Research / Strategies / Volatility Regime Filter
Study № 07 VIX1D · US Equity ETFs Regime Filter Daily Cadence

VIX1D regime filter on SPYD.

A binary risk-on / risk-off overlay on SPYD, gated entirely by the VIX1D regime. When next-day implied volatility signals stress, the book steps to cash; otherwise it holds one unit of equity.

Asset
SPYD SPDR S&P 500 HIGH DIV.
Signal
VIX1D REGIME STATE
Window
Apr 2023
— Apr 2026
Sample
739 days OPEN-TO-OPEN
Total Return01
45.88%
vs B&H +26.15pp
CAGR02
13.74%
vs B&H +7.41pp
Sharpe03
1.16
vs B&H +0.67
Max DD04
−10.20%
vs B&H +9.09pp
Calmar05
1.35
vs B&H +1.02
01 / Concept

VIX1D tells you when the room is tilting.

The filter does one thing well: read the VIX1D regime at the 4:15 PM close and decide, without ambiguity, whether tomorrow's SPYD session earns equity exposure.

InputVIX1D · Daily Close
DecisionBinary · Long or Cash
Cadence4:15 PM → Next Open
StatesLong SPYD · Cash
I

Premise

Next-day implied volatility clusters. The cost of being long SPYD rises sharply inside elevated-VIX1D regimes and collapses on the other side.

II

Signal

A proprietary rolling read of VIX1D classifies each daily close into one of two regime states. The model is evaluated at 4:15 PM, after the cash session settles.

III

Action

A benign VIX1D state → long SPYD at open T+1. A stressed state → flat cash. Return is booked open[T+1] → open[T+2]. No intraday peeking.

02 / Performance

Growth of $1, filtered vs. passive.

The VIX1D filter captures the upside of passive SPYD ownership while sidestepping the material left-tail events of the sample window.

Equity Curve

Growth of $1 · Apr 2023 – Apr 2026
VIX1D Regime — SPYD
Buy & Hold — SPYD
Cash periods

Drawdown Profile

Peak-to-trough, daily · deeper is worse
Filter DD
B&H DD
03 / Metrics

The numbers, side by side.

Exposure of 83.5% — the filter is invested five days out of six — yet delivers materially higher risk-adjusted return than passive SPYD over the identical window.

StrategyTotal ReturnCAGRVolatility SharpeMax DDCalmarWin RateExposure
VIX1D Regime — SPYD 45.88% 13.74% 11.66% 1.16 −10.20% 1.35 52.35% 83.5%
Buy & Hold — SPYD 19.73% 6.33% 14.65% 0.49 −19.29% 0.33 51.44% 98.6%
Δ vs B&H +26.15pp +7.41pp −2.99pp +0.67 +9.09pp +1.02 +0.91pp −15.1pp
04 / Distribution

Monthly returns — VIX1D regime.

Loss months are shallow and rare — a direct function of cash-state coverage during the regime's worst windows.

05 / Regime

VIX1D and the regime state.

Shaded bands mark cash (disengaged) periods across the 739-day sample — 113 cash days against 626 long days.

VIX1D and Regime State

113 cash days · 626 long days · Apr 2023 – Apr 2026
VIX1D close
Cash state
06 / Methodology

How the study is run.

Signals are generated on the evening bar and executed at the next session's open. No look-ahead, no intraday peeking.

01
Data

VIX1D daily closes from IBKR (CBOE), SPYD daily OHLC from Polygon with an IBKR / yfinance fallback chain. 746 bars each, aligned to 742 common trading days.

02
Signal

A proprietary VIX1D regime model is evaluated on each day-T 4:15 PM close, producing a single binary state for the following session.

03
Execution

Benign state → LONG SPYD at open T+1. Stressed state → CASH. Return is measured open[T+1] → open[T+2]. No intraday peeking.

04
Review

Stats computed on the realized return stream against passive buy-and-hold SPYD over the identical 739-day window (28 Apr 2023 – 9 Apr 2026).

Want the full research note and position log?