Financially reviewed by Patrick Flood, CFA.
The All Weather Portfolio is a well-diversified, low-risk portfolio from Ray Dalio designed to “weather” any environment. Here we'll look at the All Weather Portfolio's components, historical performance, ETFs to use in 2024, and various leveraged strategies.
Interested in more Lazy Portfolios? See the full list here.
Disclosure: Some of the links on this page are referral links. At no additional cost to you, if you choose to make a purchase or sign up for a service after clicking through those links, I may receive a small commission. This allows me to continue producing high-quality, ad-free content on this site and pays for the occasional cup of coffee. I have first-hand experience with every product or service I recommend, and I recommend them because I genuinely believe they are useful, not because of the commission I get if you decide to purchase through my links. Read more here.
Contents
Video
Prefer video? Watch it below. But note the video is just a general overview of the “base” portfolio and does not include the discussions on leverage or replacing Commodities with Utilities, so you'll have to scroll down further for those.
What Is the All Weather Portfolio and Who Is Ray Dalio?
First, we'll take a brief look at what this portfolio is comprised of, and why. The All Weather Portfolio is an available-to-the-masses portfolio modeled somewhat after the risk-parity-based All Weather Fund from the famous hedge fund Bridgewater Associates. The portfolio idea was created by the legendary Ray Dalio, founder of Bridgewater, and was then popularized by Tony Robbins. Dalio has become almost like a god in the world of finance and investing, and rightfully so. I would highly recommend his bestselling book Principles, as well as his more recent book Big Debt Crises. His newest book from November, 2021 is Principles for Dealing with the Changing World Order: Why Nations Succeed and Fail.
Bridgewater is perhaps most famous for their focus on and analysis of different economic cycles in “the economic machine.” Their All Weather Fund – and subsequently the All Weather Portfolio – is designed to survive all economic environments, using different types of assets that perform differently during those different “seasons.” Appropriately, it is also sometimes referred to as the “All Seasons Portfolio.”
While Bridgewater is constantly watching the market and the economy, Dalio himself admits that he can’t predict the future, thus the need for a portfolio that mitigates the financial impact of unexpected economic events, known as black swan events. The All Weather Portfolio becomes especially attractive during periods of market turmoil, particularly for investors who have a low risk tolerance and/or are primarily concerned with capital preservation.
It has long been known that portfolio diversification mitigates risk and volatility. The All Weather Portfolio seemingly maximizes diversification using a variety of asset classes. This diversification benefit is borne of the inherent uncorrelation of these assets, e.g. when stocks go down, bonds tend to go up. This conveniently obviates the need for any attempt at or anxiety over market timing or guesswork. Dalio actually explicitly maintains that attempting to time the market is a fool's errand.
Note that the All Weather Portfolio as it is prescribed is not based on true risk parity of the assets, but it is pretty close. It is simply the product of an interview between Tony Robbins and Ray Dalio in which Dalio suggested that these weightings, without leverage, would be suitable and easy to manage for the average investor. Dalio even suggested that these weightings “would not be exact or perfect.” I explore options for actual risk parity below later on in this post. In fairness, the real All Weather Fund at Bridgewater – after which this portfolio is modeled – seems to be based on risk parity of the 4 economic seasons listed below.
As the name suggests, the All Weather Portfolio is designed to be able to “weather” any storm. It uses asset class diversification based on seasonality in the interest of limiting volatility and drawdowns. The holdings and the allocations thereof correspond to Dalio's view on economic “seasons.” Dalio's strategy and expertise are so pervasive that the phrase “all weather” is now used to describe other portfolios that behave like his in surviving any economic climate, e.g. “investing in an all weather portfolio.”
Dalio proposes that the following four things affect asset value:
- Inflation
- Deflation
- Rising economic growth.
- Declining economic growth.
Based on these, Dalio expects we can see 4 “seasons” of the economy:
- Higher than expected inflation.
- Lower than expected inflation.
- Higher than expected economic growth.
- Lower than expected economic growth.
Dalio chose asset classes that performed well in each of these different seasons, with the goal being diversification that allows for consistent growth and small drawdowns. To minimize volatility, the portfolio is mostly bonds, and only allocates 30% to stocks.
