One of the interesting side benefits of my old engineering life was the rare opportunity to see parts of the world many people may read about but never experience first hand. From dim February days in Stockholm to colorful night markets in Taipei, the world is so much bigger than the place you’re sitting right now.
If you only judge the world by what you experience in your normal daily routine, that myopic paradigm is probably going to hold you back at some point. We are all at the center of our own experience, but the universe does not revolve around us. The world is a big place, and sometimes it requires exploring its far reaches to truly appreciate our tiny home in the larger global stage.
I have been thinking about this a lot recently, such as my last article talking about the difference in investing outcomes in various home countries. That exercise reminded me that it might be a good time to update my old Global Withdrawal Rates research with new site charting capabilities.
As often happens when I start tinkering with new tools, the creative process opened up exciting new possibilities that weren’t previously practical but now are finally within reach. And the result this time ended up being a significant enough improvement that it justified an entirely new chart!
I call it the Global Explorer, and it’s a uniquely powerful tool for learning about asset allocation from a truly global perspective.
The Global Explorer is about way more than just ranking famous portfolios. It explores a wide world of options, tens of thousands of portfolios at a time, that your own close circle of financial friends may not have ever considered. So if you have ever wondered what types of portfolios are best suited for certain goals like accumulation, retirement, and risk mitigation — not just in the US or your own home country, but everywhere — then pull up a chair and allow me to share my full travel guide to universal portfolio construction.
How the Global Explorer Works
The Global Explorer is like a mashup between the Optimizer wide-net approach and classic Withdrawal Rates concepts regarding worst-case scenarios.
Like the Optimizer, instead of looking at just one portfolio at a time, it studies every possible combination of assets in 10% intervals. With 9 assets, that’s 43,758 different portfolios under consideration.
Borrowing the idea from withdrawal rate studies of identifying worst-case scenarios for safe strategy building, the Global Explorer also calculates the numbers since 1970 for each of those 43,758 portfolios across 12 different home countries. That’s just over 525,000 iterations for every metric. Then it identifies the worst-case scenario for each portfolio across all 12 countries and ranks every portfolio by that number. The end result is a full list of tens of thousands portfolio options ranked by the best outcomes even in the very worst global case on record.
If that process sounds familiar, it’s the same methodology I used in the Global Withdrawal Rates study I put together a few years ago. The biggest differences here are the asset options and portfolio metrics which take that analysis to the next level. We’ll get to that in just a second, but for now here’s a fully working embed. Take a moment to poke around to get a feel for the scope.
You can immediately see one big difference between this chart and every other chart on the site — there is no asset allocation interface. Instead, there is simply a collection of asset and home country checkbox options. They are all enabled by default, which includes every possible asset and country in the resulting calculations.
Another unique thing about the Global Explorer is the scope of the chart. The main visual is a browsing tool that displays 100 portfolios at a time. Every column is an individual portfolio with the assets color coded to the above asset option checkboxes. It is sorted from best portfolio on the left to the worst on the right, and you can even browse all 43k+ portfolios at your leisure.
Click on any one column, and it will display portfolio details at the bottom. The donut chart on the left visualizes the asset allocation in the same style as other portfolios on the site. You can hover over the slices to see a description of each asset, and it shares the overall rank within the larger set. The column chart to the right displays the individual results for that one portfolio across all countries, which reveals where the worst-case number came from. And if you click on the “Find a specific portfolio” button, it will open a menu that allows you to quickly find any portfolio you want in the database.
I present the main chart in this way rather than simply highlighting the single best portfolio because honestly I find the spectrum of results even more interesting than any one asset allocation. Visualizing what all the best portfolios have in common teaches us something deeper than simply chasing a solitary ideal case that may change spots from year to year. Hover over the main chart, and the tooltip will help to identify each asset color to help you see the patterns in the bigger picture.
That’s a LOT of calculations feeding a seemingly simple chart. Pretty cool, right? Data visualization FTW.
Unique Asset Considerations
Designing a tool that studies a dozen different country-specific portfolio interpretations at once also requires a slightly different mental framework than the rest of the site tools. That’s why there are generic “Domestic” stocks, bonds, and bills on the list. When you see a domestic asset in a portfolio, the calculations assume that the market for that asset changes for each home country. So domestic stocks uses US assets for the US and Japanese assets for Japan. You get the idea. The domestic category allows the algorithm to automatically switch domestic contexts for any portfolio.
Everything not labeled domestic is international by definition, and the assets can be more specific. So every asset is clearly labeled and does not switch at all based on the home country.
That non-switching of international context is actually a nice upgrade from my old global withdrawal rates research. One of the limitations of that work was that “international” was defined as the developed world for most countries, and world ex-US for the United States. That introduced one methodology inconsistency and one blind spot.
