A gold strategy finishes above a broad equity index. Has it demonstrated skill? The comparison alone cannot answer. It might reveal a difference in asset exposure, leverage, currency or accounting rather than anything the strategy contributed.
This research note develops a comparison method. It contains no measured returns and makes no claim that a Monolith system beats a benchmark.
Give each comparison a job
An opportunity-cost benchmark asks what a stated alternative would have delivered. An exposure-aware benchmark asks whether the strategy contributes beyond the risks it takes. A mechanism baseline removes the special feature while keeping the rest of the construction comparable. These are useful but distinct editorial categories: one chart need not pretend to answer all three questions.
Start with the investment universe and the claimed contribution. For a directional commodity system, passive exposure to the same instrument may be a more diagnostic baseline than equities. Cash can clarify the value of taking risk at all, but does not explain where a risky strategy’s return comes from. Sharpe’s discussion of differential returns and portfolio risk positions makes the reference investment and scale substantive parts of performance analysis.[1]
Write down the currency, assets, leverage or exposure rule, rebalancing schedule, fees, financing and return convention. Distinguish a price-only series from a return series including income. The GIPS standards require relevant benchmarks and prohibit price-only benchmarks in GIPS composite and pooled-fund reports.[3] That is a cited reporting standard, not a claim that this notebook or its simulations comply with GIPS.
Conceptual example
Consider a hypothetical equity strategy whose accounting credits dividends, compared with a price-only index that omits them. Even with matching dates, the gap mixes portfolio decisions with inconsistent income treatment. Rebuild the comparison on a stated common income basis before attributing the difference to skill.
Next suppose the strategy uses borrowing while the benchmark is unlevered. Matching the income convention still leaves a risk mismatch. A separate, predeclared exposure-scaled baseline can investigate that mismatch, including its funding costs. It should not replace the original opportunity-cost comparison silently, especially if the scaling was chosen after inspecting the outcome.
Cheap check
Create a comparison specification before calculating the gap. Trace the opening valuation, closing valuation and every intervening observation for both series. Check the same decision window, currency and income treatment; document any unavoidable difference in implementation costs. A return starting on a date may require the preceding eligible close as its opening mark, so date labels alone are insufficient.
Pass: both series reconcile to their documented construction, windows align, and the caption states which question the comparison answers. Exposure differences are either controlled by a predeclared rule or explicitly left uncontrolled.
Fail: the apparent advantage depends on a missing dividend, omitted funding charge, shifted opening mark or retrospectively chosen comparison. Withdraw the attribution claim and correct the identified mismatch. Do not search for a more flattering index.
Limitations
Matching historical volatility does not match tail risk, liquidity or time-varying exposure. An index is not automatically an investable portfolio. A correctly reported cumulative difference is still not proof of independent alpha; selection and sampling uncertainty remain after accounting is repaired.
Read the estimator contract first, then examine drawdown units. The risk and reporting guide sets the wider evidence boundary.
Sources
[1] https://web.stanford.edu/~wfsharpe/art/sr/sr.htm — William F. Sharpe — The Sharpe Ratio (1994) [3] https://www.gipsstandards.org/wp-content/uploads/2021/03/2020_gips_standards_firms.pdf — CFA Institute — 2020 GIPS Standards for Firms