Ray Dalio All Weather Portfolio Asset Allocation
The All Weather Portfolio asset allocation looks like this:
- 30% U.S. Stocks
- 40% Long-Term Treasury Bonds
- 15% Intermediate-Term Treasury Bonds
- 7.5% Commodities
- 7.5% Gold
How To Build the Ray Dalio All Weather Portfolio
M1 Finance would be a good choice for U.S. investors to implement the All Weather Portfolio so that you can easily and seamlessly rebalance as often as you'd like, and it has zero transaction fees. I wrote a comprehensive review of M1 Finance here.
Using some low-cost Vanguard funds and my “best in class” ETFs, we can build the All Weather Portfolio (here's the pie) like this:
30% VTI
40% VGLT
15% SCHR
8% GLDM
7% BCI
To add this pie to your portfolio, just click this link and then click “Invest in this pie.”
Canadians can find the above ETFs on Questrade or Interactive Brokers. Investors outside North America can use eToro or possibly Interactive Brokers.
To diversify internationally with the All Weather Portfolio above, simply replace VTI (Vanguard's total US stock market ETF) with VT (Vanguard's total world stock market ETF). That pie can be found here.
All Weather Portfolio Performance vs. S&P 500
With live fund data for Commodities only going back to 2003, here's the All Weather Portfolio vs. a traditional 60/40 portfolio and the S&P 500 through 2022:
As we'd expect, the All Weather Portfolio has had less than half the volatility and, consequently, a much higher risk-adjusted return (Sharpe) and significantly smaller drawdowns versus the S&P 500. This shows the All Weather Portfolio doing precisely what it's intended to do: weather any storm.
Personally, I probably wouldn't adopt the All Weather Portfolio unless I were near or at retirement age, or if for some reason I absolutely couldn't mentally and emotionally endure volatility and drawdowns (which is a very real case for some). The gold, commodities, and heavy bond allocation would likely just drag down long-term total return since it only has 30% allocated to stocks.
We can see the All Weather Portfolio had a slightly lower return with a nearly identical risk-adjusted return compared to a 60/40 portfolio for this period.
I'm also usually not even a fan of gold or commodities period; they have no place in my portfolio at this time. That said, this would be a good set-and-forget portfolio, making it attractive for investors who want to be hands-off, and it is probably my favorite of the volatility-minimizing, any-economic-climate portfolios like the Permanent Portfolio, Golden Butterfly, and Ivy Portfolio.
However, as I'll explore later below, levering up this same asset allocation may dramatically improve returns while still maintaining a sensible level of portfolio risk similar to that of an unlevered 100% stocks position, depending on the amount of leverage used.
But first, let's talk about swapping out those Commodities for something else.
Using Utilities Instead of Commodities (and REITs)
Commodities have been used in the past for their purported diversification benefit from their inherent low correlation to the total stock market, and the nature of the asset class being physical necessities on which futures are traded. The same can be said for real estate (REITs).
Commodities are materials like gold, oil, copper, livestock, agriculture, etc. Their value depends on their usage in production, and is directly related to supply and demand. Any position in commodities is therefore a bet on the short-term future, rather than long-term growth associated with stocks, bonds, real estate, etc. As such, investors have historically turned to commodities as a hedge against uncertainty.
Unfortunately, commodities themselves are unpredictable by their very nature. Crops go bad. The weather changes. Macroeconomic policies shift. Alternatives to things like copper are found. Ownership, storage, and transportation of commodities increase costs.
Stock ownership is a claim on a company’s future earnings. A bond is a contractual obligation between a lender and a borrower, with interest payments going from the latter to the former. Ownership of a commodity is not value-producing; it involves no earnings or cash flow and is simply a bet on production and/or consumption at that time.
Commodities are obviously useful to your everyday life, but not so much as an investment in securities markets. Think of owning a commodities fund as just paying for their storage somewhere.
With technological advances, we would expect commodity prices to fall or stay flat over the long term. After fees (commodities funds are expensive; the popular PDBC costs 0.62%), commodities are likely losing to inflation. And indeed they have historically; even though they soared in 2021, commodities have had negative real returns over the last 100 years.
