Tool Tutorials

ETF Decay Projection Tutorial

Chapter 1

The All-Time Low Anchor

The Decay Projection Tool anchors its analysis to an asset's running all-time low (ATL) to model future 45-session price trajectories.

Leveraged and inverse volatility ETFs (e.g., UVXY) reset their exposure daily. On range-bound or mean-reverting underlying assets, this daily compounding creates severe volatility drag (decay), causing these instruments to drift toward new record lows over time.

1. The Running ATL Floor

The anchor used by the analyzer moves, but only ever downward: it steps down whenever a new record low is printed and stays flat until a lower low arrives. It never steps upward.

Price Path Running ATL Floor Current Spot ($25.09) +11% Spike Active ATL Floor ($22.66) Last New Low (7 Sessions Ago)
The running ATL floor steps down at each new record low and holds flat until the next one. The spike reading measures the current spot against whichever level the floor is holding today.

Why Measuring Against Running ATL Matters

Because decay continuously drives leveraged volatility products down, measuring price relative to a fixed historic dollar level (e.g., a low from two years ago) is non-informative.

Tracking percentage distance above the running ATL in force at that specific point in time keeps historical comparisons normalized across the entire life of the ETF.

2. Core Inputs & Formula

The tool converts raw market data into two primary variables:

Spike % from ATL = (Current Spot Price / Running All-Time Low − 1) × 100
Sessions Since Low = Trading days elapsed since the active ATL was printed

Analytical Reading Example

MetricSample ValueDescription
Spot Price$25.09Current closing price of the security.
Running ATL$22.66Lowest price recorded in the asset's history up to today.
Spike % from ATL+11%Normalized distance above the active floor.
Age of Low7 SessionsNumber of trading bars elapsed since the $22.66 low was set.
Note

These two figures form the complete input criteria for the historical path-matching engine. No manual lookback windows or arbitrary date filters are required.

See the current reading
Chapter 2

Historical Matching & Sample Construction

Rather than evaluating absolute dollar levels, the engine matches the asset's current percentage distance above its running all-time low (ATL) against all historical instances where that exact distance was first reached.

1. Matching Rules & De-Duplication

To construct a statistically clean sample, candidate sessions must meet two primary criteria:

  • Dynamic ATL Adjustment: The baseline all-time low is evaluated using the floor in force at that exact point in history. A +11% move above a $60.00 historic ATL carries identical weight to a +11% move above a $22.66 current ATL.
  • One Match Per ATL Epoch: To prevent multi-week consolidation at a level from cluttering the dataset with collinear data points, only the first session to reach the target threshold under a given ATL is recorded.

Trigger Condition

Match Event when Bar High ≥ Running ATL × (1 + Target Spike % / 100)
New ATL Established First High ≥ Target Recorded as Valid Match Forward 45-day path tracked Subsequent Sessions ≥ Target EXCLUDED (Prevents sample crowding) Count resets only when a NEW ATL forms
Each new all-time low opens one window in which a single match can be recorded. The first session to reach the target is kept, and everything after it is excluded until a lower low resets the count.

2. Low-Volatility Baseline Exception (<5% from ATL)

When an asset trades within 5% of its all-time low, directional distance matching becomes non-informative. In this regime, the system automatically transitions to a VIX-Regime Sampling Engine:

  • The dataset switches from tracking ATL percentage spikes to analyzing all historical 45-session windows that began under equivalent cash VIX regimes: <20, 20-30, or >30.
Chapter 3

The Price Cone (Percentile Distribution)

The Price Cone aggregates all forward trajectories generated by the historical sample across a 45-trading-day window (roughly 9 calendar weeks) and converts them into distribution bands.

Inner Band (66% Confidence) Outer Band (80% Confidence) Median Path Upper 90th Percentile ($30.28) Spot ($25.09) ATL ($22.66) Median Path (−32% at Day 45) Lower 10th Percentile ($11.96) Day 0 (Today) Day 45 (9 Weeks)
The bands open out from the current spot as the sessions pass. On this sample the median path drops below the ATL line inside the first three weeks, while the upper boundary still reaches $30.28.

Median vs. Mean Path

Unlike standard probability cones that display arithmetic averages, this model utilizes the median (50th percentile) path. Because volatility products experience asymmetric upward spikes, extreme outlier events skew the mean upward.

Sample Divergence Example
Historical Mean (45-Day): −19% (Distorted by hyper-spikes)
Historical Median (45-Day): −32% (Reflects typical path)

Band Definitions

ElementPercentile BoundariesIncluded Sample ShareDescription
Spot LineN/AN/ACurrent trading price of the asset.
ATL LineN/AN/AActive running all-time low reference floor.
Median Line50th Percentile50% above / 50% belowCenter of historical path outcomes.
Inner Band17th to 83rdCentral 66%Primary expected outcome corridor.
Outer Band10th to 90thCentral 80%Expanded outcome corridor.
Tails<10th & >90thOuter 20%Split evenly: 10% extreme bull / 10% extreme bear.
Open the price cone
Chapter 4

Probability Graph & Extended Multi-Year Projections

1. The Probability Graph: Touch vs. Close Logic

The Probability Graph evaluates how frequently matched paths hit or exceed a user-defined threshold (e.g., a 10% move) across weekly intervals.

PATH CROSSES THRESHOLD (−10%) BY TOUCH EVALUATION Triggered once High/Low breaches level State is PERMANENT for all future weeks Ideal for: Stop Losses / Exit Targets BY CLOSE EVALUATION Evaluates SOLELY on weekly close price If price recovers above line on Friday, path is NOT counted for that week
One threshold, two ways of counting it. By Touch keeps a path from the moment it breaches the level, while By Close asks the question again at each weekly settlement.

Differences in Execution

ModeDefinitionPrimary Analytical Use Case
By TouchCumulative count of paths whose intraday High or Low reached the target at any point up to that expiration date.Assessing execution probability for stop-loss orders or take-profit limit orders.
By CloseDiscrete count of paths whose Friday settlement price closed beyond the target level.Assessing exposure for options expiration, weekly mark-to-market, or strategic holding periods.

2. Extended Path Table (Multi-Year Projections)

The distribution table extends the cone's statistical engine out across a 2-year time frame (one row per weekly expiration).

Time Frame90% Probability Above83% Probability Above50/50 Median83% Probability Below90% Probability Below
Current (Spot)$25.09$25.09$25.09$25.09$25.09
1 Year OutLow Tail BoundaryLower Inner Edge$4.36Upper Inner EdgeHigh Tail Boundary
2 Years OutLow Tail BoundaryLower Inner Edge$1.09Upper Inner EdgeHigh Tail Boundary
Analytical Note on Long-Time Frame Decay

Because leveraged volatility products are worn down by roll yield contango, long-time frame distributions trend relentlessly lower. Projections beyond 6 months indicate the overall decay vector of the sample rather than an absolute price target.

Open the probability graph Open the table

Tutorial complete

You answered all 4 questions. Final score: 0 / 4

Open the ETF Decay Projection tool →