A.I.M. Users Bulletin Board

[ls7550]

Another noteworthy aspect is that AIM tends to be the least worst. If for instance you start with a regular 50/50 portfolio and draw income from whichever of the two is the higher value at the time, otherwise left as is, then that is a form of partial rebalancing; Or you might rebalance back to 50/50 after withdrawing income (full rebalancing). Across all such sample runs those alternative styles yield different overall outcomes, one or the other isn't the consistently better approach. The differences can be significant, such as ending with 120% of the inflation adjusted start date portfolio value instead of 80%. To reduce the risk of being in the worst case choice you might start with 50/50 allocations to each of those, in effect run two separate portfolios for 30 years, same assets, just different styles, each providing half the income. Running AIM solely alone however and the results tend to be closer to the better of the two choices - is inclined to avoid being the worst case.

Fundamentally rebalancing is good if the prior trend reverses, would have been better to leave as-is rather than have rebalanced if the trend continues. Sometimes non rebalanced is better than rebalanced and in other cases vice-versa. AIM has a knack of being a bit of both, magically identifying the more appropriate choice of the two. Robert Lichello called it Automatic Investment Management, a more appropriate name in the present era however might be a Artificial Intelligence Machine. From a 30 year SWR risk measure perspective AIM has served well, inclined to be better than aimlessly :) picking and sticking with a rigid asset allocation/style.

From the Trinity study guidance withdrawal rates (SWR's) for different lengths of time



AIM is inclined to yield 100% when others might yield 95% probabilities, so as your horizon shortens so AIM might better achieve acceptable higher rates of withdrawal rates, perhaps such as 6% SWR with a 15 year horizon.