The inconsistency was in what “foreign” means to US investors and everyone else. Because the numbers for American investors represented a different set of foreign returns, it potentially skewed the domestic vs international conclusions. And the blind spot was the idea for all non-US investors that a global cap-weighted fund is as diversified as it sounds. Since the historical average percentage of the US in a developed world fund is about 50% (sometimes quite a bit more), “global” numbers arguably say more about investing in the US than about truly broadly diversifying market exposure.
The Global Explorer tool solves both problems by separating US and ex-US stocks and bonds.
- Allowing US stocks in both the domestic and international slots lets the domestic/international calculations seamlessly work with no conflict and just minor interpretation required by the reader. If you see a portfolio with 10% domestic stocks and 30% US stocks, a UK investor can interpret that as a 10/30 UK/US split while a US investor can think of it as 40% US. The results are transparent and the math still works the same.
- Including non-overlapping US and ex-US slices also allows us to study exactly what type of international diversification works best. If it’s truly the global market weight, then the resulting portfolio will show about 50/50 US/ex-US. But removing the constraint opens up the possibility of 100% US or 100% everything else, which can be quite educational.
For investors using this tool outside of the US, I also understand that true ex-US funds are quite uncommon outside of North America. Just think of it as “every developed country but the US”. An EAFE fund would qualify. Mixing a Europe and Japan fund would get you very close. And if you want to keep it simple, a Europe fund will already cover most of the countries and market cap in an ex-US index. So think of the spirit of the portfolio that you can build around rather than the literal asset names.
The History of Global Withdrawal Rates
With that bit of technical and interpretive background out of the way, I thought it would be fun to demonstrate the types of unique research tasks it’s capable of. One particularly fun use case is replicating a variety of different famous retirement study methodologies.
The OG paper about safe withdrawal rates was Bill Bengen’s “Determining Withdrawal Rates using Historical Data”. Bengen studied the historical data for US stocks and bonds and simulated retirements in the US using varying percentages of each asset. His goal was to identify the withdrawal rate that safely sustained a portfolio for 30 years even in the worst case on record. Here are the Global Explorer settings that best represent the methodology.
Original Bengen Settings
That’s two assets and one home country. There are 11 possible combinations of those two assets, and the one with the best 30-year safe withdrawal rate was a 50/50 split of stocks and bonds. See that 4.1% SWR? Bengen’s number was 4.2%, which is well within the normal margin of error for different data sources for the same asset. That’s where the famous 4% retirement rule comes from.
Before we move on, check out the worst two portfolio options — 100% stocks and 100% bonds. That alone is a nice lesson in the importance of asset diversification.
One of the first researchers to think about broadening one’s perspective beyond the United States was Wade Pfau. Instead of looking at only stocks and bonds in the US, Pfau expanded the countries under consideration and also added domestic bills to the list of options. Here’s what that looks like.
Pfau Settings
There are now 66 possible portfolios with 3 asset options. With 9 more countries under consideration for worst case scenarios, you can see how not only did the overall 30-year safe withdrawal rate change, but the ideal portfolio composition did as well. Look at the column chart at the bottom right, and it lists the SWR numbers by country for that one portfolio. That variance of data for each country is what Pfau focused on most. Let the range sink in, and you’ll start to appreciate the issue with only looking at data from the United States.
While we’re here, also take a moment to look at the block image of all the portfolios together. Each color represents a different asset, and you can clearly see a pattern in the portfolios. Generally speaking, when looking at these 3 asset options from a global perspective the more bills the higher the SWR. Cash is a valuable asset in the worst cases.
Stocks also exhibit a similar inverse pattern even though there is some interesting noise where portfolios with similar SWRs can have very different percentages of stocks. Keep that in mind if you have been taught to evaluate portfolio returns primarily on the stock percentage.
Another popular retirement study that has made the rounds recently is a paper called Beyond the Status Quo: A Critical Assessment of Lifestyle Investment Advice by Aizhan Anarkulova, Scott Cederburg, and Michael O’Doherty. I’ll refer to it as the Cederburg paper simply because he has been the face of it in various interviews. The Cederburg paper made waves by looking at the returns from a large number of countries, including an international stock option, and concluding that not only do bonds have no place in a retirement portfolio but also that the true global SWR was in the 2s. While Cederburg has some unique methodologies like dumping all country returns into the same randomization hopper, one can simulate the spirit of the study using these settings.
Cederburg Settings
Cederburg fans may note that his paper used cap-weighted global stocks instead of US stocks. Because a global cap-weighted fund is primarily driven by US returns, the closest asset in the Global Explorer is actually not a broad international ex-US fund but a simple USA stock fund. Note that while my 2.84% SWR is still a bit higher than Cederburg’s in the 2.2s (he looked at way more worst case scenarios like defaulting Greece and Germany recovering from WWII), the Global Explorer closely replicates his ideal all-stock portfolio recommendation with 35% domestic stocks and 65% foreign. So we’re on the same page.
