If an ETF Spikes 20%, What Are the Odds It Reaches 50%?

Volatility ETFs like UVXY, VXX, and SQQQ don't trend. They run in episodic spike cycles. Each cycle starts at a fresh all-time low, climbs some unknown distance as fear hits the market, and ends when a new all-time low prints and the next cycle begins. The decision-relevant question for anyone trading these names is conditional: my ticker has already spiked +20% off its ATL. What are the odds the same cycle keeps climbing to +50% before it resets?

Eyeballing it from a chart is a coin flip. The honest answer needs a bar-by-bar walk of every cycle in the ticker's history, separated by ATL prints, with each cycle's peak measured against the ATL it was anchored to. That's the Spike Probability Ladder.

The Spike Probability Ladder

Spike Probability Ladder showing the From → To probability with median, average, shortest, and longest days to reach the target plus a distribution chart and event log

Pick a ticker from the 15-ETF universe (UVXY, UVIX, VXX, SQQQ, SOXS, ZSL, BOIL up front; DUST, FAZ, GLL, LABD, SCO, SPXU, TSLQ, TZA in the More dropdown). Set From Spike % (the starting threshold a cycle has already crossed, e.g. +20%). Set To Spike % (the target, e.g. +50%). The tool reports back five numbers and a chart.

The headline reading is the probability with both numerator and denominator visible. 61.2% (49/80) means: across the entire history, 80 cycles printed at +20%, and 49 of them carried on to +50% before the next ATL reset. That's a real conditional probability, not a hand-wave.

Below the probability you see four day-counts: median days from the From hit to the To hit (typical pace), average (mean, tail-skewed), shortest (the fastest cycle ever recorded), and longest (the slowest). Anything beyond longest hasn't happened historically.

Below the cards: a distribution chart showing every cycle plotted by days-to-target, and an expandable event log listing each From hit by date, with the ATL it was anchored to, and whether it reached the target.

What the Tool Shows You

Six readings off one chosen ticker and one From → To pair. Run it on a different ticker by clicking another pill; run it on a different range by stepping the inputs.

Headline

Conditional Probability

From → To probability shown with the underlying counts: 61.2% (49/80). The denominator is every From-hit cycle in history; the numerator is cycles that also hit To before the next ATL reset.

Days

Median, Average, Shortest, Longest

How long the From → To climb actually takes. Median is the typical pace; average is mean (tail-skewed); shortest and longest are the floor and ceiling of the dataset. Anything beyond longest has never happened.

ATL Anchor

Cycle-Bounded Math

The clock resets every time the ticker prints a new all-time low. From and To percentages are measured off the running ATL, so the probability is genuinely conditional on this cycle, not on some arbitrary lookback window.

Distribution

Days-to-Target Chart

Every reaching cycle plotted by how many days it took to climb from From to To. Shows whether the average is dominated by a few outlier slow climbs or sits inside a tight cluster of fast spikes.

Audit

Per-Event Log

Click to expand a flat table of every From hit in the ticker's history: date, price at the hit, the ATL it was anchored to, and whether the cycle reached To (and how many days it took if so). Useful for spotting regime drift.

Universe

15 Vol & Inverse ETFs

UVXY, UVIX, VXX, SQQQ, SOXS, ZSL, BOIL up front. DUST, FAZ, GLL, LABD, SCO, SPXU, TSLQ, TZA in the More dropdown. The full set of leveraged shorts and inverse vol products that move in episodic ATL-anchored spike cycles.

How It Works

The simulation runs as a single bar-by-bar walk over the chosen ticker's full daily history.

Cycle bookkeeping

Track a running ATL. Every bar whose low is below the current ATL becomes the new ATL and starts a fresh cycle.
Hold two flags per cycle: fromHit (false until this cycle's high crosses ATL × (1 + From%)) and toHit (false until the high crosses ATL × (1 + To%)).
When a new ATL prints mid-cycle and toHit was still false, the cycle goes into the dataset as a non-reaching event. The ticker hit From but failed to reach To before the reset.

Probability and day stats

Numerator: count of cycles where toHit fired before the cycle reset.
Denominator: count of cycles where fromHit ever fired.
Probability = numerator / denominator × 100.
For every reaching cycle, record the days from the bar that triggered fromHit to the bar that triggered toHit. Those become the median / average / shortest / longest stats.

A worked example

Say UVXY just hit a new all-time low at $5.00. Three weeks later it spikes to $6.00 (+20% off ATL), and fromHit fires for this cycle. Two weeks after that it climbs to $7.50 (+50% off the same ATL), and toHit fires. The cycle gets added to the dataset as a 14-day reach. If instead UVXY had printed a new ATL at $4.80 before reaching $7.50, the cycle would close as a non-reaching event, and a fresh cycle anchored to $4.80 would begin.

This is the tool to open mid-spike, when you're holding a position or thinking about entering one and want to know whether the move you're betting on has historical precedent or is asking the ticker to do something it has rarely done.

How Traders Are Using It

Five workflows that lean on the cycle-conditional probability + days-to-target stats.

Mid-spike hold-or-trim

You're long UVXY and it's spiked +30% off its recent ATL. From=30, To=50 reads back at, say, 38% (24/63). Better than a coin flip, but not great. Now check From=30, To=40. If it's 70%+, that's where the realistic exit lives. Use the ladder to find your honest target.

Sizing a short before the spike completes

You want to short the spike but the ticker has only reached +20%. Run From=20, To=50. If that's a 50% probability, you're sizing into a coin flip on continuation. Either tighten the stop, wait for a higher From, or accept the risk. The probability + days-to-target tell you what you're actually trading.

Pacing the trade by median days

If the median From → To climb takes 12 days, an option short 7 DTE is in a different timing risk than a 30 DTE position. The median + longest stats let you match contract duration to the typical move's pace, not your gut.

Cross-product comparison

Run the same From / To pair across UVXY, UVIX, VXX, SQQQ, SOXS, ZSL, BOIL one ticker at a time. The probabilities differ: UVXY (1.5x VIX) and UVIX (2x VIX) cycle differently than the inverse-equity products. Pick the product whose conditional profile fits the trade you want.

Auditing the sample with the event log

A 60% probability on 8 cycles is noise. The same probability on 80 cycles is signal. Open the per-event log to see how the From-hit count was built and whether the recent few years look anything like the older history. Regime drift hides in the dates.