulation study of the strategy whose parameters were fitted to the historical data; the comparison to VXX is therefore not an independent out-of-sample result but follows directly from the calibrated inputs.
full rationale
The paper derives a dynamic daily-rebalancing strategy from a discrete-time model that is calibrated to historical VIX futures data, then reports superior tracking performance versus VXX via simulation study on that same data. No held-out test period or walk-forward validation is described in the abstract, so the reported superiority reduces to performance of the fitted model rather than independent evidence. This matches the fitted_input_called_prediction pattern but is not fully self-definitional or self-citation based. The rest of the derivation (static portfolios, optimization) appears independent.
Axiom & Free-Parameter Ledger
free parameters (1)
- discrete-time model parameters
axioms (1)
- domain assumption Discrete-time model assumptions accurately capture VIX futures dynamics
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.