ESG Ratings Methodology: Indicators, Normalisation, Weights & Scores

What you'll build

What you will be able to do

Each lab ends with a concrete result that you build yourself and the checker confirms, cell by cell.

• Measure how complete the raw ESG data is per company and per indicator, flag companies with too little data, and fill one gap with a median estimate.
• Rescale one indicator three ways, min-max to 0–100, z-score and percentile rank, and see how an outlier affects each method.
• Score two lower-is-better indicators and one higher-is-better indicator on 0–100 so that 100 always means best, then average them.
• Build the E, S and G pillar scores from six indicator scores, combine them with pillar weights into one overall score and find the leader.
• Look up each company's E, S and G weights from a sector materiality table, compute the sector-weighted score and compare it with equal weights.
• Translate each company's controversy level into a point deduction, apply it with a floor at zero, flag severe cases for a watchlist and measure the average impact.
• Convert scores into letter grades with an approximate-match lookup, find each company's band number and the points it needs for the next grade, and count the companies at BBB or better.
• Recompute the scores with a second set of weights, measure how ranks move, quantify agreement with another method using CORREL and plot the two methods against each other.