# Volatility targeting

Volatility targeting is a position-sizing approach that adjusts exposure according to estimated volatility so the portfolio or position aims for a specified level of return variability.

## What it means

Volatility targeting changes the size of an exposure when its estimated volatility changes. Higher estimated volatility generally means a smaller exposure; lower estimated volatility generally means a larger one, subject to whatever limits apply. The target is a level of *return variability*, not a target return or a maximum loss.

The approach can apply to one stock, a strategy, or an entire portfolio, but the volatility estimate must match the exposure being sized. A stock’s volatility is not necessarily the volatility of a basket containing that stock: the basket also depends on position weights and how its holdings move together.

## How it is calculated

A simple sizing rule is `target weight = baseline weight × (target volatility / estimated volatility)`.

- **Target weight:** the fraction of portfolio value assigned to the exposure after scaling, before any cap; express it as a percentage or decimal.
- **Baseline weight:** the fraction assigned before volatility scaling, in the same units as target weight.
- **Target volatility:** the desired standard deviation of returns over a stated horizon, such as an annualized percentage.
- **Estimated volatility:** the exposure’s estimated return standard deviation over that *same* horizon, expressed in the same units as target volatility.

In this simple rule, both volatility inputs must be positive. The ratio is unitless. It scales a baseline position; it does not necessarily make the *whole portfolio* hit the target, particularly if other holdings remain unchanged. For a single-exposure portfolio with cash as the remainder, the corresponding approximation is `target exposure weight = target portfolio volatility / estimated exposure volatility`, assuming cash has negligible volatility and the estimate holds. A result above 100% implies leverage; an unleveraged account would need to cap it. Position limits or other constraints can also reduce the calculated weight.

## Worked example

Suppose, hypothetically, an investor has a US$100,000 portfolio and a baseline allocation of 20% to one stock. The target and estimated volatilities for that *stock exposure* are both annualized: 12% and 24%, respectively.

1. Calculate the scaling ratio: `12% / 24% = 0.5`.
2. Scale the baseline weight: `20% × 0.5 = 10%`.
3. Convert the weight into dollars: `US$100,000 × 10% = US$10,000`, versus the baseline US$20,000.

This example halves the stock allocation; it does **not** establish that the resulting portfolio has 12% annualized volatility. That would require an estimate for the entire portfolio, including the other holdings and their correlations with the stock. Dollar amounts describe exposure, not an expected gain or loss.

## Why it matters before you trade

A planned quantity might leave a position larger than intended relative to its current volatility. Comparing the intended weight with a volatility-scaled weight makes that difference visible before an order is placed. The check should use a stated estimation period and horizon, and should account for existing holdings rather than treating a new trade in isolation.

Volatility estimates can change, and transaction costs can accumulate if frequent changes lead to repeated trading. Even a position that fits a volatility target can suffer a sudden gap, a liquidity shortfall, or a loss much larger than its typical daily move. The target is a sizing convention, not a guarantee that realized risk will stay fixed.

## How PreTrAIde uses it

PreTrAIde does not claim to run a volatility-targeting sizing rule. Its pre-trade analysis evaluates a user-built stock basket before an order is placed. Its quantitative layer estimates daily volatility using blended estimators, calculates 1-day 95% value at risk, and produces 0–100 volatility scores for stocks and the basket. It also provides risk assessments and portfolio breakdowns. These outputs can inform a separate volatility-targeting check, but a PreTrAIde verdict is decision support, not an instruction to change a position. Users place trades themselves through their own brokers.

## Common mistakes

- **Mixing horizons:** dividing annualized target volatility by daily estimated volatility produces an invalid scaling ratio.
- **Confusing an individual stock with the portfolio:** correlations and other position weights matter to portfolio volatility.
- **Treating the calculated weight as unconstrained:** leverage, position limits, available cash, and execution costs may prevent it from being implemented.
- **Equating volatility with total risk:** standard deviation does not rule out gaps, extreme losses, or hard-to-exit positions.
- **Assuming the estimate is stable:** a volatility estimate is based on a method and available data, not on known future returns.

---
Source: https://pretraide.com/glossary/volatility-targeting