FUEGO
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
- Joined
- Nov 13, 2007
- Messages
- 7,746
Portfolio survivability success, that is. I hope I got your attention!
I've been thinking about the firecalc results and how it would be nice to know at some point early in the game whether you are destined for success or failure. That is, at some point when you can still do something about it. I did a little study of a hypothetical 40 year retirement period with a 4% SWR of 60,000 on a 1.5 million portfolio w/ a little social security about 30 years into it. This is loosely based on my own circumstances in what I hope will be ~10 years or so. I'll be late 30's, maybe 40 by the time I can FIRE.
I was searching for a magic rule that would tell me at some point in the first 5 to 15 years whether my portfolio was destined for failure or success.
Apologies to PhD level finance professors, but I'm not a PhD researcher. I just wanted a quick peek at firecalc results to test my hypothesis that there may be a good indicator of future success/failure at some point in time significantly preceding the end date of the portfolio (in this case, 40 years). I followed a basic methodology as follows: Get spreadsheet of year by year inflation adjusted portfolio balances. There were 97 results. For simplicity's sake, let's call it 100 results. There happened to be 10 "failures" in firecalc, giving me a ~90% success rate. Failure is defined as zero or lower portfolio value at any point in the simulation according to firecalc.
I wanted to look at what I call a "soft failure". I would define that as the point at which I would have trouble sleeping at night (because my wife would be kicking me for losing most of our retirement savings). For this analysis, I assumed any time my end of year portfolio balance was 50% or less of the original balance (adjusted for inflation), I experienced a soft failure. For each analysis year, I took a minimum value of all year-end balances and summarized the results at the far right hand side of my spreadsheet (it's attached here). I sorted the analysis periods by these minimum portfolio values.
Then I examined the worst three deciles of analysis periods. The worst decile were the traditional failures where a zero portfolio balance was obtained at some point in the 40 year analysis period. The next to worst decile had a minimum value at some point in the analysis period of between 5% and 39% of the initial portfolio value. The third worst decile had a minimum value at some point in the analysis period of between 40% and 52% of the initial portfolio value. The best 7 deciles are what I would call "successful" - that is, the end of year portfolio value was never less than ~50% of the initial portfolio value.
I pose the following question: is it possible to determine whether the portfolio will fail by looking at the portfolio balance as a percentage of the initial balance at 5, 10, and 15 years from the start?
By looking at the results, I would answer "Yes".
The portfolio balance as a percentage of the initial balance at 5 years wasn't very useful. The 10 and 15 year percentages were increasingly more useful (see table below). The worst decile had an average of 59% and 39% of the initial portfolio remaining after 10 and 15 years, respectively. The 2nd worst decile fared better with an average of 75% and 60% of the initial portfolio remaining after 10 and 15 years, respectively. The third worst decile, consisting of marginal soft failures, fared even better with an average of 91% and 85% of the initial portfolio remaining after 10 and 15 years, respectively. The percentages of the initial balance remaining were fairly variable. The 15 year percentages are the most reliable predictors of portfolio success.
Table of average portfolio balance as a percentage of the initial balance:
......................5 yr 10 yr 15 yr
1st worst decile 89% 59% 39%
2nd worst decile 79 75 60
3rd worst decile 97 91 85
4th worst decile 98 108 112
Setting a rule of thumb for when to go back to work and stop portfolio draws is difficult. Serisouly, who wants to go back to work after 10 or 15 years of chillin' in the sun?!? One would have to balance their own desires and correctly weight the desirability/disutility of false positives and false negatives from a "percent of initial balance" test at 10 and 15 years.
A false positive would be an analysis period that has a 10 or 15 year "percent of initial balance" below the 10 or 15 year threshold that indicates one needs to find other sources of income to supplement portfolio withdrawals (go back to work!). However the analysis period ends up as a successful run.
A false negative would be an analysis period where the 10 or 15 year "percent of initial balance" exceeds the 10 or 15 year threshold (ie - you're headed to success) but ends up falling in the lowest two deciles (either a hard failure or a soft failure).
I'll throw out 60% and 50% as proposed thresholds I might personally use for the 10 and 15 year "percent of initial balance" thresholds, respectively. The 60% threshold at 10 years would catch 6 out of 10 hard failures and 3 out of 10 2nd worst decile soft failures. Only one potential false positive was observed out of the remaining ~80 trials (and that false positive was in the 3rd worst decile, or marginally a soft failure).
The 50% threshold at 15 years would catch 9 out of 10 hard failures (zero balances) and 4 out of 10 soft failures. No false positives were observed.
Application of the 10 or 15 year "percent of initial portfolio" tests would significantly reduce the chances of a catastrophic failure of the portfolio (ie reaching a zero balance) and would moderately reduce the chance of a soft failure (reaching less than 50% of the initial balance).
Here's the link to my firecalc run that generated this data.
