I performed 2 more tests.
Test 1
I wanted to see again the effect of rebalancing alone in comparison with 'AIMing.' This time I chose two fake stocks that were exactly negatively correlated as I did previously, but this time one of the 'stocks' had half the amplitude of the other one.
To create the fake stock with half the amplitude, I started with a 'stock' at 4. The next price in Lichello's series is 5 - for a difference of 1. I halved the difference and added it to 4 - for a price of 4.5. Next price in the Lichello series is 8 - for a difference of 4 from 4. I halved this difference and added it to 4 - for a new price of 6. I repeated this process for 10 - for a new price of 7.
I started the Lichello series at 10 (normal) and started the less volatile fake stock at 4.
Rebalancing alone produced: $246,565.
AIMing both produced: $957,457 + $44,479 = $1,001,936
Test 2
I created two fake stocks. One based on the Lichello series and the other the Lichello series beginning at 8 (that is, skewed backward 1 month). This represents two stocks that are closely positively correlated (the opposite end of the spectrum from the exactly negatively correlated example in my previous post.)
Rebalancing alone produced: $25,150
AIMing both produced: $957,457 + $528,380 = $1,485,837
The conclusion I draw is that rebalancing alone only works when there is an exact negative correlation of equally volatile stocks. Synergy is the next question I want to look at.
By the way, my thought now is that I should also investigate AIMing individual stocks and rebalancing the portfolios rather than rebalancing within a basket and AIMing the basket.
I thought I'd share this food for thought.
Matt