How to calculate a 6-month backward looking moving average matrix?

I have monthly data and would like to calculate a 6-month backward moving average. I'm looking at movavg function. I tried the following:

[Short, Long] = movavg(returns_list, 0, 6);

I got this error "Lead and lag arguments must be positive <= 588."

Should I use "[Short, Long] = movavg(returns_list, 1, 6);" instead?

Thanks, J

Answer 1

[Image Analyst]

I don't have whatever toolbox has movagv() in it, but I think you could use convolution:

filteredSignal = conv(signal, [0 0 0 0 0 1 1 1 1 1 1]/6, 'valid')

Remember, convolution flips the kernel so that's why you want to have the 1's on the right side of the array to get the average of the current and 5 prior values (assuming each element is one month).

Answer 2

[Julio]

Thank you.. I actually figured out like this

[Short, Long] = movavg(return_list, 6, 6);

[Andrei Bobrov]

As by IA

filteredSignal = conv(signal, [0 0 0 0 0 1 1 1 1 1 1]/6, 'same')

[Julio]

Will this work if I have missing data? Especially, since it will try to average always by 6...

[Image Analyst]

We don't know what your data are. Months have different numbers of days in them. If you have missing days or missing months, you should first run interp1 on it to fill them in. Conv() has options 'same', 'valid', and 'full' depending on how you want to handle the "edge effects" where your window reaches the beginning and end of your data where you don't have the full 6 months of data.