(sorry folks, I accidentally deleted the spreadsheet of my analysis!)
I've been thinking about the firecalc results and how it would be nice to know at some point early in the game whether you are destined for success or failure. That is, at some point when you can still do something about it. I did a little study of a hypothetical 40 year retirement period with a 4% SWR of 60,000 on a 1.5 million portfolio w/ a little social security about 30 years into it. This is loosely based on my own circumstances in what I hope will be ~10 years or so. I'll be late 30's, maybe 40 by the time I can FIRE.
I was searching for a magic rule that would tell me at some point in the first 5 to 15 years whether my portfolio was destined for failure or success.
Apologies to PhD level finance professors, but I'm not a PhD researcher. I just wanted a quick peek at firecalc results to test my hypothesis that there may be a good indicator of future success/failure at some point in time significantly preceding the end date of the portfolio (in this case, 40 years). I followed a basic methodology as follows: Get spreadsheet of year by year inflation adjusted portfolio balances. There were 97 results. For simplicity's sake, let's call it 100 results. There happened to be 10 "failures" in firecalc, giving me a ~90% success rate. Failure is defined as zero or lower portfolio value at any point in the simulation according to firecalc.
I wanted to look at what I call a "soft failure". I would define that as the point at which I would have trouble sleeping at night (because my wife would be kicking me for losing most of our retirement savings). For this analysis, I assumed any time my end of year portfolio balance was 50% or less of the original balance (adjusted for inflation), I experienced a soft failure. For each analysis year, I took a minimum value of all year-end balances and summarized the results at the far right hand side of my spreadsheet (it's attached here). I sorted the analysis periods by these minimum portfolio values.
Then I examined the worst three deciles of analysis periods. The worst decile were the traditional failures where a zero portfolio balance was obtained at some point in the 40 year analysis period. The next to worst decile had a minimum value at some point in the analysis period of between 5% and 39% of the initial portfolio value. The third worst decile had a minimum value at some point in the analysis period of between 40% and 52% of the initial portfolio value. The best 7 deciles are what I would call "successful" - that is, the end of year portfolio value was never less than ~50% of the initial portfolio value.
I pose the following question: is it possible to determine whether the portfolio will fail by looking at the portfolio balance as a percentage of the initial balance at 5, 10, and 15 years from the start?
By looking at the results, I would answer "Yes".
The portfolio balance as a percentage of the initial balance at 5 years wasn't very useful. The 10 and 15 year percentages were increasingly more useful (see table below). The worst decile had an average of 59% and 39% of the initial portfolio remaining after 10 and 15 years, respectively. The 2nd worst decile fared better with an average of 75% and 60% of the initial portfolio remaining after 10 and 15 years, respectively. The third worst decile, consisting of marginal soft failures, fared even better with an average of 91% and 85% of the initial portfolio remaining after 10 and 15 years, respectively. The percentages of the initial balance remaining were fairly variable. The 15 year percentages are the most reliable predictors of portfolio success.
Table of average portfolio balance as a percentage of the initial balance:
......................5 yr 10 yr 15 yr
1st worst decile 89% 59% 39%
2nd worst decile 79 75 60
3rd worst decile 97 91 85
4th worst decile 98 108 112
Setting a rule of thumb for when to go back to work and stop portfolio draws is difficult. Serisouly, who wants to go back to work after 10 or 15 years of chillin' in the sun?!? One would have to balance their own desires and correctly weight the desirability/disutility of false positives and false negatives from a "percent of initial balance" test at 10 and 15 years.
A false positive would be an analysis period that has a 10 or 15 year "percent of initial balance" below the 10 or 15 year threshold that indicates one needs to find other sources of income to supplement portfolio withdrawals (go back to work!). However the analysis period ends up as a successful run.
A false negative would be an analysis period where the 10 or 15 year "percent of initial balance" exceeds the 10 or 15 year threshold (ie - you're headed to success) but ends up falling in the lowest two deciles (either a hard failure or a soft failure).
I'll throw out 60% and 50% as proposed thresholds I might personally use for the 10 and 15 year "percent of initial balance" thresholds, respectively. The 60% threshold at 10 years would catch 6 out of 10 hard failures and 3 out of 10 2nd worst decile soft failures. Only one potential false positive was observed out of the remaining ~80 trials (and that false positive was in the 3rd worst decile, or marginally a soft failure).
The 50% threshold at 15 years would catch 9 out of 10 hard failures (zero balances) and 4 out of 10 soft failures. No false positives were observed.
Application of the 10 or 15 year "percent of initial portfolio" tests would significantly reduce the chances of a catastrophic failure of the portfolio (ie reaching a zero balance) and would moderately reduce the chance of a soft failure (reaching less than 50% of the initial balance).
Here's the link to my firecalc run that generated this data.
(sorry folks, I accidentally deleted the spreadsheet of my analysis!)
Last edited: