Prime Algorithms: Data & Methodology

 At Prime Algorithms, our investment strategies are driven by strict, rules-based quantitative engines. This page outlines the mathematical definitions, data-handling procedures, and performance-calculation methodologies used across our strategy tearsheets and performance dashboards. 

 Our algorithmic trading models are built to systematically balance risk and reward. We rely on quantitative signal strength, multi-strategy consensus, and rigorous risk-off triggers rather than discretionary human trading. 

 All performance metrics are calculated based on the weekly closing prices of the underlying equities and benchmark indices.

• Rolling Returns (1-Yr, 3-Yr, 5-Yr, 10-Yr): Calculated on a trailing basis from the specific 'Data as of' date listed on the strategy profile. These figures represent the cumulative compounding return over the stated period.
• Positive Year Percentage: The historical probability that the strategy will close a full calendar year with a positive absolute return, calculated across all full calendar years since the strategy's inception.
• Inception Date: The date of the earliest available historical data point for the strategy's underlying algorithmic signal. For many of our core strategies, data extends back to 1990.

 We utilize standard quantitative measures to evaluate the risk-adjusted performance of our strategies.

• Lifetime Volatility: The annualized standard deviation of the strategy's weekly returns since its inception date.
• Lifetime Max Drawdown: The largest single historical peak-to-trough drop in the strategy's total portfolio value, expressed as a percentage.
• Sharpe Ratio (52W): A measure of risk-adjusted return over the trailing 52 weeks. It is calculated by subtracting the risk-free rate from the strategy's return, then dividing the result by the strategy's annualized volatility.
• Sortino Ratio (52W): Similar to the Sharpe ratio, but penalizes only downside volatility. It measures the trailing 52-week return relative to the strategy's downside deviation.

 To provide context for a strategy's performance, we evaluate it against a strict cohort of its direct peers (strategies that use the exact same position-sizing and weighting methodology).

• Consistency Ranks: Rather than highlighting a single fortunate week, we measure the total percentage of a strategy's lifespan spent ranked in the Top 10 of its peer group across Return, Volatility, Sharpe, and Sortino metrics.
• The Efficient Frontier: A mathematical curve representing the set of portfolios that offer the highest historical return for a given level of risk.
• Efficient Frontier Reign: We track the total number of weeks and the longest uninterrupted streak of weeks that a strategy has maintained an optimal mathematical position on the Efficient Frontier.

 Our models are built on varying algorithmic mandates, ranging from raw, unconstrained momentum to highly controlled multi-strategy consensus.

• Combined vs. Merged Architectures: Our "Combined" models aggregate sub-strategies by prioritizing peak individual algorithmic conviction. In contrast, our "Merged" models utilize a penalized average-ranking system to enforce multi-algorithmic consensus, inherently dampening idiosyncratic risk.
• Turnover Calculation: Strategy turnover is calculated on a per-cycle (weekly) basis. It is defined mathematically as the average of the absolute changes in target position weights between rebalancing periods. The figure displayed on our tearsheets represents the historical average of this weekly rebalancing friction.

• Hypothetical & Backtested Data: The historical performance data presented on this website and in our strategy tearsheets include hypothetical and backtested results. These results were prepared with the benefit of hindsight and do not reflect the impact of material economic or market factors that might have affected decision-making if client funds were actively traded during those historical periods.
• No Guarantee of Future Results: Financial markets are dynamic and constantly evolving. Strategies, signals, and mathematical models that have demonstrated success in historical backtesting may fail to perform in different future market regimes. Past performance is strictly not indicative of future results.
• Algorithmic Risk: Prime Algorithms relies heavily on proprietary quantitative models, software execution, and third-party data feeds. Anomalies in market data, coding errors, or software interruptions could result in unintended portfolio allocations and material financial losses.