ISG Personal Retirement Income Planner

[mark]

Anyone looked closely at this post retirement income planner?
It is a freebie download at:

http://www.isgplanning.com/Individuals.htm

I just downloaded it and played with it for the last hour, and don't quite understand it yet.

Modern Portfolio Theory provides an allocation of assets based on a client's risk tolerance---which is an appropriate approach prior to retirement. But Modern Portfolio Theory does not provide a specific plan for withdrawing income from a portfolio during retirement. ISG Personal™ fills this gap and provides an allocation of assets that is specifically designed for maintaining steady and dependable income during retirement.

Just playing with it so far, I cannot get close to allowable withdrawls that FireSeeker allows.

[cute fuzzy bunny]

Well for starters the program uses monte carlo simulations, which have already been found in some cases to produce lower returns than actual market action...

See: http://early-retirement.org/cgi-bin/...num=1111612523

Also, when they show "failures" in their presentation, they were using a 6% withdrawal rate.

The program also uses some sort of built in "switching strategy" that sells stocks when they become 'fully valued'. I dont see anyone with a sure-fire method of measuring 'full value', nor a perfect way of 'switching' assets that isnt selling snake oil. Every study I've seen shows that market-timers dont do as well as a good buy-and-hold asset allocation scheme.

Combining 3 or more not particularly valid strategies doesnt cause the flakey ideas to cancel each other out

[cute fuzzy bunny]

Another good calculator besides firecalc is ORP. Very comprehensive.

http://www.i-orp.com/about.html

[mark]

I think the point of their program is that one should keep a fixed income portfolio large enough so that he/she will not have to sell stocks in a down market for income. Thus, the variable "holding period" for stocks.

[cute fuzzy bunny]

Thats one approach, the other is to hold a 2-3 year cash buffer and little or no fixed income. Some folks follow that approach.

I've yet to see a "stock switching" scheme that stood the test of time. Someone that used to be here was hawking such a scheme, only problem is it failed twice in the last 10 years to the tune of missing a 60% increase in stocks the first time and a 30% increase the second time.

I've also yet to see a monte carlo simulation that could be backtested to demonstrate reasonable correlation with actual historic results. About the best I've seen is that they could create worse worst case scenarios. I would imagine that a good monte carlo sim that had some correlations built in to help guide the year-to-year results as are actually played out in real life might be interesting. I heard Raddr made up some monte carlo simulation tools that incorporated some correlative factors...I havent looked at them though.

Pretty easy to factor for worse than worst case scenarios. Do the firecalc historic return, take the worst resulting series and cut that by however worse you expect the worst case to be...

[mark]

Check out the J&L Planner in the recent post. It has random historical simulations.

[cute fuzzy bunny]

If you followed that link I supplied above, it pretty much shows how random historical simulations produce false-low results.

In other words, interesting to look at, but I wouldnt base any decisions on the results.

[mark]

What do you mean by a falsely low result?

[cute fuzzy bunny]

Maybe it would help if you explained what you're trying to accomplish. It appears you're throwing calculators at the wall to see if they'll stick. Theres a million of them. All have their ups and downs. Not one of them will tell you whats going to happen over the next 30 years.

As I said just above your post, if you read the link I supplied way up top, it should explain the 'lower returns' comment.

In brief, its been discovered that taking a historic chain of annual returns and swapping any two years at random produces an end return result that is lower than the original series. In other words, there is something beneficial going on from year to year that creates a correlation that produces better returns than if that chain is broken, albeit in a small way. Maybe its Fed influence. Maybe its investor psychology. Maybe its a bazillion different positive influences by everyone being exerted on the market at the same time.

Monte Carlo strings together series of completely uncorrelated and unrelated data at random. Unless you try to imply some sort of rules to the way the data is constructed. Since nobody (that I'm aware of) knows what exactly makes the actual data series produce better results than randomizing parts of that actual series, I would think making up rules to improve the monte carlo results would not work out so well.

So what you can get out of such a product is a whole range of possible returns that have no correlation with the way the market has worked in the past, nor any certainty towards the way they will work in the future.

In other words, its interesting, but I dont know what the heck you would get out of it.

I suppose as a comparative tool between two different styles of portfolio, it might tell you possible differences in returns. That depends on knowing the long term rates of return on certain asset classes. Which we dont. You can use historical rates of return for asset classes, in which case you would probably be just as well off using the historical data itself rather than making up a random or semi-random data series that may or may not be any more or less predictive of tomorrows actual market action.

Really short answer: every calculator is hooey. Run them all and if they all say your plan is great, then is it great ?

[mark]

Monte simulations are ok, not great. J&L Financial Planner enables the user to change the monte carlo analysis and historical return analysis factors.
I find some calculators useful, but all are lacking to some degree. It's like predicting long term weather.
But lets not digress to software simulations of human induced climate change.

[Kevin]

Just to clarify... The ISG Personal program does not use Monte Carlo analysis. It simply determines how much capital must be allocated to fixed-rate investments in order to avoid having to sell stocks before the entered holding periods. The idea is to create the same situation that most people have prior to retirement, whereby one portion of the portfolio is invested in reasonably guaranteed investments to replace salary, while another portion of the portfolio is invested for the longer-term future. Some people will be comfortable assuming shorter stock holding periods. Others will want a bigger safety net.

We don't advocate waiting until the end of the holding periods before selling stocks any more than we advocate waiting until one's car is on empty before looking for a gas station. The goal is to sell stocks when they are at or above their long-term trendlines rather than being forced to sell at bargain prices.

The goal is not to predict the future, but rather to build enough flexibility into the plan so that one can adjust to it.