That depends on your calibration strategy. If you are fixing shares there as well, the answer is yes.
All of my parameters (even distribution param) are indeed calibrated from long-run ratios.
Okay, I actually find good results for the production side of the economy, but I have issues on the consumption side. I have defined output from the production approach, meaning output is the sum of labor revenues plus capital revenues plus energy consumption. Summing these three does not give me the same output when I sum consumption, capital investment and public spendings. This is quite normal that my output from the production approach differs from output from the spending approach in data. The statistical discrepancy is about 200 billion euros, which corresponds to “tax less subsidies” that I did not take into account in the output approach. Should I add it? If I do, as an example, my capital factor share will drop capital + labor factor share will not sum up to 1 (less than 1).
National accounts correspondance with the model can be tricky…
Thank you.
Bets regards
That depends. 200 billion is a big number, but relative to the Euro area GDP, it may be negligible.
I’m studying France, so 200 billion on 2’000 billion is quite a lot. But I don’t get how most of papers take output as GDP, and have labor and capital share summing up to 1 … Since GDP is labor comp + capital comp + taxes less subsidies … What would you suggest?
Thank you
[/quote]
I would simply split the taxes less subsidies part according to the capital and labor share.
1 Like
About the BGP, I have computed all my variables for the BGP. However some variables give me BGP values that are far from today’s data, such as the debt to output ratio which is 0.6 (according to the growth and stability European Pact) for my BGP and 0.95 in data. As I reckon, the initial block sets the BGP (or SS) values for which the economy converges. Where can I specify initial values for my system at t=0, where some variables are not at BGP?
Should I specify in the initial block that debt is at 0.6 and “shock” debt to 0.95?
Thanks.
Best regards
Thank you.
[/quote]
No, you set the BGP as an endval and the initial condition in initval
1 Like
Thank you.
Why in your paper with Born, bonds have a different timing convention than capital? As I reckon, both accumulate through time, so both should have to same timing convention, shouldn’t they?
Best regards
Good morning,
why do papers have a different timing convention for bonds and capital? I often see, such as yours with Born, Bonds entering the period at B(t-1) while capital enters the period as K(t), with choices on B(t) and K(t+1).
Best regards
That is just poor and inconsistent notation. Both should be predetermined.
Ok, so I should define every stock variables in my papers such as Dynare’s time convention ?
Ideally, yes. Otherwise, you get a suboptimal mixture of timing conventions and predetermined_variables.
1 Like
For the case of durables, I don’t see why most papers define it as a predetermined variable as when you buy a durable (car, dishwasher etc) let’s say in t, it is effective directly in t and not (t+1).
In my point of view the stock of durables used for durable services in t should be the stock of durables bought in t plus the one bought in (t-1) (minus depreciation), not just the stock of durables of (t-1).
Am I wrong with this ?
Thank you,
It’s a timing convention. The model is discrete. Say you buy two cars over the course of the year. But you still need to decide whether you measure your number of cars for the year on January 1st or December 31st.
1 Like