Volatility Targeting: Sizing Different Assets to Comparable Risk

The same invested amount can leave one asset quiet all day while another shakes the account within an hour. The issue is sizing only by money. Volatility targeting first asks how much daily fluctuation to accept, then works backward to quantity.

Equal Amounts Do Not Mean Equal Risk

Fixed notional produces very different risks across assets. Put $800 into three assets and calculate the daily dollar fluctuation.

In an equal-dollar portfolio, C determines most account fluctuations while the others contribute relatively little. The feeling of diversification differs from actual risk allocation. See volatility and why crypto varies so much for the causes.

The Calculation Is One Division

Reverse the sequence: choose a tolerable daily dollar fluctuation, then divide it by the asset's volatility.

These are amounts that equalize risk, not judgments that one asset deserves more investment. The calculation contains no forecast of which is more promising. It separates expected-return judgment from risk allocation.

Two Ways to Measure Realized Volatility

The denominator uses recent observations, not future information. Two common methods are:

ATR% is convenient if working entirely from charts. See ATR for stops and position size for the procedure. Fix the window, such as 20 days or 14 bars, and do not change it between assets; unequal windows invalidate comparison.

The Exposure Multiple Needs a Cap

Unrestricted division keeps increasing size as volatility falls. A long quiet period can make the formula produce dangerous amounts.

A quiet market is not guaranteed to remain quiet. Low- and high-volatility periods cluster, and transitions can arrive without warning. A cap arithmetically limits exposure during the transition. Choose it by whether the worst day is tolerable, not preference.

Rebalancing Frequency and Fees

Volatility changes daily, so target notional changes daily. Matching it every day accumulates fees.

This sacrifices some precision to reduce costs substantially. The certain fee can exceed the benefit of matching notional to decimal-level accuracy every day. Profit factor calculations show how even a few basis points of fees can materially change gross-profit-to-loss ratios.

When It Conflicts with Stop-Based Sizing

Stop distance provides another common sizing method. The two calculations can give different answers.

They answer different questions. Stop-based sizing controls loss if this trade is wrong; volatility-based sizing controls fluctuations while the position is held. If a trade remains open for days without reaching its stop, the latter better reflects the experience. Taking the smaller value is the more conservative choice. See position-sizing calculations for working backward from stop distance.

Ignoring Correlation Lets Risks Add Up

Equalizing each asset's risk does not resolve total risk if they move together.

Several same-direction altcoin holdings can resemble a triple-sized position in one asset. Keeping each asset's original target can triple the account-level target. Define the account budget first, then allocate it across assets. Volatility drag explains the asymmetric recovery burden after large fluctuations.

What This Method Cannot Solve

Its limitations should be explicit.

The fourth point is often misunderstood. Uniform loss sizes can smooth the equity curve, but streaks still occur. Losing-streak probability shows how they remain statistically normal even for a reasonable win-rate strategy. The exchange leverage setting is separate: notional determines exposure, while the setting determines locked margin. Read isolated and cross margin for that distinction.

Key Points

Choose the daily fluctuation budget before choosing the invested amount. Once fixed, quantity follows by division as assets and conditions change. Without it, quiet assets tend to be undersized and risky assets oversized.