Reality Check

Let the VIX set your position size, so a reading of 20 means 20% of the account in stock. We tested the 1x and 1.5x versions on 4,850 daily closes. Both beat holding a fixed slice of the same size, and both left you at your largest while the market fell hardest.

A weak open that closes strong is supposed to show real buying. We tested both halves of that rule on 33 years of S&P 500 sessions under twelve different definitions of weak and strong. Every one leans the opposite way to the rule.

The weekly MACD cross is one of the most watched trend signals. We tested every bullish and bearish cross on 56 years of S&P 500 weekly closes. The week that follows looks the same either way, and the bearish cross is the one followed by the better week.

Alphabet gaps up on earnings about 60% of the time. Across 77 reports since 2007, going in 10% below its 52-week high leaves that rate unchanged, and 20% below drops it to 42%. Being beaten down first has not improved the odds, only widened the gap.

Why Probability of Touch is often the more useful number for both options sellers and buyers.

Buying the S&P 500 only when the VIX spikes above 35 and selling once it calms below 12 captured +457% over 20 years, close to buy-and-hold's +513%, while invested only 64% of the time and with roughly half the average drawdown. A backtest of the buy-panic, sell-calm family across entry and exit thresholds.

Goldman Sachs has gapped up at least 3.74% on an earnings reaction eight times since 2008. In seven of those eight, GS was trading higher 30 days later, averaging +9.59% with the gap itself excluded. Every instance is here, along with why eight observations cannot carry a high conviction trade.

A 20-year backtest of the Sell in May adage on the S&P 500. The winter half (Nov to Apr) really has beaten the summer half (May to Oct), averaging +5.88% versus +3.78%. But summer was still positive in 74% of years, and selling every May trailed buy-and-hold by a wide margin, +191% versus +513%.

We tested two classic overbought signals on 20 years of the S&P 500: a close above the upper Bollinger Band, and RSI above 70. On its own, an upper-band close cuts the market's usual weekly gain to roughly a wash (54.8% green). With RSI above 70 as well, the next week becomes a coin flip at 50.6%, while oversold closes below the lower band still bounce 61.9% of the time.

We checked the trader belief that three down days in a row is the sweet spot for a reversal, across 55 names and 20 years. The fourth-day bounce is real but small on the steady giants and the index (the S&P 500 rises to 59% green), and it turns negative on the hottest names like Rivian, SoFi and Palantir.

On 20 years of S&P 500 breakouts (a close above the prior day's high), heavy volume did not improve short-term follow-through. Low-volume breakouts held better the next day (53% vs 38%) and over five days (63% vs 44%), the opposite of the common rule that a breakout needs heavy volume to be real.

We logged 90 trading days of 0DTE dealer gamma walls on the S&P 500. At least one wall was touched on 77% of sessions, and the direction of the open predicted which one: the wall on the side of the opening gap was reached about 62% of the time, the far wall about a quarter.

A study of ten major assets shows that 2x-volume spikes reliably widen both tails, while next-day and 30-day directional results are mixed.

We tested the volatility squeeze four ways on the S&P 500, Nasdaq, and semis across decades: consecutive range contraction, NR7, a range-percentile squeeze, and a Bollinger band-width squeeze. A tighter range does not precede a bigger move, and the sharper the squeeze, the quieter the day and week that follow.

On steadily rising indexes, the engulfing candle provides minimal additional information.

AI-driven infrastructure spending has decoupled two of technology’s most popular sector ETFs, dropping their same-direction move rate from 85.1% in 2022 to 55.2% in 2026.

We pulled every 15-minute bar of the S&P 500 from 2022 to 2026 and found which window holds each day’s high and low. The opening 15 minutes set the low one day in five, the closing 15 minutes set the high one day in six, and the lunch hour rarely sets either.

Across 1,471 S&P 500 sessions since 2020, we tracked every cross of the prior-day close. A first flip in the lunch hour, 12:15 to 1:30 PM, held to the close just 44% of the time, worse than a coin flip, while flips at other hours leaned toward sticking. The midday recovery was the weakest of all.