Check out the shape of the green color trend in the chart and compare it to the previous Pfau chart. Very different, right? That’s what made the Cederburg conclusions about bonds so provocative.
Now look at the opposite trends of the two red colors (domestic vs. US stocks) in how they affect withdrawal rates. Higher percentages of US stocks helps, while higher percentages of domestic stocks hurts. Clearly the specific stock types matter.
This is a good opportunity to point out the utility of the checkboxes. If you study the numbers by country for long enough, you may notice that Spain in the early 1970s shows up often in the worst case scenario. If you’re like me and have a few basic common-sense investing rules like “I choose to not invest my money in the local stocks and bonds of a country with a crumbling dictatorship” and “high-quality foreign bonds are OK, too”, then here’s what the same data looks like excluding Spain and including US bonds.
Alternate Cederburg Settings
Excluding Spain and including foreign bonds
Tweak the settings just a little, and you get a notably higher withdrawal rate along with a very different portfolio composition. Importantly, it’s not that the Cederburg research is inaccurate. Just like how Pfau and Cederburg expanded the possibilities from the original Bengen research, there’s just more to the story.
Contributing to that progress is why I have built many retirement calculation tools over the years including things that focus on Withdrawal Rates, Financial Independence and Retirement Spending. The most recent was my spin on the multi-country worst-case approach that I called Global Withdrawal Rates.
The Global Explorer not only folds that calculation system into a fancy new chart layout, but it also levels up the portfolio options with two more assets. Instead of just using straightforward cap-weighted global stock and bond funds for “foreign” options, it separates international investing into two non-overlapping buckets — US, and ex-US. That not only increases the total portfolio options from 8008 to 43,758, but it also allows us to finally study what the ideal breakdown of international investing looks like that may diverge from normal market weights.
Global Explorer Settings
Expand the options to a nicely comprehensive range of assets for modern investors, and retirement portfolio research starts to get really interesting. That top international stock allocation really was 100% US after all. And it may be hard to believe, but worst-case global retirement portfolios that lasted even in Franco Spain with a SWR higher than 4% were actually possible! You just need the right tools to find them.
That search process is what the Global Explorer is all about.
Early Exploration Tips
Truly understanding the lessons of the Global Explorer takes time, and to be honest I’m just getting started with the new data as well. But to give you a head start, here are a few early noteworthy takeaways that I think you’ll find interesting.
First, I know a lot of you look at this chart and immediately cringe at the dark blue gold allocations dominating the top portfolios. Gold is a uniquely difficult pill for a lot or people to swallow, but it’s easy to exclude. Just uncheck that one box and see what happens.
Beyond the effect on the top SWR number (gold is just that helpful in the worst retirement scenarios), check out the colors in the chart and how the general portfolio composition of the top portfolios changes.
First, ex-US bonds become much more important in picking up the role that gold previously filled. And second, ex-US stocks actually become more important than US stocks. The root causes for those two things are open for discussion, but my educated guess is that they both tie back to currency. If you’re not depending on gold as your diversifier for currency risk, then things like broadly diversified stock and bond funds become more important. Keep that in mind as you build your favorite retirement portfolio without gold.
Next, let’s go back a previous point about Spain. Here is how simply removing Spain from the consideration list affects the top retirement portfolios.
In this case, removing Spain bumps the global SWR a little but doesn’t have a major effect on the top asset allocation on the surface. But look at the medium-dark green domestic bonds. Spain sets so many low points for domestic bonds because if its own unique historical situation that removing that one country makes them look way more appealing than before. Depending on how conservative you want to be, maybe that doesn’t change your mind. But I think understanding the influence of unique situations is an important part of balanced historical research.
Finally, let’s check out the opposite end of the spectrum and check out the absolutely worst portfolio combinations.
Remember how all of the top retirement portfolios contained a big slice of gold? The worst possible thing you can do as a retiree is invest all your money in gold. So don’t go overboard! Like salt in your food or chlorine in your drinking water, the dosage matters.
But c’mon now. Don’t stop there. The second worst portfolio was 100% commodities. 100% domestic stocks is also right there at the bottom. Scroll over one page and 100% domestic bonds is also in the 1st percentile. Even famously productive 100% US stocks is ranked #42,889 in the lowly 3rd percentile.
The lesson here is that the problem with 100% gold is much more than a story about the value of a shiny rock. When planning for a safe retirement, investing all your money in almost any single asset class is likely to be a poor choice. And not only does diversification reduce risk, but by doing so it also drastically improves returns for things like safe withdrawal rates. It’s highly counter-intuitive, but the right combinations of assets can generate results measurably better than the weighted average of each asset in isolation.
