How do i predict future price by applying moving average?

Hello Everyone,I have a house price data with two rows, 1st: Date (monthly) and 2nd: Price of house, my data has the price record of houses from 1990-2012, I would like to predict the price of house in 2013 by calculating the moving average of the prices i have to forecast and predict the 2013 house price.Please any idea on how to carry out simple time series analysis (step-by-step)in MATLAB? I would appreciate sample codes.

>> Q = ones (1,3)./3; 
>> G = conv2(c1,Q,'same').';  
>> Moving_average=conv2(c1,Q,'same').';

I tried finding the moving average of my price row (c1) but the results i am getting is invalid, Any suggestions please? This is exactly what i am trying to replicate in matlab. Thanks!

How To... Calculate Simple Moving Averages in Excel 2010

Answer 1

[Roger Wohlwend]

I always use the function filter to calculate a moving average of length N.

N = 3;
MA = filter(ones(N,1),N,Y);

However you cannot really forecast a time series with a simple moving average. Try an ARMA model or a linear regression model instead.

[Ahmad]

Thanks, I made use of loop to calculate the moving average: for i=4:253 myhousedata(i,4) = mean(myhousedata([i-2 i-2 i-1],3)); please any idea on how to get the predicted next month's price from there? I am looking to find a price value.

[Ali]

Hi month ,i am facing the same issue.Have you predicted the fututre price/forecast?

Answer 2

[Image Analyst]

What I would do is to take the most recent data and assume some polynomial model, like a line or quadratic. Then fit that most recent data to the model. Then extrapolate into the future. But I'm not really sure how the financial analysts do it. Perhaps there's something in the Financial Toolbox. I often see ranges where there is an upper limit, best guess, and lower limit when they do the stock price estimates into the future.