Performance & Backtesting Policy
How MCH classifies performance outputs and the disclosures that accompany backtests, simulations and forecasts.
This policy governs how MCH Analysis classifies and discloses performance information: model returns, simulated research baskets and strategy indices, historical alpha, backtests, basket and benchmark comparisons, return forecasts and options scenario payoffs. It forms part of the MCH Analysis policy framework and is one of the research-governance standards, alongside the Research Integrity Policy.
This policy forms part of the MCH Analysis policy framework. Your use of the Service is governed by the MCH Analysis Subscriber Agreement and, for personal information, the Privacy and Cookie Notice; where this policy and either of them differ, the Subscriber Agreement or the Notice prevails.
Scope
This policy applies to every page and publication that shows:
- model returns
- simulated research baskets and strategy indices
- historical alpha
- backtests
- basket and benchmark comparisons
- return forecasts
- options scenario payoffs
Required classifications
Every performance output should be classified as one of:
- actual/live
- audited actual
- hypothetical
- simulated
- backtested
- paper
- forecast/expected
Different classifications must not be merged into one apparent historical-performance series without clear disclosure.
Backtest disclosures
Backtests are hypothetical unless explicitly identified otherwise. Where relevant, a backtest should disclose its assumptions, including:
- starting universe
- delisting treatment
- survivorship controls
- data availability
- point-in-time treatment
- rebalance date
- transaction costs
- slippage
- liquidity assumptions
- taxes
- dividends
- corporate actions
- execution assumptions
- leverage
- shorting costs
- options pricing assumptions
Expected returns
Expected returns, target prices and probability-weighted outcomes must be described as estimates, not as realised performance.
Historical returns
Historical model or simulation results must not be labelled in a manner that suggests they are returns actually earned by clients or by MCH, unless that is demonstrably true.
Charts
Every performance chart or table carries one of these classification labels beside it, not only in a distant footer (wording as approved by counsel):
- BACK-TEST — hypothetical, not actual trading results
- SIMULATED — hypothetical, assumptions apply
- LIVE — actual results of [strategy/account] since [date]
- FORECAST — estimate, not a result
Actual, paper, simulated, back-tested and forecast results are never combined into a single track record without explicit separation. Each presentation states or links to its universe, period, rebalance assumptions, dividend and corporate-action treatment, transaction costs, slippage, fees where relevant, taxes, and the benchmark and its material differences.
Questions about this policy: research@mchadvisoryservices.com. All policies: MCH Analysis legal.