Essentially, even in inflationary environments, investors have historically been better off over the long term holding just about anything other than commodities. And we now have assets like REITs, TIPS, etc. as alternatives. Even a narrow gold fund should be a better choice than broad commodities. Commodities may offer a tiny diversification benefit to lower volatility and risk, but we still want our diversifiers to have positive future expected returns.
Andrew Tobias, in The Only Investment Guide You'll Ever Need, maintains that “it is a fact that 90% of all people who play the commodities game get burned. I submit that you have now read all you ever need to read about commodities.”
I don't bet on sectors, but over the past 20 years, the Utilities sector has had the lowest correlation to the total stock market of any sector, lower than that of both Commodities and REITs – specifically, 0.38 for Utilities compared to 0.53 for Commodities and 0.59 for REITs. Yet REITs and Commodities are both treated as classes of their own and are used as diversifiers, while Utilities are simply thrown under the equities umbrella at their market weight and forgotten.
Using live fund data for Commodities going back to 2002, through 2022, the results corroborate this idea. I backtested portfolios using 90% total stock market with respective 10% tilts using Utilities, Commodities, and REITs. Utilities provided the lowest volatility, smallest drawdown, highest return, and highest risk-adjusted return:
Some other macroeconomic things to consider are:
- We'll always need energy of some form, whether that's solar, wind, natural gas, etc.
- Utility operating costs are passed on to the ratepayer.
- Demand for utilities stays relatively constant (it is said to be noncyclical), which is a significant reason why utilities perform comparatively well during market downturns.
- Perhaps most importantly, using Utilities instead of REITs lets you avoid the idiosyncratic, uncompensated risk of the real estate market.
- Of all sectors, Utilities are the least explained by the known equity factors that explain the differences in returns between diversified portfolios. Their R-squared ratio is notably lower than that of REITs. From this paper:
So, at least for the relatively small 7.5% slice for the All Weather Portfolio, I would submit that Utilities are a fine – and likely superior – replacement for Commodities.
Here's how that swap would have worked out historically:
Notice how the AWP with Utilities beat the “normal” AWP on almost every important metric.
For those wanting this All Weather Portfolio using Utilities instead of Commodities, I've created that pie here. It looks like this:
30% VTI
40% VGLT
15% SCHR
8% FUTY
7% GLDM
You can invest in this portfolio here.
Applying Leverage To the All Weather Portfolio
Applying leverage to those same allocations gets you enhanced exposure to what is traditionally a low-risk, low-volatility portfolio. Most people don't realize that Dalio and Bridgewater themselves deploy leverage in their in-house All Weather Fund. So perhaps this could be a solution for a risk-averse investor who still wants to get in on the leverage game in an attempt at higher returns to beat the market.
Of course, this isn’t a new idea. Stemming from the foundation of Markowitz's Modern Portfolio Theory is the demonstrable idea that we can allocate assets based on their relative “risk” – not dollar amount – and then lever up to increase exposure, thereby constructing a more efficient portfolio than one of a single asset. For example, a classic 60/40 portfolio levered up to match the volatility and risk of 100% equities has delivered higher returns historically. As an aside, this is exactly what the fund NTSX does and is also the basis of the famous Hedgefundie strategy.
Leverage is the necessary key to generating the requisite return for any given efficient (“optimal”) portfolio for which the expected return is too low to meet the investor’s objective. After all, the goal is to fund future consumption, not just achieve the best Sharpe ratio. By “efficiency,” we mean greater return per unit of risk, or what we call risk-adjusted return, of which Sharpe ratio is one definition. This concept holds true across multiple definitions of “risk,” including volatility, tail risk, and even Dalio’s economic “seasons.”
Allocating multiple assets based on risk and levering up makes use of the benefits of diversification, on which I’m always harping. Remember, diversification is the closest thing we have to a free lunch in investing. Leverage allows us to make use of that free lunch while also increasing expected return. We see this demonstrated here in Dalio’s famous retail portfolio, in which return-driving stocks take up a relatively low allocation, while comparatively lower risk bonds comprise a higher allocation, which is uncharacteristic of traditional growth portfolios. Conventional portfolios increase expected return not with leverage but with concentration in riskier assets, e.g. stocks.
