Avoiding Survivorship Bias in Options Backtests

Avoiding Survivorship Bias in Options Backtests

Survivorship bias can quietly invalidate an options backtest.

If you build a strategy using today’s list of optionable stocks and apply it backward in time, you are almost certainly overstating historical performance.

Why?

Because not all stocks had options in the past — and not all stocks had weekly expirations.

If your backtest assumes they did, your universe is incorrect.

What Is Survivorship Bias in Options Trading?

Survivorship bias occurs when a backtest includes only securities that exist today, ignoring:

• Stocks that delisted

• Stocks that lost options eligibility

• Stocks that did not yet have options listed

In equities, this is already a known problem.

In options trading, it’s even more dangerous — because the existence of an options chain itself changes over time.

The Hidden Layer: Option Availability Bias

Most traders focus on price history, but options strategies depend on structural features:

• Whether the stock had listed options

• Whether weekly expirations existed

• How dense the expiration calendar was

If you backtest a weekly options strategy in 2012 using a universe of stocks that have weeklies today, you are introducing structural lookahead bias.

Many of those stocks did not have weekly expirations at that time.

Example: Weekly Options Expansion

Weekly listings have expanded significantly over the past decade.

If you:

1. Pull today’s list of stocks with weekly options

2. Apply a 2010–2015 backtest

3. Assume those stocks had weekly expirations throughout

You are overstating:

• Trade frequency

• Liquidity

• Strategy feasibility

This can materially distort Sharpe ratios and drawdowns.

Why This Matters for Systematic Traders

Survivorship bias in options backtests can lead to:

• Inflated historical returns

• Unrealistic trade counts

• Misestimated slippage

• Invalid risk modeling

For cross-sectional options strategies — especially short-dated volatility or weekly expiration systems — correct universe construction is foundational.

The Correct Approach: Point-in-Time Optionable Universes

To avoid survivorship bias, you must reconstruct the universe as it existed at each historical date.

That means:

• Only including stocks that had listed options at that time

• Confirming expiration structure (monthly vs weekly)

• Avoiding forward-looking universe filters

The cleanest solution is to use point-in-time snapshots.

Programmatic Retrieval of Historical Optionable Tickers

The Alphanume Historical Optionable Tickers dataset provides monthly point-in-time snapshots of stocks with listed options.

Endpoint:

Each snapshot corresponds to the first trading day of the month.

Example request:

import requests

url = "https://api.alphanume.com/v1/optionable-tickers"
params = {
    "date": "2026-02-02"
}

r = requests.get(url, params=params)
print(r.json())

import requests

url = "https://api.alphanume.com/v1/optionable-tickers"
params = {
    "date": "2026-02-02"
}

r = requests.get(url, params=params)
print(r.json())

import requests

url = "https://api.alphanume.com/v1/optionable-tickers"
params = {
    "date": "2026-02-02"
}

r = requests.get(url, params=params)
print(r.json())

import requests

url = "https://api.alphanume.com/v1/optionable-tickers"
params = {
    "date": "2026-02-02"
}

r = requests.get(url, params=params)
print(r.json())

Example response:

{
  "count": 1,
  "data": [
    {
      "date": "2026-02-02",
      "ticker": "NVAX",
      "avg_days_between": 7.0,
      "has_weeklies": 1
    }
  ]
}

{
  "count": 1,
  "data": [
    {
      "date": "2026-02-02",
      "ticker": "NVAX",
      "avg_days_between": 7.0,
      "has_weeklies": 1
    }
  ]
}

{
  "count": 1,
  "data": [
    {
      "date": "2026-02-02",
      "ticker": "NVAX",
      "avg_days_between": 7.0,
      "has_weeklies": 1
    }
  ]
}

{
  "count": 1,
  "data": [
    {
      "date": "2026-02-02",
      "ticker": "NVAX",
      "avg_days_between": 7.0,
      "has_weeklies": 1
    }
  ]
}

Of course — picking up where it left off.

Returned fields include:

• date — Snapshot date

• ticker — Stock symbol

• avg_days_between — Average spacing between the next six consecutive expirations

• has_weeklies — Indicator for consecutive weekly expirations

This allows you to:

• Construct historically accurate optionable universes

• Filter by weekly expiration availability

• Control for structural changes in expiration density

• Backtest strategies without assuming forward knowledge

Understanding Expiration Density

The avg_days_between metric reflects how frequently options expired at that point in time.

Values near 7 indicate dense weekly expiration structures.

Higher values suggest:

• Biweekly spacing

• Monthly-only listings

• Less consistent short-dated coverage

If your strategy depends on weekly expirations, filtering on:

or

ensures your universe reflects structural reality — not today’s listings projected backward.

A Simple Illustration of the Bias

Assume you are testing a short-dated volatility strategy from 2012 to 2020.

Incorrect method:

1. Pull today’s list of stocks with weekly options

2. Apply the strategy across historical prices

3. Assume weekly expirations always existed

Correct method:

1. For each month in your backtest

2. Retrieve the optionable universe as of that date

3. Filter for stocks that actually had weekly expirations

4. Run your strategy on that point-in-time universe

The difference can materially affect:

• Trade count

• Capital deployment

• Return dispersion

• Tail risk

Why Monthly Snapshots Are Sufficient

The optionable universe does not typically change daily in large increments.

By capturing the universe on the first trading day of each month, you obtain:

• Stable point-in-time references

• Reduced noise

• Efficient storage and retrieval

• Sufficient granularity for most systematic strategies

For most research workflows, monthly universe reconstruction eliminates the majority of survivorship distortion.

Common Mistakes in Options Backtests

Even experienced traders unintentionally introduce bias by:

• Using current index constituents

• Assuming weekly expirations existed historically

• Ignoring when stocks first became optionable

• Applying liquidity filters based on today’s market structure

The result is often subtle — but statistically meaningful — overstatement of performance.

When Survivorship Bias Matters Most

You should be especially careful if your strategy involves:

• Short-dated options

• Weekly expiration rolls

• Earnings-based volatility trades

• Cross-sectional ranking of optionable stocks

• Small- and mid-cap equities

These areas are highly sensitive to structural availability.

Summary

Avoiding survivorship bias in options backtests requires more than clean price data.

You must reconstruct:

• Which stocks had options

• When they had options

• Whether weekly expirations existed

• How dense the expiration structure was

Assuming today’s optionable universe existed historically can significantly distort results.

Point-in-time optionable snapshots provide the structural layer necessary for serious options research.

If you are building systematic options strategies, accurate universe construction is not optional — it is foundational.