Smart diversification creates portfolios greater than the simple sum of its parts. That’s the beauty of portfolio theory in practice. And the real-world result beyond the cold academic theory is a more prosperous retirement for you and your family.
Exploring New Metrics
Retirement is a topic near to my heart, but the reason that I named this tool the Global Explorer rather than simply sticking with Global Withdrawal Rates is that it is capable of so much more than retirement research. Why not apply that same worst-case global scenario mentality to portfolio growth and risk mitigation as well?
The last interface option that we have not covered yet is that bright red Explore By dropdown button. The Global 30-Year SWR is just one of three options.
Global 30-Year SWR
The minimum 30-year safe withdrawal rate looking at every country. The 30-year SWR is a standard measure for retirement portfolios.
Global 15-Year Baseline CAGR
The minimum 15th percentile 15-year rolling real CAGR looking at the compound returns for every country. The baseline return is my favorite measure for a conservative long-term return to plan around without trying to predict the future.
Global Ulcer Index
The maximum ulcer index looking at every country. The ulcer index measures the overall pain of holding a portfolio by looking at the depth, length, and frequency of all drawdowns. It is my favorite measure of the emotional side of investing risk.
In all three metrics, the Global qualifier signals that it’s not just looking at one country at a time, but expanding the concept to find the worst-case global scenario. Withdrawal rates are not the only measure that can benefit from that framing, as finding portfolios that also grew and protected your money even in the worst cases can be extremely educational.
Flip between the exploration options, and the first thing you’ll learn is that your primary goal may change your ideal asset allocation.
Portfolios designed to maximize withdrawal rates tend to be highly balanced between all three major asset classes. When making your money last, coving your bases is key.
Portfolios designed to maximize growth even in the less-than-ideal situations tend to reduce bonds but still have a lot more real assets than many of you probably anticipated. Sustainable portfolio growth clearly requires more than fixating on the expected returns of individual assets.
And portfolios designed to minimize pain lean heavily on cash while still mixing in highly volatile assets like stocks and gold in greater amounts than you might naturally think. Sometimes the safest plan isn’t to avoid volatility, but to properly balance the useful volatility of one asset with another.
So while the mindset behind withdrawal rates was a major inspiration for this tool, please don’t get the wrong idea that the Global Explorer is just another retirement calculator. It’s way more than that, and even has room to grow.
Embrace the Explorer Mindset
The Global Explorer is one of those tools that admittedly is so information-dense that I can understand if it doesn’t immediately sink in with everyone. I get it. Like offering someone a ship equipped with the finest new compass and pointing them towards the open ocean, not everyone is as eager as Magellan to sail over the horizon to undocumented shores.
But if you’re like me and find that explorers mindset invigorating, the ability to quickly sort more than half a million datapoints at a time for each metric is pretty exciting. There’s a ton of research and technical work that went into making it possible, and I’m just happy to have the venue to openly share the results.
Before you dive in head-first, I think it’s important to point out that one quality that the best explorers all share is a healthy sense of self-preservation. Nobody will remember your journey if it is cut short by impulsive and reckless decisions. So as with any optimization tool, immediately falling in love with the “best” option often isn’t the wisest choice. Sometimes it’s best to watch with open eyes and build a solid plan by studying the big picture.
That mental image of surveying the landscape actually played a part in an important design decision. Instead of sticking with a cloud of datapoints like the Optimizer that focuses on numbers, I purposefully created the Global Explorer to help you think about asset allocation first. And not just one portfolio, either. But a full horizon of possibilities.
Study the image like an abstract painting of a landscape. Look for patterns and what they imply about what a robust portfolio design looks like. And pause on the realization that every single one of these top-100 portfolio options is an objectively great choice. Like 99th percentile excellent that any investor even in the worst case global scenario would have enjoyed. In fact, we’re talking about so many portfolios that you can scroll many pages of options and still not go wrong.
Truly wise investing isn’t about constantly chasing the one top option on the far left. The best investors embrace the variability in the brush strokes and focus on the vision. If you were to paint your own portfolio on the canvas for your goal, would it blend right in or stand out like a sore thumb?
If you can’t pick your portfolio out from the noise, you’re probably already on the right track regardless of the specific percentages for every asset. Don’t sweat the small details. And if it draws attention to itself, just think of it as a sign that your current portfolio maybe doesn’t match your desired performance landscape. You don’t necessarily have to change it all at once, but carefully adjusting the color balance can get you back on track.
So as you play with the Global Explorer, I encourage you to approach it from a true explorer’s mindset. Think big, keep an open mind, and focus in the big picture. Look beyond a single allocation, beyond assets you may normally consider, and even beyond one country at a time.
There’s a huge new sea of data waiting.
Join the conversation
