I am thinking of rebalancing my investment portfolio and I've become
increasingly interested in evaluating life as a range of possibilities
(distribution) instead of a fixed number (average).

I'm not a professional investor and I've yet to find a professional
for guidance that seems to think in these terms - so I'm thinking of
tinkering around with it myself. I don't want to be active, I'd like
to buy and hold some strategy for a fairly long period of time as a
long term investment.

A popular view is: assume an investment of 10K on X date, what would
the terminal value be on Y date.

My concerns with this are:
1. What is the right X date?
2. What is the right Y date?
3. Why do I have any reason to believe that the future will be like
the span between those dates?


I'm thinking of running a series of simulations that will create a
distribution of potential gains and then I can compare the
distributions across the different strategies.

Example simulations:

1. Evaluate the performance over a 5 year horizon. Pick a rolling
start date between X and Y for each 5 year interval and calculate the
returns. (e.g. 1/1/2010->1/1/2015, 1/2/2010->1/2/2015,
1/3/2000->1/3/2015).

2. Repeat for each 5 year span possible (e.g. 1/1/2000->1/1/2005,
1/1/2010->1/1/2015... and all in between )

3. Repeat for 3, 5, 10, 15 year spans

4. Instead of shifting the dates, sample with replacement from the
spans for each of the above (e.g. between 1/1/2010 and 1/1/2015, the
dates could be 1/1/2010, 1/2/2010, 1/1/2010)

If these different simulations were plotted on a histogram, I am
thinking something might pop out.

Questions:
1. Has anyone done anything like this in J?
2. Are there any investors (armchair or professional) that would care
to opine on the usefulness of this analysis?

Thanks,
Joe
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to