RSI 14

Jolam0524 · #1 (**permalink**) 05-04-2009, 07:43 AM

Anyone could help to illustrate how RSI in the MarketScope be obtained?

FrAnton · #2 (**permalink**) 07-26-2009, 01:05 PM

Quote:

Originally Posted by Jolam0524

Anyone could help to illustrate how RSI in the MarketScope be obtained?

I tried to calculate RSI 14 manually taking open/close values from FXCM trading station marketscope 2.0 chart, and could not get result to be even close.

I have used the standard formula

RSI = 100 - (100 / (1 + RS)); where
RS = Average Gain / Average Loss for 14 periods.

So I also wonder what formula is in use to calculate this Indicator?

Nikolay.Gekht · #3 (**permalink**) 07-26-2009, 10:00 PM

RSI is implemented according Perry Kaufman "Trading System and Methods" chapter 6 "Momentum and Oscillators".

The pseudo code:

Code:

sumup := 0
sumdown : = 0
for i = 1 to n begin
    if close[i + 1] - close[i] > 0 then
        sumup := sumup + close[i + 1] - close[i];
        else sumdown := sumdown + close[i] - close[i + 1];
    end;
end;
RSI = 100 - (100 / (1 + sumup / sumdown));

FrAnton · #4 (**permalink**) 07-26-2009, 11:46 PM

Quote:

Originally Posted by Nikolay.Gekht

RSI is implemented according Perry Kaufman "Trading System and Methods" chapter 6 "Momentum and Oscillators".

The pseudo code:

Code:

sumup := 0
sumdown : = 0
for i = 1 to n begin
    if close[i + 1] - close[i] > 0 then
        sumup := sumup + close[i + 1] - close[i];
        else sumdown := sumdown + close[i] - close[i + 1];
    end;
end;
RSI = 100 - (100 / (1 + sumup / sumdown));

thank you for the sample. I guess this is just for the first period.
For the next periods it seems that it calculates Avg as:

diff = source[period] - source[period - 1];
if (diff > 0) then
sump = diff;
else
sumn = -diff;
end
positive = (pos[period - 1] * (n - 1) + sump) / n;
negative = (neg[period - 1] * (n - 1) + sumn) / n;

This is what I found in RSI body. Is that right?

Nikolay.Gekht · #5 (**permalink**) 07-27-2009, 02:08 PM

Exactly so! pos and neg are hidden streams which contains the previous values for the data.