At 1.8% withdrawal rate if you are risk averse you may want to consider just laddering out your annual payments using TIPS. This would allow you to spend more without worry of failure. You can afford to drive the failure rate to zero, while having more to spend if that is your wish.Thank you!! Right now I'm at around 1.8% withdrawal rate, so I guess I'm goodThat's a 1% failure rate for a 50 year retirement with a 3.2% withdrawal rate while waiting 20 year for social security. 3.2% withdrawal rate is not crazy if you study other withdrawal studies for more than 30 years or if you look at international SWR results. Nobody can tell you the actual future failure rate so it's more about how making changes affects your failure rates.Thank you so much. This went over my head, tbh, but I do appreciate your explanation and I got the bottom line, which is that for now I can stick to what i have. I'm a little concerned about your statement of a 1% failure rate at 31.1 times expenses, even though I don't spend that much.I am using my own Monte Carlo simulations using the VTSAX and VBTLX data from the Simba Spreadsheet. The above results were for a 50 year retirement sequence with social security kicking in at year 20 and accounting for 50% of income for years 20-50.
This is super-helpful. Thank you! I'm surprised to see any failure rate at 50-52x expenses. What simulation are you using? I've mostly used firecalc.
1. pick a year at random
2. have an 85% chance of grabbing the next year in the sequence
3. repeat until we have to pick a new random starting year
4. do this 1,000,000 times so the simulation results are consistent if repeated
5. repeat for each AA (100/0 to 0/0)
6. repeat for each withdrawal rate (2.00%, 2.01%, ... 9.00%) and compute the success rate
Now I can investigate the relative safety of any AA or portfolio size choice. The 85% chance means that I grab 5 year sequences on average but there will be also be 1, 2, and 10+ year sequences. The results are like history but they allow for variations along with a a slim chance of near-market failure or a Japan-like run up.
You could think of the 0.01% failure rate as hyperinflation risk for the 0/100 AA and deep market risk for a 100/0 AA. For 0.01% failure, the 100/0 AA needs 75.2x expenses, the 55/45 needs 50.3x expenses, and the 0/100 portfolio needs 73.0x expenses. For large portfolios, risks are balanced around 55/45 AA where you have equal risk of hyperinflation or market decline. If we want no risk we need a ladder of TIPS big enough that we can't outlive it.
History clusters at 4% being safe but that is not the case if we get a different series of events.
The good news for you is that a 2.5% spend rate drops this to 0.1% failure and a 2% spend rate drops it to 0.01% failure. The important result is that your portfolio risk is defined by how much you choose to spend and not significantly by your AA choice. I'd be very confident with 1% failure if I had the ability to cut spending by 30%. That turns a 3.2% withdrawal into 2.24% withdrawal rate and will still be very safe.. It will probably go up later, so this is all really good to know.
If you want to leave behind a bunch of money or spend a bunch more in future years then 1.8% spend rate is the way to go.
Statistics: Posted by abc132 — Sat Apr 06, 2024 1:46 pm — Replies 90 — Views 14816