The legendary Cliff Asness maintains that risk parity “works” precisely due to investors’ irrational aversion to – or the disallowed use of – leverage, and due to lower risk assets having higher risk-adjusted returns than they theoretically should. The story goes that because some investors are averse to or are not allowed to use leverage and thus go the route of concentration, those assets in which they concentrate have lower expected returns, leaving higher risk-adjusted returns to be had by lower-risk assets. This is consistent with the famous “low beta anomaly” in stocks. There’s also another theoretical underpinning that favors the risk parity and leverage approach: Past volatility is more correlated with future volatility than past returns are with future returns. That is, we can make more reliable guesses about future asset volatility than about future asset pricing.
The knee-jerk aversion to the word “leverage” usually comes from those who have a poor understanding of its utility in multi-asset portfolios. Anyone who has a mortgage is using leverage. If you’ve taken out student loans, you’ve levered up your human capital. Even if you’re 100% equities, the companies you’re holding carry debt on their balance sheets.
Moreover, leveraged scenarios that blew up historically were usually the result of combining leverage with concentration in a single asset. “Return stacking,” as it’s been called, is simply adding different sources of returns within a leveraged framework to avoid sacrificing performance. In doing so, the leverage ratio per se tells us little about the riskiness of the portfolio. Again, products like PSLDX and NTSX are examples of this concept.
Make no mistake that increasing leverage is increasing risk. But Asness submits that, compared to concentration, the use of leverage is a more manageable, more rewarding, and more reliable assumption of additional risk. I’ve said sort of the same thing when discussing stock picking with people, noting that the investor willing to take on the greater risk of picking a handful of stocks should almost certainly prefer to simply apply leverage to a broad index commensurate with the same level of risk instead. Even Warren Buffett, the ever ardent proponent of an S&P 500 index fund, has long used leverage in the form of low-cost insurance float, which he noted has always given Berkshire “quite an edge,” to amass a fortune.
Some obligatory preliminary warnings:
- I'm not a financial advisor and this is not financial advice. It's for informational and recreational purposes only. Products discussed below are for illustrative purposes only. This is not a recommendation to buy, sell, or transact in any of the products mentioned. Do your own due diligence.
- Past results do not indicate future performance.
- Research and read up on the fundamentals of leverage and the nature of leveraged ETF products before blindly buying in. It can potentially be a very dangerous game. Using leverage increases the potential for greater returns but also the potential for greater losses. The use of leverage increases portfolio risk, and investors face a real possibility of losing all money invested. Specifically with leveraged ETFs, these are relatively new, exotic products that behave differently than “regular,” unleveraged index ETFs. Do your due diligence and read the fine print.
- Similarly, don't put your entire portfolio in a strategy like this. If you want to play with something like this, do so with a small piece of your total portfolio, and definitely not with money you'll need in the next 5-10 years. If you're using M1 Finance, make this a sub-pie within your larger pie.
- I had a hard time finding leveraged commodity index ETF's (probably for good reason) so I somewhat arbitrarily chose a couple oil-and-gas ones I found. There may be better options. This is more just meant to be a discussion starter and a first iteration from which to improve. I discuss improvements using the Utilities sector below.
- Moreover, I realize this may not be the optimal approach for what it's trying to achieve. Again, I'm not a fan of gold and commodities; I think I've fared better playing around with tilting with utilities and consumer staples in my cauldron of leverage: UPRO, TQQQ, TMF, UTSL, and NEED. This is simply taking the unedited All Weather Portfolio and applying leverage to it. More on this later, with variations using Utilities in place of Commodities.
- Of course, I also realize that applying leverage to the All Weather Portfolio sort of defeats its whole purpose, but the historical volatility, return, and drawdown KPI's compared to the S&P 500 index and a 60/40 portfolio were much more impressive than I anticipated. More on that later. I'm viewing it as a way to diversify and limit volatility and drawdowns with the underlying assumption that one has already chosen to invest with leveraged products.
- The linked backtests using Commodities only go back to 2006. We get to see performance through the GFC, but not during a runaway period like the late 1970's in the US. The All Weather Portfolio's large allocation to long-term treasury bonds still expose it to interest rate risk.
