Calculating RSI 2 in excel

Okay so here's a post on RSI on request left by an anonymous user in comments to my post on RSI calculation. He was facing an issue with the calculation of RSI 2.

First What is RSI 2 ? In normal RSI (Wilder's RSI) we use data from past 14 days to calculate RSI. In RSI 2 we use just the data for past 2 days to calculate RSI.

So now the issue is with the computation of RSI 2 in Excel. The Average gain or avg loss often turns out to be zero as there could be two continuous days of gains or losses. The formula for RS is avg gain/avg loss. So in case avg loss is zero it so happens that then this value is not computable, or rather infinite.

What does one do in that case? Cause excel cannot do division by zero or infinite has on meaning for it. In reality if one looks at the formula for RSI = 100- 100/(1+RS). So the larger the value of RS, larger is the value for RSI. For value of RS = infinite, the value for RSI is 100.

So to be able to do the same calculation in excel what we will do is simply add a if statement before the formula for RS, such that if avg loss were equal to zero, RS would be a very high number, which in turn would mean that RSI would be 100 in all such cases. This should solve our problem. Look at the excel sheet to understand the concept better.Leave a comment if you have any doubts or issues


Dow Jones RSI Calculation in Excel

I have done the RSI calculation for Dow Jones in excel so that any anyone can see and learn from it how to calculate RSI in excel for any stock or indice.

You can download the excel spreadsheet from here

Nifty RSI Calculation in excel


Here's an example of RSI calculation done in excel spreadsheet with real life data of Nifty index. You can download that from here -



In fact you can understand from the excel sheet how exactly to compute RSI and then you can apply the formulas to data of any stock or indices. Remember each value of RSI takes in an unlimited number of previous values in consideration, therefore it makes sense to have atleast 50 data points before your calculation date so that the values you get are more or less accurate.