Mixing
![We say that the complement of a strip aâ¤xâ¤b is connected in the extended complex plane (or Riemann Sphere) and disconnected in a complex plane. Why? [closed]](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhZCTB4-bb-s-32SAm6ytjzFOU06cH9o8q-a_wq_UnH6lcd8hvws23CmBWHM6g256LzTX9yK1EiCCzTC1NSgtxlNk_J9hdr9bA3tD0Nqntbl521j4bLAm_uXPANuWLBhcCKTKzns2gaUodx/s72-c/1.jpg)
Handling outliers in estimated forward rates
Clash Royale CLAN TAG#URR8PPP
up vote
0
down vote
favorite
I am estimating forward interest rates. Historically i observe that the daily change in interest rate is less than 0.05% at 90 percentile.
If my estimated rate for a given day is 8.50% and previous day is 8.00% can i revise the estimated rate to 8.05% considering the maximum daily change is 0.05%?
Any thoughts please?
statistics regression estimation
asked Jul 16 at 13:07
onkar
1
add a comment |Â
up vote
0
down vote
favorite
I am estimating forward interest rates. Historically i observe that the daily change in interest rate is less than 0.05% at 90 percentile.
If my estimated rate for a given day is 8.50% and previous day is 8.00% can i revise the estimated rate to 8.05% considering the maximum daily change is 0.05%?
Any thoughts please?
statistics regression estimation
asked Jul 16 at 13:07
onkar
1
This is not at all clear. Obviously, old estimates of future values have to get updated as new information arises. Is that all you are asking?
– lulu
Jul 16 at 13:43
It might indicate that your model is bad; i.e. it might have overfit and is throwing out very unreasonable answers. However, if you truly think that the max daily change is 0.05%, and your model is spitting out 0.5%, and you still believe that your model is accurate, then obviously you should trade on this information.
– Kevin Li
Jul 16 at 13:54
1
Voting to close the question as it is unclear what you are asking. If you can, please edit your post to provide more detail. What model are you using to make the estimate? As a general note, if you find yourself in the habit of overruling your own model because the results it gives feel badly wrong to you, that's a sure sign that you need a better model.
– lulu
Jul 16 at 14:07
In a situation like this when you’re making predictions of the next day with information on what happened the previous day, you should generally 1) have the previous day value (and longer lags) in your model. 2) Have compared your model’s performance to simply using yesterday’s value to predict today. (And simple smoothing methods.) If your model makes an extreme prediction, you should focus on understanding why. So you can either find how to improve the model or become comfortable with “believing†the prediction.
– spaceisdarkgreen
Jul 16 at 14:43
add a comment |Â
up vote
0
down vote
favorite
up vote
0
down vote
favorite
I am estimating forward interest rates. Historically i observe that the daily change in interest rate is less than 0.05% at 90 percentile.
If my estimated rate for a given day is 8.50% and previous day is 8.00% can i revise the estimated rate to 8.05% considering the maximum daily change is 0.05%?
Any thoughts please?
statistics regression estimation
asked Jul 16 at 13:07
onkar
1
I am estimating forward interest rates. Historically i observe that the daily change in interest rate is less than 0.05% at 90 percentile.
If my estimated rate for a given day is 8.50% and previous day is 8.00% can i revise the estimated rate to 8.05% considering the maximum daily change is 0.05%?
Any thoughts please?
statistics regression estimation
asked Jul 16 at 13:07
onkar
1
asked Jul 16 at 13:07
onkar
1
asked Jul 16 at 13:07
onkar
1
asked Jul 16 at 13:07
onkar
1
1
This is not at all clear. Obviously, old estimates of future values have to get updated as new information arises. Is that all you are asking?
– lulu
Jul 16 at 13:43
It might indicate that your model is bad; i.e. it might have overfit and is throwing out very unreasonable answers. However, if you truly think that the max daily change is 0.05%, and your model is spitting out 0.5%, and you still believe that your model is accurate, then obviously you should trade on this information.
– Kevin Li
Jul 16 at 13:54
1
Voting to close the question as it is unclear what you are asking. If you can, please edit your post to provide more detail. What model are you using to make the estimate? As a general note, if you find yourself in the habit of overruling your own model because the results it gives feel badly wrong to you, that's a sure sign that you need a better model.
– lulu
Jul 16 at 14:07
In a situation like this when you’re making predictions of the next day with information on what happened the previous day, you should generally 1) have the previous day value (and longer lags) in your model. 2) Have compared your model’s performance to simply using yesterday’s value to predict today. (And simple smoothing methods.) If your model makes an extreme prediction, you should focus on understanding why. So you can either find how to improve the model or become comfortable with “believing†the prediction.
– spaceisdarkgreen
Jul 16 at 14:43
add a comment |Â
This is not at all clear. Obviously, old estimates of future values have to get updated as new information arises. Is that all you are asking?
