The original paper was in 2023, but they revised it a week ago.
Unless they drastically changed the methodology, I believe it is highly flawed still. Big ERNâs article does a good job of illustrating it. An exerpt:
1 Like
I also had some questions about the block bootstrap method, but intuitively, I feel it is better than sampling independent data points or sampling from a distribution which doesnât reflect the patterns within the data.
I havenât seen any strong academic challenge to the block bootstrap method. Thereâs one which notes potential bias and ways to mitigate:
have you seen this post?
Its not necessarily a challenge but comes out with different answer to the same question.
1 Like
Probably because it uses different data. How many scenarios does that website cover?
I think many simulations either do:
- Look at historical time periods, which limit the data available and have only a fixed return pattern matching history
- Use randomized returns matching an idealized return distribution. It gives an âaverageâ answer but doesnât account for the patterns that happen in real life
- The block bootstrap method tries to address both. It takes many historical periods and combines them in randomized blocks to generate more scenarios than are available in method 1, but also contain long enough blocks to reveal patterns in time that you donât get with #2.
What i can see is following in terms of method. My personal view is that there is a lot of research around this topic, but retirement portfolios are extremely difficult decisions because you do not have much time to correct (as you are not working anymore).
Similar to my Portfolio Finder method of exploring asset allocation possibilities, I calculated every possible combination of those 7 assets (repeats allowed) using 10% intervals. Thatâs 8008 portfolios ranging from simple options like 100% domestic stocks to more complex mixes like 60% domestic stocks, 20% foreign bonds, 10% bills, and 10% gold.
On the scenario side, the countries represent the 10 largest developed markets as defined by most index funds. These established economic contemporaries cover 4 continents, 6 currencies, and some particularly unique events in places like Japan and Spain. By looking at 40 retirement start years since 1970 in 10 different countries, the analysis studies 400 unique retirement scenarios.
Put it all together, and my new withdrawal rate database contains not only 8x the number of scenarios of the typical long-term retirement study but also 10x the country diversity and many thousands more portfolio options. Thatâs a lot of new data to work with.
For each of the 8008 possible portfolio options, the analysis looks at all 400 historically accurate scenarios around the world and calculates the withdrawal rate that depleted the portfolio in 30 years. Note that it includes my withdrawal rate projection method that allows more recent start dates to also be studied even when theyâre not quite to the 30-year mark. Thatâs 3.2 million combinations of portfolios and scenarios, and yes, I could use a break from spreadsheets.
The minimum number across start dates for a portfolio in a single country is what Bengen referred to as the SAFEMAX withdrawal rate, or simply the safe withdrawal rate or SWR for short. The minimum number across both start dates and countries for each portfolio is what I call its global withdrawal rate. Itâs the worst case scenario not just in the United States but in the larger global record.
3 Likes
I think argument #7 is valid, and maybe you could adapt the simulation for this by conditioning the selection of the next block based on valuations to get a more ârealisticâ price evolution.
I am just glad that I am not retiring soon 
This is very tough decision.
My plan is to build a portfolio I can stick from now until forever so that I donât even need to make this decision when time comes.
I think it is a challenge to get comfortable financially retiring now as you have quite high valuations, we are seemingly at the top of a long bull market and to top it all off, we have the Trump administration making massively disruptive changes to the global trading system.
Imagine retiring now and having a worldwide trade war happen right after retirement.
Thatâs similar to what I am doing in a way because Iâve been filling out my Pillar 2 with voluntary payments the last year and it has a low fixed return profile.
Iâm not sure one can define the optimal asset allocation depending on age.
The last check should be for the undertaker and it should bounce!
Or something like that. But fortunately we donât know.
I think all depends on risk tolerance. The rule of age in bonds came from my banker, the same guy that told me not to buy my house because it had too much land and ânobody pays the landâ. Whenever a banker tells me something I was much better off doing the exact reverse!
I have a high risk tolerance and 45 years of experience in the stock market. I have 2 nice pieces of self-used real estate that give me quiet some comfort and insurance. So I invest all I have (even a bit more on credit) in stocks, what historically gives the best return but also the highest volatility.
I will not change this with age. Even if my brain is probably not as good as a few years ago. I have written down every detail of my mechanical strategies and yes, sometimes I have to check those writings when I donât know what exactly to do. I will do that as long as I can, hopefully until the day I die.
1 Like
Yeah I actually changed my plan into this 
Uh oh. I hope no Cortana effect!
1 Like
| % | |
|---|---|
| Gold | 5 |
| CHF | 5 |
| Energy stocks | 10 |
| Mining stocks | 5 |
| US REIT | 5 |
| General stocks | 60 |
| CH REITs DRPF | 10 |
| 100 |
So. I was thinking again about asset allocation upon retirement. And I tried to sketch out something and came out with the above.
The idea was to have the bulk of assets in stocks (60%) but then have some invested in particular assets/sectors and then periodically re-balance.
1 Like
What bonds actually are you buying? I would guess only chf denominations as swiss investor? Gov bonds are near 0 and corp bonds dont seem very appealing either atm?
If you account for taxes on dividends and interest rate risk, how attractive are these really?
I feel the swiss speciality of no capital gain tax is weightibg strong on the plus side for equities?
A post was merged into an existing topic: SPHQ vs JQUA vs?
Iâm with you there. All about risk tolerance - specially psychologically. I feel if one can stomach the volatility (stay invested in downturns), does not need the cash anytime soon for a large purchase; a very high equity allocation can be quite reasonable for any phase in life, even through retirement.