How to View Weighted Average Returns over time.

KiwiFI

Dryer sheet wannabe
Joined
Nov 28, 2024
Messages
23
Location
Australia
Seeking some views and advice on this.

Like many, I've run dozens of calculations in FireCalc and other calculators around my numbers and the projections. I like the "portfolio with random performance option" in FireCalc and setting a mean return with standard deviation and inflation rate. Note, I don't live in the US so the default market options aren't all that useful for me, even if I do have some exposure to US equities.

I plan to do a Geo-arbitrage and relocate to a LCOL area from a HCOL part of Aussie - reducing living expenses considerably.

My portfolio tracker tools give me really accurate data about average returns year on year. I don't know why I only started tracking with these tools 2-3 years ago as they're just the best thing! I have a broad mix in my portfolio across a number of asset classes, which looks a like a 70/30 split currently. 30 cash and 70, Equities, precious metals, property funds, & some alternative investments.

Running weighted average return calculations across these I get a pretty healthy figure over the last 5 years of 18%.

Obviously, 18% pa compounds very nicely over time, but what I'm concerned about is 5 years worth of data enough to project this out 10,15 or 20 years? I'm 12 years away from a pension (social security), which is actually means-tested in Australia.

As the saying goes, the long run is really just a series of short runs you need to navigate well.

The upshot is that I would have a withdrawal rate closer to 6% but FireCalc reports this isn't an issue, given the inputs. I'm pretty conservative on the whole so what do people think about that rate and what should I be looking out for or reassessing here? I'm mid 50s, no dependants. There will be modest inheritances at a point in time, but I tend to discount this in calcs.

One thing that lurks in my mind a bit is what if country relocation doesn't pan out long term and I move back to a M/HCol area in my 60s or 70s. I guess the answer is an obvious one and something to think about and / or plan for.

Thanks.
 
Do I understand you correctly that you are inputting (into Firecalc) a portfolio with random performance, but you are using the mean return of 18% (based on your last 5 years of experience)?

If I am wrong about that, what value are you using?
 
Do I understand you correctly that you are inputting (into Firecalc) a portfolio with random performance, but you are using the mean return of 18% (based on your last 5 years of experience)?

If I am wrong about that, what value are you using?
Effectively, yes. Though I'm projecting out 20 and 25 years also and running those calcs too.

I play around with a standard dev number between 5-10%.

I'm not sure if I'm using the calc the right way to be honest.

I expect the way to think about it is the old adage of past performance isn't a predictor....
 
Well, it just seems very optimistic to think that you can expect a mean return of 18% in the long run with ANY portfolio. You had a good run recently, but I am skeptical that those numbers will continue.
 
I use about 1.75% ROR on my forecasts. My actual ROR is ~5% annual net of spending since I retired (4 years ago). I use the lower number to forecast. As I get older, I may use higher numbers as I determine what I might want to spend before I die. I am 65yo.
 
The answer to the OPs question is an emphatic, No, extrapolating from 5 years of data is not prudent.

Kudos for being willing to do and think about monte-carlo simulation, but as with all models, so much relies on the assumptions made for the inputs.

I think it is useful to think in terms of years of funded retirement. Compute (retirement assets) / (annual expenses) and you'll get the years of funded retirement with no inflation and zero investment growth. Then the question begins to center on the actual problem, how to grow assets to outpace inflation for the length of retirement with a high degree of certainty.

Using historical returns like fire calc does (pretty much the entire recorded financial history) helps to avoid wishful thinking from short term patterns. But it still assumes that past performance is indicative of future results. In the end some assumption about the future is necessary.
 
The answer to the OPs question is an emphatic, No, extrapolating from 5 years of data is not prudent.

Kudos for being willing to do and think about monte-carlo simulation, but as with all models, so much relies on the assumptions made for the inputs.

I think it is useful to think in terms of years of funded retirement. Compute (retirement assets) / (annual expenses) and you'll get the years of funded retirement with no inflation and zero investment growth. Then the question begins to center on the actual problem, how to grow assets to outpace inflation for the length of retirement with a high degree of certainty.

Using historical returns like fire calc does (pretty much the entire recorded financial history) helps to avoid wishful thinking from short term patterns. But it still assumes that past performance is indicative of future results. In the end some assumption about the future is necessary.
Thanks. Point well taken re using the actual historical returns. And yes, there are so many assumptions that can change over time & course correction is needed.
 
I use the US Stock Market's average annual return, after inflation. That's about 7%. 18% is way too high, and IMHO, is unlikely to continue for 25 years. Or 5.
 
IMHO, it is extremely foolish to think that anyone can predict what the market will do at any time for any period in the future. You just happened to have started tracking in one of the best periods of market growth for two years. How about we use the closing value of the S&P 500 on last trading day before Christmas 2021 through 2023 instead. Looks like a gain of 29 points (0.6%). Project that for the next 20 years and see how you do. In my case, I retired as soon as I could and crossed my fingers. Yes, it was around Christmas 2021 and I quickly lost 30% of my net worth. I spent 25 years studying charts, reading financial newsletters, etc and in the end it meant absolutely nothing.
 
^^^^^^^^^

Yeah I retired not too long before the unpleasantness of '08. Kinda scary, but I survived and thrived. I just no longer think in terms of 2 or 3 years as a planning period.
 
^^^^^^^^^

Yeah I retired not too long before the unpleasantness of '08. Kinda scary, but I survived and thrived. I just no longer think in terms of 2 or 3 years as a planning period.
I actually didn't sweat it too bad. Up double from end of 2022 so I guess you just hang in there and everything will be fine.
 
I haven't followed all the responses in the thread but I have couple of comments:
* Returns of 5 years is no where close to a reasonable data set. You need a data set over decades for the securities you have invested in before you can perform any meaningful simulations (which is what FireCalc does).
* An remember, with any calculator/spreadsheet, garbage in = garbage out

Sorry, I don't have any helpful suggestions for your case.
 
IMHO, it is extremely foolish to think that anyone can predict what the market will do at any time for any period in the future. You just happened to have started tracking in one of the best periods of market growth for two years. How about we use the closing value of the S&P 500 on last trading day before Christmas 2021 through 2023 instead. Looks like a gain of 29 points (0.6%). Project that for the next 20 years and see how you do. In my case, I retired as soon as I could and crossed my fingers. Yes, it was around Christmas 2021 and I quickly lost 30% of my net worth. I spent 25 years studying charts, reading financial newsletters, etc and in the end it meant absolutely nothing.
Useful insight, thanks.

As mentioned in the Psychology of Money, if a 30% retreat in net worth would cause sleepless nights, one might not be ready. Sage advice, generally speaking, I sense.
 
Back
Top Bottom