Performance Attribution
Learn how to attribute crypto portfolio performance to allocation, selection, carry, rewards, fees and execution rather than relying on total return alone.
Reading progress — saved on this device
Performance attribution explains where portfolio returns came from—asset weights, security selection, yield, funding, fees and execution—so investors can distinguish repeatable decisions from favourable market moves.
Learning objectives
- Calculate simple single-period return contributions from beginning weights and returns.
- Distinguish contribution analysis from benchmark-relative attribution.
- Classify crypto-specific sources such as staking rewards, funding, airdrops, fees and execution costs consistently.
What it is
Contribution analysis asks how much each asset, sleeve or strategy added to the portfolio's total return. In a simple single period without external-flow complications, beginning weight multiplied by asset return provides an intuitive contribution estimate.
Attribution goes further by comparing the portfolio with a benchmark or policy allocation. It asks whether relative performance came from overweighting a segment, selecting better assets within it, or other effects. Brinson-style frameworks often separate allocation and selection, but exact interaction conventions differ.
Crypto adds return sources that ordinary spot attribution can miss: staking rewards, lending income, perpetual funding, futures basis, airdrops, validator penalties, protocol fees, gas, exchange fees and execution slippage.
How it works
Beginning-period weights are commonly used for simple contribution because end weights already embed the period's return. Using ending weights can therefore mix cause and outcome.
For a one-period, no-flow example, portfolio return is approximately the sum of asset or sleeve contributions. Multi-period attribution is more complex because weights change, returns compound and cash flows may occur. Arithmetic contributions cannot simply be added across many periods without an appropriate linking method.
Benchmark-relative attribution requires a clear reference. If a portfolio outperforms because it held 40% in a high-return satellite sleeve while the benchmark held 20%, that is partly an allocation decision. If both held the same sector weight but the portfolio chose the better token, that is closer to selection effect.
Return attribution should be paired with risk attribution. A sleeve contributing most of the return may also contribute most of the drawdown, volatility or tail exposure.
Measurement methodology
| Check | Purpose | What to verify |
|---|---|---|
| Classification | Defines attribution buckets | Use stable sector, strategy and income categories. |
| Beginning weights | Supports clean contribution math | Use pre-return weights for simple single-period contribution. |
| Benchmark | Defines active decisions | Match the portfolio's mandate and constraints. |
| Crypto cash flows | Prevents hidden return sources | Separate price P&L, staking, funding, rewards, fees and execution. |
Worked example and thought exercise
A portfolio begins with 60% in its core sleeve and 40% in satellites. During the period, the core returns 10% and satellites return 30%.
Approximate contribution from core = 60% × 10% = 6 percentage points. Satellite contribution = 40% × 30% = 12 percentage points. Total simple single-period return is therefore approximately 18% before any additional fees, flows or income adjustments.
Suppose the benchmark held 80% core and 20% satellites. Because satellites were the stronger segment, some of the portfolio's benchmark-relative outperformance came from the decision to overweight satellites—not merely from security selection.
Thought exercise: if a strategy earned 3% from token-price appreciation but paid 1% in funding and execution costs, what would be misleading about attributing the full 3% to strategy skill?
Common mistakes and practical workflow
- Using ending weights to explain a return that created those ending weights.
- Calling contribution analysis benchmark-relative attribution.
- Ignoring staking rewards, funding, gas, fees or slippage.
- Adding single-period attribution effects across long periods without a linking method.
Practical workflow
- Define the portfolio and benchmark classification structure.
- Capture beginning weights and reliable periodic returns.
- Separate price return from rewards, carry and costs.
- Calculate contribution, then benchmark-relative allocation/selection effects using one documented convention.
- Review return attribution alongside risk contribution and drawdown.
Knowledge checkpoint
- How is simple single-period return contribution calculated?
- What is the difference between contribution and attribution?
- Why are beginning weights generally preferable for simple contribution?
- Which crypto-specific return and cost sources should be separated from price P&L?
FAQs
❓ Does attribution prove investment skill?
No. It describes where realised relative return came from under a chosen methodology; repeatability must be assessed separately.
❓ What is allocation effect?
It broadly captures the impact of holding different segment weights from the benchmark, subject to the specific attribution convention used.
❓ Should staking rewards be included?
Yes, but classify them separately enough to distinguish price return from earned protocol rewards.
❓ Why include risk attribution too?
A return source can look attractive while consuming a disproportionate share of volatility, drawdown or tail risk.
Summary
Performance attribution turns total return into an auditable set of drivers. In crypto, useful attribution needs consistent beginning weights, a suitable benchmark, explicit treatment of rewards and costs, and a parallel view of the risk required to generate each return source.
Want this in a personalised order?
Take the crypto assessment and get a custom path of 10 modules matched to what you already know. Free, no card required.
Build my path →