– lulu
Jul 16 at 13:43
It might indicate that your model is bad; i.e. it might have overfit and is throwing out very unreasonable answers. However, if you truly think that the max daily change is 0.05%, and your model is spitting out 0.5%, and you still believe that your model is accurate, then obviously you should trade on this information.
– Kevin Li
Jul 16 at 13:54
1
Voting to close the question as it is unclear what you are asking. If you can, please edit your post to provide more detail. What model are you using to make the estimate? As a general note, if you find yourself in the habit of overruling your own model because the results it gives feel badly wrong to you, that's a sure sign that you need a better model.
– lulu
Jul 16 at 14:07
In a situation like this when you’re making predictions of the next day with information on what happened the previous day, you should generally 1) have the previous day value (and longer lags) in your model. 2) Have compared your model’s performance to simply using yesterday’s value to predict today. (And simple smoothing methods.) If your model makes an extreme prediction, you should focus on understanding why. So you can either find how to improve the model or become comfortable with “believing†the prediction.
– spaceisdarkgreen
Jul 16 at 14:43
This is not at all clear. Obviously, old estimates of future values have to get updated as new information arises. Is that all you are asking?
– lulu
Jul 16 at 13:43
This is not at all clear. Obviously, old estimates of future values have to get updated as new information arises. Is that all you are asking?
– lulu
Jul 16 at 13:43
It might indicate that your model is bad; i.e. it might have overfit and is throwing out very unreasonable answers. However, if you truly think that the max daily change is 0.05%, and your model is spitting out 0.5%, and you still believe that your model is accurate, then obviously you should trade on this information.
– Kevin Li
Jul 16 at 13:54
It might indicate that your model is bad; i.e. it might have overfit and is throwing out very unreasonable answers. However, if you truly think that the max daily change is 0.05%, and your model is spitting out 0.5%, and you still believe that your model is accurate, then obviously you should trade on this information.
– Kevin Li
Jul 16 at 13:54
1
1
Voting to close the question as it is unclear what you are asking. If you can, please edit your post to provide more detail. What model are you using to make the estimate? As a general note, if you find yourself in the habit of overruling your own model because the results it gives feel badly wrong to you, that's a sure sign that you need a better model.
– lulu
Jul 16 at 14:07
Voting to close the question as it is unclear what you are asking. If you can, please edit your post to provide more detail. What model are you using to make the estimate? As a general note, if you find yourself in the habit of overruling your own model because the results it gives feel badly wrong to you, that's a sure sign that you need a better model.
– lulu
Jul 16 at 14:07
In a situation like this when you’re making predictions of the next day with information on what happened the previous day, you should generally 1) have the previous day value (and longer lags) in your model. 2) Have compared your model’s performance to simply using yesterday’s value to predict today. (And simple smoothing methods.) If your model makes an extreme prediction, you should focus on understanding why. So you can either find how to improve the model or become comfortable with “believing†the prediction.
– spaceisdarkgreen
Jul 16 at 14:43
In a situation like this when you’re making predictions of the next day with information on what happened the previous day, you should generally 1) have the previous day value (and longer lags) in your model. 2) Have compared your model’s performance to simply using yesterday’s value to predict today. (And simple smoothing methods.) If your model makes an extreme prediction, you should focus on understanding why. So you can either find how to improve the model or become comfortable with “believing†the prediction.
– spaceisdarkgreen
Jul 16 at 14:43
add a comment |Â
active
oldest
votes
active
oldest
votes
active
oldest
votes
active
oldest
votes
active
oldest
votes
Â
draft saved
draft discarded
Â
draft saved
draft discarded
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2853390%2fhandling-outliers-in-estimated-forward-rates%23new-answer', 'question_page');
);
Post as a guest
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
This is not at all clear. Obviously, old estimates of future values have to get updated as new information arises. Is that all you are asking?
– lulu
Jul 16 at 13:43
It might indicate that your model is bad; i.e. it might have overfit and is throwing out very unreasonable answers. However, if you truly think that the max daily change is 0.05%, and your model is spitting out 0.5%, and you still believe that your model is accurate, then obviously you should trade on this information.
– Kevin Li
Jul 16 at 13:54
1
Voting to close the question as it is unclear what you are asking. If you can, please edit your post to provide more detail. What model are you using to make the estimate? As a general note, if you find yourself in the habit of overruling your own model because the results it gives feel badly wrong to you, that's a sure sign that you need a better model.
– lulu
Jul 16 at 14:07
In a situation like this when you’re making predictions of the next day with information on what happened the previous day, you should generally 1) have the previous day value (and longer lags) in your model. 2) Have compared your model’s performance to simply using yesterday’s value to predict today. (And simple smoothing methods.) If your model makes an extreme prediction, you should focus on understanding why. So you can either find how to improve the model or become comfortable with “believing†the prediction.
– spaceisdarkgreen
Jul 16 at 14:43