I want to understand better how different asset allocations in US vs ex-US equities affect the success of portfolios in retirement. Most FIRE calculators/simulators that I've seen only use US equities, I assume because the original Trinity study only used US equities. However, this simulator: https://www.portfoliovisualizer.com/monte-carlo-simulation does have data for ex-US going back to 1986 (not as far back as I had hoped, but I'll take what I can get).
If you spend much time in investment forums you'll hear various percentages recommended for ex-US equities. 0%, 20%, and 38-40% are the most commonly recommended (see, for example: https://www.bogleheads.org/forum/viewtopic.php?t=409214 )
So I plugged these percentages into the Monte Carlo simulator to see what has the least chance of failure. If we assume our hypothetical investor is withdrawing 4% and adjusts for inflation, with a simulation period of 30 years, and NO bonds or any other investments (keeping this solely focused on the question of equity allocation), we get some interesting results:
100% US, 0% ex-US: 89% success rate
80% US, 20% ex-US: 93% success rate
62% US, 38% ex-US: 90% success rate
Earlier this year, I made a gut decision during a time of fear about the current US administration, and reallocated my equities to a higher weighting in international, at 40%. Based on these simulations above, I'm seriously considering reallocating back down to 20%. There are many factors to consider in such a decision, both economic and geopolitical, and I'm still reading and thinking, but at the very least I can have some comfort in knowing that even if I remain at the market-weighted asset allocation (38ish percent international), it will do no worse than if it had been 100% US all along. (Which is a different claim than doing better than 100% US!)
International diversification in Monte Carlo simulations
byu/bemusedly infinancialindependence
Posted by bemusedly
2 Comments
>I made a gut decision during a time of fear about the current US administration, and reallocated my equities to a higher weighting in international, at 40%. Based on these simulations above, I’m seriously considering reallocating back down to 20%
I think you are putting far too much stock in your analysis. It’s possible you may be subconsciously looking for a way to justify another ‘gut feeling’ from the other direction. Perhaps you should consider moving much of your stock portfolio to a fund like Vanguard’s Total World ETF (Ticker: VT) so you don’t have to fret about and/or second-guess your US v. Int’l allocation. That fund is currently ~63% US and 37% Int’l.
> but at the very least I can have some comfort in knowing that even if I remain at the market-weighted asset allocation (38ish percent international), it will do no worse than if it had been 100% US all along.
A 30-year retirement simulation that goes back 39 years (to 1986) is far too short to accurately estimate chance of failure with different US / international balances. It sounds like you searched the main Boglehead forum for US/international balance discussion. You can also find many such discussions on this sub. Prior to 2025 ~20% international was the most common selection in user surveys on this sub. After international performed well this year during the start of 2025, increasing to market cap (VT) seems to be most common among sub members. It sounds like you made this switch as well. Nobody knows for certain which market will perform best in the future, and one can make arguments for/against many allocations.