Serving the Quantitative Finance Community

 
TheCorpFinanceQuant
Topic Author
Posts: 4
Joined: May 16th, 2019, 8:52 pm

Pricing commodities via indiicators

September 12th, 2019, 7:03 am

Hi All,

I am interested in gaining what I believe to be the 'fair' price of a commodity. For example, if the price of the top 5 correlated assets all, but my commodity stays the same, I would like to think that my commodity is undervalued and the fair price is higher than represented. However, what other inidcators for commodities could be used? 
 
User avatar
tags
Posts: 3159
Joined: February 21st, 2010, 12:58 pm

Re: Pricing commodities via indiicators

September 15th, 2019, 8:59 pm

("if the price of the top 5 correlated assets all go higher" ?)

I'm not sure of what kind of replies you're expecting, but from a practioner's standpoint, one angle to assess the fair value of a commodity on its fundamentals. For instance at an oil major you would try to get some sense of the fair price a the crude oil barrel in terms of days of cover (that is the number of consumption allowed by reserves) that you will try to relate to flat price/forward structure and at an ABCD you would use a grains stock-to-use measure (simple stock / total demand ratio).    

Issue is that commodity fair value (1) changes all the time for countless reasons (e.g. price inflation in the long-run), (2) it is difficult to trade with and (3) the fundamentals of a commodity are only one component of its flat price.

About (3), on top the fundamentals of a commodity, one has to assess a risk-premium e.g. maybe the oil industry has to price the risk of drone disruption a petroleum industrial site, and corn traders have to consider weather market (price premium related to weather) during the growing season.

Also, commodity people look closely to CFTC's COT reports. For instance, you may want to remove trend and seasonality of the net position of managed money for a given commodity and add this as an input to your price model.

But yeah... it is a difficult exercise and no one gets it right, e.g. look backward to past crude oil price forecast of any IB and you will realize we were not there at all. Anyways, the good news is that "fair value" is somewhat a flawed concept, so you're never really wrong.

Maybe you can have a look at H.Geman's paper on soybean forward curve and volatility, even if it is not exactly what you're looking for. Her paper is very interesting but notice that it is useless in practice. Moreover, World soybean market has totally changed with the rise of Brazil and Argentina as top suppliers in recent years, probably invalidating some of her reasonings (I haven't read of paper for years now, so apologies I also may be wrong....).

[in italics above are terms in the parlance of the respective industries]