Stock scores are widely used thorough out our platform. The idea behind stock scoring and ranking is to enable users to quickly analyse a stock’s standing with respect to other stocks in the desired universe.
Stock scores and ranks can be calculated for any metric over any desired universe. In fact, in our scorecard for stocks, we show stock’s score for all metrics over 4 universes – the entire market, stocks in the sector, stocks in the industry group and stocks in the industry.
The idea behind creating stock scores is to create a uniform comparison methodology for all metrics irrespective of the metric.
For the sake of uniformity, ranks are normalized to scores from 0 and 100 in such a way that 0 is the worst and 100 is the best.
Normalization is important because, for some metrics, lower value is better (P/E ratio) while for others, higher value is better (EBITDA margin). Normalizing each metric rank to scores between 0 and 100 in such a way that higher value is always better helps in better visualization and comparison.
Let’s take an example. Suppose a stock has a P/E ratio of 15.6 and is ranked 25th out of all stocks in the market (say 1600 stocks). This could translate into a high P/E score of 90.2 with respect to the market.
The stock looks relatively cheap when compared with other stocks in the market. Now suppose this stock is ranked 25th out of 34 stocks in the sector. This will translate into a low P/E score of 26.5.
Possible interpretation – While the stock looks cheap with respect to the broader market, the entire sector is cheap and within the sector, the stock is pretty expensive.
We also calculate the non-normalized ranks. Ranks are just the opposite of scores and are not normalized between 0 and 100. Methodology for calculating the rank is similar. Ranks are calculated in such a way that lower rank is always better.
User should note that when ranking (and normalizing) the stock on a particular metric, stock where the metric value is not available is not considered in the ranking.
For example, assume that there are 1800 active stocks on NSE and we are calculating the rank of Reliance Industries P/E ratio. If the P/E ratio is not available for 100 stocks (due to data availability or negative EPS), ranks and scores will be calculated using 1700 stocks.
It is important to note that scores and ranks are relative measures calculated using the cross section of the universe. They are not to be used as absolute measures.
If the market is overvalued, it is quite possible that stocks with high P/E score are overvalued on a standalone basis.