| Indicator | Now | Bucket | Measured forward return |
|---|---|---|---|
| Implied volatility (VIX) | 19.0 | normal (15-20) | fwd 12m +12.5% (t=0.0, not significant) |
| Drawdown from high | -0.8% | at/near highs | fwd 1m +0.7% (t=-1.3) |
| Price vs 200-day average | +6.4% | above 200dma | no significant effect either side (|t| < 1) |
| Volume vs 20-day average | 1.33x | busy (1.2-1.6x) | no measurable effect at any horizon (all |t| < 1.1) |
| Pay-cycle position | trading day 15 of 15 | turn of month | inside the turn-of-month window — real pre-1988, statistically dead since 2008 (t=0.29) |
trend intact Index is 6.4% above its 200-day average — trend intact.
Holding stock: A 200dma exit rule would hold. Historically this rule cut max drawdown in every era tested (1929: -86% to -49%; GFC: -57% to -9%), but its return edge came entirely from 1929-49 and 2000-08 and it has cost about 5.3%/yr since 2009 before tax. Treat it as insurance with a premium, not as alpha.
Selling premium: For selling premium this signal does not apply, and inverting it is worse. Over 428 monthly short puts, selling below the 200dma out-returned selling above it (+7.6% vs +6.4% annualised), and six of the ten worst months began ABOVE the average — two with VIX in the bottom third of its range. Crashes start from calm, so a backward-looking trend filter cannot see them coming. Position sizing limits this damage; timing does not.
Implied vol is low. Option sellers are collecting thin credits; buyers of premium are paying less than usual for convexity.
None of these should gate a stock purchase. Every waiting rule tested over 33 years underperformed simply buying on schedule — including VIX>30, the strongest predictor, which cost 11.7%. Deploy on your own cadence; use this panel for context, not permission.
Measured over 97 years of S&P 500 data and 33 years of VIX, with Newey-West standard errors. Volume showed no effect at any horizon. The turn-of-month pay-cycle effect was real before 1988 (t=5.21) and is statistically dead today (t=0.29). Waiting for VIX>30 — the strongest predictor found, t=5.0 for forward 12-month returns — cost 11.7% of terminal wealth over 33 years of simulated paycheck investing, because acting on it meant up to 1,598 consecutive days in cash.