Forums  > Basics  > What is the daycount convention for commodities?  
     
Page 1 of 1
Display using:  

pj


Total Posts: 3317
Joined: Jun 2004
 
Posted: 2016-10-04 14:39
Is it ACT/360 everywhere?

I saw a dead fish on the pavement and thought 'what did you expect? There's no water 'round here stupid, shoulda stayed where it was wet.'

Patrik
Founding Member

Total Posts: 1333
Joined: Mar 2004
 
Posted: 2016-10-04 14:43
For what situation are you looking at daycount convention?

In general it's not really a concept that comes up much in commodities.

Capital Structure Demolition LLC Radiation

pj


Total Posts: 3317
Joined: Jun 2004
 
Posted: 2016-10-04 14:45

I am staring at the commodity forward curves.

I saw a dead fish on the pavement and thought 'what did you expect? There's no water 'round here stupid, shoulda stayed where it was wet.'

tbretagn


Total Posts: 243
Joined: Oct 2004
 
Posted: 2016-10-04 14:51
Hi pj. Same as Patrik, not sure why that would be relevant for forward curves as this is the price to get a physical delivery.
I'd say use act/365 if anything, as you can get delivery any day? In case your are looking at the cost of warehousing the material.

Et meme si ce n'est pas vrai, il faut croire en l'histoire ancienne

pj


Total Posts: 3317
Joined: Jun 2004
 
Posted: 2016-10-04 15:22
I have to admit my total lack of knowledge on the area.
I am looking at the hard-coded (not by me) stuff.
And there are commodity forward curves which are supposedly
playing the role equivalent to that of the dividends.

I saw a dead fish on the pavement and thought 'what did you expect? There's no water 'round here stupid, shoulda stayed where it was wet.'

Patrik
Founding Member

Total Posts: 1333
Joined: Mar 2004
 
Posted: 2016-10-04 15:23
People don't think of forward curves in terms of relative time, we think in terms of absolute delivery month/time. As we have seasonality etc relative points have no meaning to us in commod.

EDIT: on the dividend part - in practice, no. some academics talk about convenience yield etc, but practitioners don't.

Capital Structure Demolition LLC Radiation

pj


Total Posts: 3317
Joined: Jun 2004
 
Posted: 2016-10-04 15:29
> some academics talk about convenience yield etc, but practitioners don't.
Spot on!
So basically nobody IRL gives a damn. That explains and
assuages my doubts.
Thank you!

I saw a dead fish on the pavement and thought 'what did you expect? There's no water 'round here stupid, shoulda stayed where it was wet.'

deeds


Total Posts: 346
Joined: Dec 2008
 
Posted: 2016-10-04 15:29
@patrik - doesn't cost of carry have to be estimated from prices/forwards for use in contracts with optionality (or for interpolating or extending forward curves themselves)? admit i may have this wrong

Patrik
Founding Member

Total Posts: 1333
Joined: Mar 2004
 
Posted: 2016-10-04 15:36
deeds: for extrapolation (assume unhedgable) you'd more likely make a judgement call what you think is sensible rather than use some implied costs from the front of the curve. for interpolation, sure you could do that, but you could also use any other interpolation method that just cares about the prices. It wont make much difference. It's also worth noting that most contracts would only reference actual futures contracts, or platts/argus publicated prints. So you don't tend to care so much about the tenors in between the published prices.

You may care ABOUT the implied cost of storage for trading point of view though. i.e. if you know that grain elevator cost for storing wheat between time X and Y will cost you Z cents/bushel, and the market is pricing something else - then that's interesting. What I meant is that it doesn't tend to come in as a concept for constructing curves per se (like I imagine in e.g. FI or Eq).

Capital Structure Demolition LLC Radiation

ronin


Total Posts: 201
Joined: May 2006
 
Posted: 2016-10-04 15:51
The reason people use daycount conventions in fixed income is because they trade rates, but settle numbers.

In commodities, you trade numbers and settle numbers. You don't need a daycount convention for that. If you want to convert those numbers to some rates for your own convenience, that is up to you - use what ever convention you need.

Same thing for interpolation and extrapolation. That is just your choice of how you hedge delivery on a day for which there is no standardised contract or no liquidity. Either you hedge with one nearby contract (piecewise constant interpolation), or with two surrounding contracts (piecewise linear interpolation), or something complicated involving more than two contracts (no, me neither).

"People say nothing's impossible, but I do nothing every day" --Winnie The Pooh

pj


Total Posts: 3317
Joined: Jun 2004
 
Posted: 2016-10-04 16:45
I understand now why the day count isn't figuring in the official commodity conventions.
Nobody cares.

In my case the day count was linked with the
interpolation of the convenience yield,
thus it was meant to inherit the day count of the according interest rate curve.
(thank you, deeds, for pointing it out!)

Some smart-aleck simply decided to hard-code day count as ACT360
since it is the default day count of the according interest rate curve.
Default, independently of the currency. Ouch.




I saw a dead fish on the pavement and thought 'what did you expect? There's no water 'round here stupid, shoulda stayed where it was wet.'

deeds


Total Posts: 346
Joined: Dec 2008
 
Posted: 2016-10-04 20:06

thanks, @ronin, @patrik, @pj for clarification

Previous Thread :: Next Thread 
Page 1 of 1