This example helps to illustrate the difference between short term and long term, taken from Stata.com:

Suppose that we compare housing markets between two cities. If the housing markets become too dissimilar in the short run, people and businesses will migrate, bringing

the average housing prices back toward each other. We therefore expect the series of average housing in one city to be cointegrated with the series of average housing prices in the other city, in the long run.

It depends on particular data. When there is some cointegration relationship, it is there usually because there is some force pushing two time series together. What is "long run" depends how "strong that force is". For example, when you have one stock traded at two different exchanges, prices will be very similar. If they were not, somebody would buy stock at one exchange and sell it in another exchange. Therefore, in this case "arbitrage force" will work very quickly, for example within seconds. However, in most other cases in economics, similar arbitrage is not possible. Therefore, in some other cases, even 10 years might not be enough for 2 time series to come together.

Dear Naveed Ahamad, 

Usually the short and long-run referes to periods in days, where short-run covers the 1 to 30 days and long run covers the more than 30 days.

Dear Naveed Ahamed,

If you have monthly data for ten years & suggest that you have variable B as dependent and Variable A as independent,

1.       Example for the short run -  The impact of first month of A on the twelfth month of B: Short term impact: You could run Granger Causality Test in order to estimate the same. Here you could estimate a feedback also, for ex :  A’s impact on B & B’s impact on A

2.      Example for the Long Run : The impact of first month of A on the 48th month (4th  year) of B : Long Term impact: You could run Vector Auto Regression or Hong’s Model (2001) in order to estimate the same.

Thanks a lot, Jayantha Dewasiri Narayanage , Naveed Raza,Peter Molnár, Peter Molnár 

 Hello,

Naveed Ahamed Lone

I think ,

the short run means the all factors of the economy can not be changed in the particular period in the meantime long rnu means all factors can be changed in the period. More explanation please see this attachment   

Dear Mr Dewasiri,

How did you indicate that the short-run is measured on 12 months and the long-run is measured on 4 years? I have a semi-annually data for 16 years, so the long run is also measured on 4 years?

Thank you in advance!

Dear Naveed, suppose you estimate an econometrics model based on quarterly data, every period (quarter) refer to short run relationship. If you use VECM for example, and the the error correction term (ECT) =0.35, the long run relationship will be three quarters , and so on for ARDL and VAR..... etc, in general short run concept refer to one year at most, and long run concept refer to more than one year.

Dear Mr Obeid,

Suppose the ECM model:

Δ𝐹𝐷𝑡=𝑐− (1−𝛼) 𝐸𝐶𝑇𝑡−1 + 𝛽1 Δ𝐹𝐷𝐼𝑡 + 𝛽2 Δ𝑅𝐺𝐷𝑃𝐶𝑡 + 𝛽4 Δ𝐾𝑡 + 𝜀𝑡

In this equation, ECT should be between 0 and 1 or not ?

(I know that (1−𝛼) should be between 0 and 1 to be significant).

When you say ECT=0.35 you mean 1−𝛼= 0.35 or ECT itself?

Thank you in advance!

What is the meaning of Short-run and long run  (Time period)

This is a fascinating issue; yet, based on the comments, it appears that the substance of long- and short-run estimation, as well as the policy implications, has been overlooked.

The short run is a period of time in which the quantity of at least one input is fixed and the quantities of the other inputs can be varied. The long run is a period of time in which the quantities of all inputs can be varied