We tracked every VIX episode between 20 and 30 since 2014 and asked which boundary breaks first: 30, or back below 20. The odds start out favoring a quick retreat, then flip the longer the episode drags on.

We tested inside days and inside weeks on the S&P 500, Nasdaq, and semis across decades. A single inside bar predicts neither direction nor a bigger move. Two in a row precede a move about a quarter larger than normal, but only two: three inside bars in a row almost never happen.

We measured the S&P 500’s return during the week of July 4th for the past 20 years, year by year. The holiday week closed higher in 13 of 20 years with a +1.26% median, and the last decade ran green 9 times out of 10.

Across 1,469 S&P 500 sessions since 2020, we tracked how much of the day’s high-to-low range is in at every 15-minute mark. By 11am it is about 56%, and by 1pm roughly three quarters, so the morning sets the boundaries the afternoon trades inside.

COR1M, the Cboe correlation index, tracks how much S&P 500 stocks move together. We lined it up against the market and the VIX: it shadows the VIX almost perfectly, moves opposite to the market the same day, and does not predict the next day.

Across 1,468 S&P 500 sessions since 2020, we checked every 15-minute mark to see how often the day's color survives to the 4 PM close. The odds climb from about 70% at the open to 94% by 3:45 PM, and green has held a few points better than red.

Across 33 years of S&P 500 sessions, an opening gap is a strong clue to whether it fills and a weak one for direction. Small gaps fill about 70% of the time, big gaps tend to run, and the day's green odds stay near a coin flip no matter how it opened.

A popular claim says overnight Bitcoin shows where the VIX opens better than S&P futures. We tested nearly 1,500 overnight sessions going back to mid-2020: the S&P move wins clearly, and Bitcoin adds nothing once the S&P move is known.

Every weekday bounces about equally after a red day, but the fades are weekday-specific: a green Monday tends to give it back Tuesday. The strongest two-day setups are a green Friday-and-Monday fading into Tuesday and a red Monday-and-Tuesday rebounding on Wednesday.

We tested every 15-minute S&P bar that jumped or dropped 0.25% to 0.75%. The close barely gets more predictable: up-moves drift on, down-moves stabilize, and the edge is tiny next to the randomness.

VIX bounces from almost everywhere, so a line drawn anywhere looks like it works. Once you control for the VIX level, the trendline adds nothing.

We set out to test whether a beaten-down VIX warns of a spike. In 22 years its RSI dropped below 30 only five times. VIX is structurally almost never oversold.

We tested every cross above the 10, 20, 50, and 200-day lines, and being above them, over 22 years. The crosses lag, and being above the average just means VIX is already elevated.

The bigger the one-day surge in VIX's RSI, the more VIX faded afterward. The signal arrives after the spike is already made.

It looks like it warned you before every big spike. We tested all 138 of them over 22 years against a plain baseline, and the forward odds match a random day.

BOIL melts and winter is its worst season, so the short wins most years. The risk is the one winter in seven that can take the whole account.

On the S&P 500, buying at RSI 30 is a real edge. On single stocks it is a coin flip, and on the best stock of the decade it lost money. Oversold means opposite things on an index and a stock.

By the time the death cross fires, VIX is already at its 84th percentile and the fear is in the tape. Then it fades, and the famous doom signal was followed by the S&P rising 80% of the time.

Natural gas is the most mean-reverting commodity, and the 50/200 cross is a trend tool. Across 42 crosses since 2000 the golden cross bought the top (down 5% the next month) and the death cross often marked the bottom.

A stock pokes above a prior high, then closes red: the classic reversal short. Across 20 years of the S&P 500 and QQQ it never paid. The indices tend to bounce; a failed breakout is usually a pause in an uptrend rather than a reversal.

UVXY decays to a new low almost every week, so held long enough every short wins. We shorted every 50% spike since 2011: a 100% win rate, and an account wiped to zero by one spike.

We ran two different tests for whether a rise in VIX9D predicts a rise in VIX, split by VIX regime (under 20, 20 and above). Neither found VIX9D leading VIX higher.