Five Pillars

The five weighted inputs behind the 7H Conviction Score: Fundamental, Technical, Model (CNN), Sentiment, and Valuation.

The Five Pillars represent a structured, multi-dimensional framework used to evaluate an asset's overall investment thesis. By breaking down an asset's profile into five distinct categories—Fundamental, Technical, Model (CNN), Sentiment, and Valuation—this methodology prevents reliance on any single analytical style. Each pillar is assigned a specific, predetermined weight reflecting its historical reliability and importance to the overall conviction score: Fundamental carries 30%, Technical 25%, Model/CNN 20%, Sentiment 15%, and Valuation 10%. Investors care about this multi-pillar approach because it enforces analytical discipline and mitigates cognitive bias. For instance, a stock might look exceptionally cheap on valuation metrics but suffer from deteriorating fundamentals and negative technical momentum. By aggregating all five dimensions, the resulting score provides a holistic view that tempers over-optimism in any single area. A key strength of this framework is its dynamic handling of missing data. If a specific pillar cannot be calculated—such as valuation metrics for a pre-revenue startup—the system does not assign a default score of zero, which would artificially drag down the asset's rating. Instead, the missing pillar is excluded entirely, and the weights of the remaining active pillars are proportionally rescaled to sum to 100%, ensuring the final score remains accurate and unbiased.

An asset has the following individual pillar scores (out of 1.0): Fundamental = 0.8, Technical = 0.6, Model = 0.7, Sentiment = 0.5, and Valuation is missing. Normally, the weights are 30%, 25%, 20%, 15%, and 10%. Since Valuation is missing, the remaining weights (30%, 25%, 20%, 15%) sum to 90%. To rescale, each remaining weight is divided by 0.90, resulting in new weights of approximately 33.3% Fundamental, 27.8% Technical, 22.2% Model, and 16.7% Sentiment. The final score is calculated as: (0.8 0.333) + (0.6 0.278) + (0.7 0.222) + (0.5 0.167) = 0.2664 + 0.1668 + 0.1554 + 0.0835 = 0.672 out of 1.0.