- The fees alone for the gold and commodities may be a major drag on the strategies' performance, with the trade-off being the extra diversification and subsequent drawdown protection and volatility reduction. They are probably not responsible for any of the actual returns of the strategies. Is it worth it? I don't know.
- This would require some regular maintenance in the form of rebalancing, as the allocations may stray quickly. Quarterly should be fine. I explore different rebalancing intervals at the end of this post.
- M1 Finance doesn't allow a 7.5% holding so I used 8% gold and 7% commodities.
Aren't Leveraged ETF's Unsuitable for Holding Long-Term?
There's no shortage of articles lambasting the idea of holding leveraged ETF's for more than one day, reciting that they should “only be used for intraday trading.”
However, I firmly maintain that the idea that leveraged ETF’s are unsuitable for long-term holding is largely overblown fear mongering that’s been wrongly perpetuated after the financial blogosphere took the scary-sounding “volatility decay” and ran with it.
“Volatility decay” sounds like a bad thing, but viewing it as such is simply a misunderstanding of what it actually is and its underlying mechanics and arithmetic. If you're curious to see the math, check out this page.
That same “decay” actually works in your favor when the market goes up with decent momentum, which it does more often than it goes down. This is why UPRO, the 3x leveraged S&P ETF, has delivered close to 5x the returns of the SPX since its inception instead of the proposed 3x.
In short, I concede that leverage can indeed be dangerous in certain situations, but can be useful in others.
2x Leveraged All Weather Portfolio
So here's the portfolio using 2x leverage:
30% SSO – 2x S&P 500
40% UBT – 2x LT treasury
15% UST – 2x IT treasury
7.5% DIG – 2x oil and gas
7.5% UGL – 2x gold
You can add this pie to your M1 Finance portfolio here, but stay tuned for the variation using Utilities below, as I don't feel completely comfortable using DIG.
3x Leveraged All Weather Portfolio
And here's that portfolio with 3x leverage:
30% UPRO – 3x S&P 5
40% TMF – 3x LT treasury
15% TYD – 3x IT treasury7.5% GUSH – 3x oil and gas7.5% UGLD – 3x gold
Update April 2020: After Direxion's changing GUSH from 3x to 2x effective March 31, 2020, I would consider using 3x Utilities, REITs, or Consumer Staples, accessed via UTSL, DRN, and NEED respectively. Note that because of this, the 2 backtests immediately below are no longer accurate. Skip to the next section to get the most recent updates and new backtests. Unfortunately, at the time of writing, M1 Finance doesn't offer the NEED (3x Consumer Staples) ETF. Below I explain why Utilities are probably the best choice, and I've included a pie link for that option.
Update June 22, 2020: With the recent market turmoil, Credit Suisse announced today that it plans to delist some of its 3x leveraged VelocityShares™ ETN's, including UGLD. With no other 3x gold funds available, the next logical choice would be UGL, the 2x gold ETF from ProShares. Keep in mind gold is already an extremely volatile asset, so this shouldn't throw anything off too much considering it only has a 7.5% allocation. This will slightly change the allocations of the leveraged risk parity portfolio at the end of this post. Also note that because of this, the 2 backtests immediately below are no longer accurate. Skip to the next section to get the most recent updates and new backtests.
Below is a backtest going back to 2006 comparing the above 2 with the Hedgefundie 55/45 strategy and the S&P 500 index.
Note the highest Sharpe ratio (risk-adjusted return), albeit only slightly, for the All Weather 2x above, with volatility almost identical to the S&P 500 but with much better Worst Year and Max Drawdown (from the GFC) stats.
Here's a backtest going back to 2006 comparing the 2x All Weather above to the unleveraged All Weather Portfolio, a traditional 60/40 stocks/bonds portfolio, and the S&P 500 index:
The 60/40 achieves the highest Sharpe but with much lower return than and a nearly identical Worst Year and Max Drawdown to the 2x All Weather.
Update April 2020: After Direxion's changing GUSH from 3x to 2x effective March 31, 2020, I would consider using 3x Utilities, REITs, or Consumer Staples, accessed via UTSL, DRN, and NEED respectively. Unfortunately, at the time of writing, M1 Finance doesn't offer the NEED (3x Consumer Staples) ETF. Below I explain why Utilities are probably the best choice.
