I am surprised to see that portfolio did well than any of its members.
The author clearly says that table is equally weighted assets i.e simply divide your money into 7 and buy 7 assets.
I am still baffled, after 40 years, the portfolio shows a positive return but most of portfolio members are in negative !?
I can see why portfolio can be positive if some of the assets were shorted but author seems to be long on all assets!
Am I missing something? How is this possible?
Perhaps author rebalanced the portfolio every month when rolling his futures contracts to next month in this 40 year period?
I wish author had a returns table for his portfolio along with 7 commodities, could some one help with a sample table (say 5 years with 3 assets) to understand this phenomenon?
Last edited by rocktrader; May 18th, 2015 at 01:25 AM.
Got my hands dirty and tried monthly rebalancing of silver:wheat:coco = 8:3:1
Using monthly returns for the period: Apr 1985 to Apr 2015
Portfolio returns: 186.7 %
Silver returns: 153.02 %
Wheat returns: 50.71 %
Coco returns: 28.00 %
My take away is:
With proper weights we can have a portfolio dominating any of its members.
In author's case it seems equally weighing the assets happened to work well and portfolio dominated the members. In the book he further goes on properly weighing the portfolio members and taking some short positions too, to further optimize the returns which are even better than the table I posted.
From the yearly returns of 30 years stats for silver, wheat coco respectively:
mean: 7.623 SD: 35.183
mean: 5.244 SD: 29.951
mean: 2.743 SD: 19.751
mean: 5.522 SD: 23.978
SD (Standard Deviation) of portfolio is lesser than silver and wheat, providing the magic required to beat the individual top performers over time! This is inline with authors explanation of table I posted. Author goes further explaining how optimally risking from the point of view of Kelly investor mindset made this achievement possible.
I think I have grasped what author is saying (although his table has some gaps in my understanding, I have constructed another equivalent table to validate/understand his ideas!).
Last edited by rocktrader; May 21st, 2015 at 05:03 AM.
To keep the calculations simple, I simply divided the money at start of month into 12 parts and allocated 8 parts to silver at that month's silver price and so on, it will not be proper lot size etc... but for self demonstration close enough is good enough. Yeah, you are right author shows portfolio outperforming by just rebalancing since there is no table of prices, I was just not able to see how this can work. If rebalancing can work with 8:3:1 well there is no surprise it can work 1:1:1 for some portfolios... that's my take away in a roundabout way.