Friday, June 16, 2017

10 Year graph of Eggs

This is for a comment I made on Seeking Alpha.

The only consistent very long term series I could find for US eggs is the PPI. So this is an index with 1993 = 100. The graph has a good fit though.
x-axis - PPI for eggs deflated for inflation and value of the $US.
y-axis - change in deflated price in next ten years.
so the last point on the graph is on May 2006.
The red line is where we are in May 2016.


Monday, April 17, 2017

dygraph of Rhodium

I'm trying out the dygraph package in R. Blogger apparently doesn't allow you to upload a webpage (if I'm wrong about that, please call or write). But I did publish it to RPubs. It's at:

http://rpubs.com/brothberg/268672

Dygraph brings all the functionality of javascript timeseries graphing to R. If you select a portion of the graph while holding down your mouse, the graph will render that section only. Double click to get back to the original. You can also do neat things like shade recessions and annotate important dates.

I did one for my rhodium position, with a few reasonably important annotations. Pretty cool.



Friday, March 24, 2017

Rhodium is going crazy!

I've been bullish on rhodium for a while and wrote an article about it in SA. It really is performing well, up almost 30% since Jan 1 and 4% in the last two days.
I really don't know what is going on short term, but I still like the long term. It is only mined as a byproduct of platinum and palladium. The miners of those metals are in trouble, so we will likely see production down. Meanwhile, demand remains good from a wide variety of industries. Here's a chart of the price:
Still, even now it's really cheap on a long run basis:

Monday, February 27, 2017

More Rain in the Cocoa Zone

So the cocoa zone in west Africa got another dose of rain a few days ago. This is pretty unusual for this time of year, and augurs for a heavy mid-crop in Ghana and the Ivory Coast. Bearish!

Having said this, I still don't see a big downside here. The end-users, of which I used to be one, realize the value here are are willing to extend coverage. I also think that we are near an inflection point in grindings.
I'm looking for another spot to buy again.

Wednesday, February 15, 2017

The Bottom in Cocoa?

I think it is. It's always hard to catch falling knives, but in this case I think the risk/reward is there.

- Cocoa is cheap on a long run basis. The forward 10-year graphs that I run have it about doubling in the next 10 years. That's not a basis for a trade, but it's a good start. Note: I no longer publish these.

- We are beginning to get some bullish info out of Africa. The rain event in early Feb was unusual and quite bearish. Since then however, dryness has continued. Also, Reuters has an article about how cocoa is being left on the trees to rot (overly dramatic, but with a grain of truth). The official Ghana cocoa agency reduced its 2016/17 crop estimate by 50K tonnes. I doubt that this is accurate, but it shows that the producers are starting to do things to move prices up.

- Commitments of traders is VERY one sided. There have been very few times in the past ten years where the specs have gone short cocoa; all have ended with bull markets. We are at record short now.

I believe the last leg down was due to defaults by some medium size cocoa middlemen. I have no knowledge of, but wouldn't discount the possibility of, a western trading company being in trouble as well. This new selling added to the decline. This is probably over, or nearly so.

I bought a half position via options in CCK two days ago. If we can get some further upside action, I will double up.


Wednesday, January 18, 2017

Trades Today

I took off about 30% of my Impala Platinum. It's had a nice run. If it comes off I'll buy it back, if not - too bad.
Going to add some Cameco (CCJ). I'm still bullish on uranium, but it's going to be a long slow climb. So only buy the dips.

Someone commented in the last post about the correlation between platinum and gold. I simply used the correlation of the daily returns of GLD and PPLT. In fact, if you strip out the overall commodity correlation, the true relationship is even lower.