Every valuation method has blind spots. Graham's formula is famously conservative for high-growth names; a DCF is only as good as a decade-long forecast; Forward P/E leans on the market's own multiple logic, which can itself be wrong. Blending them is a hedge against any one method's specific weaknesses dominating the answer.
Blended IV = (Graham x 0.30) + (DCF x 0.40) + (Forward P/E x 0.30)
Why DCF gets the largest default weight. It's the only method built from the business's actual projected cash generation rather than a multiple-based shortcut — in principle, the most direct measure of what the business is worth, even though it's also the most assumption-sensitive of the three.
Why not just pick the best method? There often isn't a single "best" method — which one is most reliable depends on the business (a mature, low-growth company might be better served by Graham's formula; a business with heavy near-term reinvestment might be better served by the DCF). Blending avoids having to make that judgment call company-by-company, at the cost of a small amount of precision for any one method.
The weights are yours to change. In advanced mode, you can reweight the three methods however you'd like for a given company — the defaults are a reasonable starting point, not a claim that this exact split is optimal for every business.