3x Leveraged All Weather Portfolio Using Utilities
Here's a pie using UTSL (3x Utilities) in place of broad commodities. Interestingly, this may actually be a better choice anyway considering the backtest below.
30% UPRO – 3x S&P 500
40% TMF – 3x LT treasury
15% TYD – 3x IT treasury
7.5% UTSL – 3x utilities7.5% UGLD – 3x gold
Update June 22, 2020: With the recent market turmoil, Credit Suisse announced today that it plans to delist some of its 3x leveraged VelocityShares™ ETN's, including UGLD. With no other 3x gold funds available, the next logical choice would be UGL, the 2x gold ETF from ProShares. Keep in mind gold is already an extremely volatile asset, so this shouldn't throw anything off too much considering it only has a 7.5% allocation. This will slightly change the allocations of the leveraged risk parity portfolio at the end of this post.
Backtests of leveraged ETFs below are using my own data series I created in an attempt to accurately simulate how these leveraged ETFs – which are relatively new products – would have behaved historically. This factors in their daily resetting, fees, and borrowing costs, so returns shown are net of those things.
The backtest below compares 1x (“normal”), 2x, and 3x versions of the All Weather Portfolio, all using Utilities instead of Commodities and all using 2x gold since a 3x gold ETF is no longer available. It goes back to 1987 and runs though February, 2021:
Interestingly, the risk-adjusted return of all of them (Sharpe) was identical, and higher than that of the S&P 500. The 2x and 3x versions of the AWP above have higher general and risk-adjusted returns than the S&P 500, though obviously that came with greater volatility.
2x Leveraged All Weather Portfolio Using Utilities
With the loss of UGLD (3x gold), you may be interested in simply using a 2x All Weather Portfolio. Incorporating Utilities in place of Commodities as described above, we arrive at the following allocations:
SSO – 30%
UBT – 40%
UST – 15%
UPW – 8%
UGL – 7%
You can add this pie to your M1 Finance portfolio here.
True Risk Parity
Update January 2023: Previously I had a section here talking about a “true” risk parity weighting scheme based on the per se volatility of each constituent asset, which is usually what we mean when we say “risk parity.”
After a lot of reflection on this initially innocuous experiment, I realized that interpretation – without trying to – basically bastardizes Dalio’s original idea of using specific assets at specific weights to achieve a rough parity of the risks of the aforementioned economic seasons, for which certain assets span multiple seasons.
I also concluded it probably didn’t make much practical sense anyway due to the inclusion of multiple funds of the same asset type as independent inputs of an optimization algorithm. Ironically, it would make more sense to do so only after trimming down the fund selection to a single fund for each asset class, but of course at that point we’re talking about another portfolio entirely. In any case, it created a pretty significantly different portfolio, relatively speaking considering the aforementioned seasons-parity thesis underpinning Dalio’s original proposition.
Lastly, it was also confusing for readers not knowing whether or not they should consider what was essentially a theoretical exercise as a viable real-world investment strategy.
For these reasons, I’ve decided to remove that discussion in its entirety. I probably should have never published it in the first place. You can perhaps take solace in the fact that it backtested worse than everything else you see on this page.
Addressing Concerns Over Bonds
I've gotten a lot of questions about the use, utility, and viability of bonds being a significant chunk of the All Weather Portfolio. I'll briefly address and hopefully quell these concerns below.
Common comments nowadays about bonds include:
- “Bonds are useless at low yields!”
- “Bonds are for old people!”
- “Long bonds are too volatile and too susceptible to interest rate risk!”
- “Corporate bonds pay more!”
- “Interest rates can only go up from here! Bonds will be toast!”
- “Bonds return less than stocks!”
So why use bonds?
- Bond duration should be roughly matched to one’s investing horizon, over which time a bond should return its par value plus interest. Betting on “safer,” shorter-term bonds with a duration shorter than your investing horizon could be described as market timing, which we know can’t be done profitably on a consistent basis. This is also a potentially costlier bet, as yields tend to increase as we extend bond duration, and long bonds better counteract stock crashes. More on that in a second.
