Docs

Benchmark & alpha

On-chain benchmark rates, read from the protocols' own contracts, with three distinct ways to compare them against your book.

A 20% return is a thin margin over a passive deposit that would have earned 18%. Making that comparison needs the rate that actually existed on-chain over the same dates, and a clear statement of what kind of comparison is being made.

Benchmark catalog#

Benchmarks are indexed once per day per market from 14 protocol integrations, plus composites defined from them.

FamilyProtocols
LendingAave V2, Aave V3 (all markets), Aave V4 (hub and spoke), Spark, Compound V3, Fluid, Euler V2, Morpho Blue markets, Gearbox V3 passive pools
Savings rateSky (SSR)
VaultsMorpho V2 curated vaults, Symbiotic deposit vaults
Operator-NAV fundskpk (kETH, kUSD), Midas mTokens
CompositeA constant-weight blend of two or more catalog benchmarks

Not every market publishes every field. Borrow rates, utilisation and TVL exist only where the protocol has them; vault-style products publish a share price instead.

How rates are measured#

  • Realised from the protocol's indexWhere the protocol exposes a cumulative index (Aave, Spark and similar pools via their liquidity index; Morpho V2, Gearbox, Symbiotic, Midas and Aave V4 via a supply or share index), the supply rate is what the index actually grew between two reads, not the instantaneous rate quoted at one block.
  • Instantaneous otherwiseOther markets publish the rate the protocol reports at the snapshot block.
  • Rewards are a separate fieldIncentive APR is stored apart from the base rate. When the reward rate is unknown, net is empty rather than silently equal to gross.
  • Intrinsic yield is separateFor wrappers such as liquid-staking tokens, the yield of the underlying is kept apart from the market rate.
Net supply
APRnet=APRsupply+APRrewardAPYnet=APYsupply+APYreward\text{APR}_{\text{net}} = \text{APR}_{\text{supply}} + \text{APR}_{\text{reward}} \qquad \text{APY}_{\text{net}} = \text{APY}_{\text{supply}} + \text{APY}_{\text{reward}}

Rewards are not compounded into the supply rate.

Intrinsic yield
1+rtotal=(1+rmarket)(1+rintrinsic)1 + r_{\text{total}} = (1 + r_{\text{market}})(1 + r_{\text{intrinsic}})
Realised supply rate from an index
APR=I1/I0−1Δt×SAPY=(I1I0)S/Δt−1\text{APR} = \dfrac{I_1 / I_0 - 1}{\Delta t} \times S \qquad \text{APY} = \left(\dfrac{I_1}{I_0}\right)^{S / \Delta t} - 1

I is the protocol's cumulative index and Δt the seconds between the two reads. S = 31,536,000 seconds in a 365-day year.

APR to APY
APY=(1+APRS)S−1\text{APY} = \left(1 + \dfrac{\text{APR}}{S}\right)^{S} - 1

Per-second compounding, matching how lending protocols accrue interest. S = 31,536,000.

7, 30 and 90-day APY
APYn=(IendIstart)365/n−1\text{APY}_n = \left(\dfrac{I_{\text{end}}}{I_{\text{start}}}\right)^{365/n} - 1

Computed only when both index reads exist exactly n days apart. The reward component over the same window is an arithmetic mean.

Share price is published as a multiplier (1.05 means one share is worth 1.05 units of the underlying), not a percentage.

Composite benchmarks#

A composite blends catalog benchmarks with fixed weights, for example 60% Aave V3 USDC + 40% Morpho USDC, to match a multi-protocol mandate.

Composite rate
APRc=∑iwi APRiAPYc=(1+APRcS)S−1\text{APR}_c = \sum_i w_i \, \text{APR}_i \qquad \text{APY}_c = \left(1 + \dfrac{\text{APR}_c}{S}\right)^{S} - 1

Weights sum to 1. S = 31,536,000. APY is derived from the blended APR, not averaged across constituents.

  • Rebalanced every dayWeights are constant, so the composite behaves as if it were rebalanced at every interval. A real portfolio drifts; part of the gap against a composite is drift, not skill.
  • All constituents or nothingIf any constituent is missing for a day, the composite has no value for that day. It is never computed from the remaining legs.
  • Supply onlyComposites publish supply and net rates only: no reward split, no borrow rates and no share price. Net is never downgraded to gross.

Benchmark risk statistics#

For benchmarks with a share price, the API computes risk statistics over 1M, 3M, 1Y or the full history, from day-over-day share-price returns.

Return and volatility
R=P1P0−1CAGR=(1+R)365/d−1σ=s(r)×365R = \dfrac{P_1}{P_0} - 1 \qquad \text{CAGR} = (1 + R)^{365/d} - 1 \qquad \sigma = s(r) \times \sqrt{365}

P is the share price. s(r) is the sample standard deviation of consecutive daily share-price returns.

Sharpe, Sortino, drawdown and Calmar
Sharpe=r−rf‾s(r)×365Sortino=r−m‾σd×365Calmar=CAGR∣MaxDD∣\text{Sharpe} = \dfrac{\overline{r - r_f}}{s(r)} \times \sqrt{365} \qquad \text{Sortino} = \dfrac{\overline{r - m}}{\sigma_d} \times \sqrt{365} \qquad \text{Calmar} = \dfrac{\text{CAGR}}{|\text{MaxDD}|}

MaxDD is the largest fall of the share price from a running peak. r_f and the Sortino MAR both default to 0 on this endpoint unless passed in. σ_d is the downside deviation below that MAR.

Three ways to compare#

These answer different questions and are not interchangeable.

1. APR spread (position drawer)
spreadt=APRtpos−APRtnet\text{spread}_t = \text{APR}^{\text{pos}}_t - \text{APR}^{\text{net}}_t

A daily difference of two annual rates, in percentage points. Benchmark net supply APR falls back to supply APR for a single-protocol benchmark. With returns set to MD, position APR is the Modified Dietz APR. Days without a benchmark value are kept and flagged, and the summary averages state how many days were compared.

2. Excess XIRR (money-weighted)
excess XIRR=XIRRposition−XIRRbenchmark\text{excess XIRR} = \text{XIRR}_{position} - \text{XIRR}_{benchmark}

The benchmark XIRR replays the position's own deposits and withdrawals as if they had earned the benchmark's daily APY. Same capital, same timing, different return. Not available for composite benchmarks.

3. Strategy excess (time-weighted)
excess=1+Rstrategy1+Rbenchmark−1\text{excess} = \frac{1 + R_{strategy}}{1 + R_{benchmark}} - 1

Geometric excess of a strategy's TWR over its benchmark's index, on days both were observed. See Strategies for how the benchmark index is built.

Strategies → Comparing with a benchmark

Fund versus BTC, ETH and SOL#

The performance view can chart the portfolio against buy-and-hold BTC, ETH and SOL. The portfolio leg is Modified Dietz (P&L over the window ÷ time-weighted average capital); a strategy subject uses its TWR. The comparables are pure price returns from daily closes. Each series is rebased to 0% at the start of the selected window.

What Syncrone means by alpha#

Syncrone does not publish a single "alpha" number. Excess XIRR and strategy geometric excess are excess returns over a chosen benchmark. Neither is risk-adjusted (no regression on a market factor), so neither is Jensen's alpha. Choose the one that matches the question: money-weighted for "what did this capital earn versus leaving it in the protocol", time-weighted for "how did the manager perform versus the mandate".

Sources#

Where this lives in the app#

Next →How Syncrone works