- Moreover, in regards to bond duration, we know market timing doesn’t work with stocks, so why would we think it works with bonds and interest rates? Bonds have returns and interest payments. A bond’s duration is the point at which price risk and reinvestment risk – the components of what we refer to as a bond’s interest rate risk – are balanced. In this sense, though it may seem counterintuitive, matching bond duration to the investing horizon reduces interest rate risk and inflation risk for the investor. An increase in interest rates and subsequent drop in a bond’s price is price risk. A decrease in interest rates means future coupons are reinvested at the lower rate; this is reinvestment risk. A bond’s duration is an estimate of the precise point at which these two risks balance each other out to zero. If you have a long investing horizon and a short bond duration, you have more reinvestment risk and less price risk. If you have a short investing horizon and a long bond duration, you have less reinvestment risk and more price risk. Purposefully using one of these mismatches in expectation of specific interest rate behavior is intrinsically betting that your prediction of the future is better than the market’s, which should strike you as unlikely.
- It is fundamentally incorrect to say that bonds must necessarily lose money in a rising rate environment. Bonds only suffer from rising interest rates when those rates are rising faster than expected. Bonds handle low and slow rate increases just fine; look at the period of rising interest rates between 1940 and about 1975, where bonds kept rolling at their par and paid that sweet, steady coupon.
- New bonds bought by a bond index fund in a rising rate environment will be bought at the higher rate, while old ones at the previous lower rate are sold off. You're not stuck with the same yield for your entire investing horizon.
- We know that treasury bonds are an objectively superior diversifier alongside stocks compared to corporate bonds. This is also why I don't use the popular total bond market fund BND. It has been noted that this greater degree of uncorrelation between treasury bonds and stocks is conveniently amplified during periods of market turmoil, which researchers referred to as crisis alpha.
- Again, remember we need and want the greater volatility of long-term bonds so that they can more effectively counteract the downward movement of stocks, which are riskier and more volatile than bonds. We're using them to reduce the portfolio's volatility and risk. In short, more volatile assets make better diversifiers. Dalio seems to be one of the few who truly understand this. Designers of many other lazy portfolios are often using short or intermediate bonds, of which the volatility contribution pales in comparison to that of stocks.
- This one's probably the most important. We're not talking about bonds held in isolation, which would probably be a bad investment right now. We're talking about them in the context of a diversified portfolio alongside stocks, for which they are still the usual flight-to-safety asset during stock downturns. Though they admittedly provided a major boost to this strategy's returns over the last 40 years while interest rates were dropping, a low interest rate environment does not suddenly mean bonds stop doing their job. Even if rising rates mean bonds are a comparatively worse diversifier (for stocks) in terms of future expected returns during that period does not mean they are not still the best diversifier to use.
- Similarly, short-term decreases in bond prices do not mean the bonds are not still doing their job of buffering stock downturns.
- Historically, when treasury bonds moved in the same direction as stocks, it was usually up.
- Interest rates are likely to stay low for a while. Also, there’s no reason to expect interest rates to rise just because they are low. People have been claiming “rates can only go up” for the past 20 years or so and they haven't. They have gradually declined for the last 700 years without reversion to the mean. Negative rates aren't out of the question, and we're seeing them used in some foreign countries.
- Bond convexity means their asymmetric risk/return profile favors the upside.
- Long bonds have beaten stocks over the last 20 years. We also know there have been plenty of periods where the market risk factor premium was negative, i.e. 1-month T Bills beat the stock market – the 15 years from 1929 to 1943, the 17 years from 1966-82, and the 13 years from 2000-12. Largely irrelevant, but just some fun stats for people who for some reason think stocks always outperform bonds.
- Again, I acknowledge that post-Volcker monetary policy, resulting in falling interest rates, has driven the particularly stellar returns of the raging bond bull market since 1982, but I also think the Fed and U.S. monetary policy are fundamentally different since the Volcker era, likely allowing us to altogether avoid runaway inflationary environments like the late 1970’s going forward.
David Swensen summed it up nicely in his book Unconventional Success:
“The purity of noncallable, long-term, default-free treasury bonds provides the most powerful diversification to investor portfolios.”
Ok, bonds rant over. If you still feel some dissonance, the next section may offer some solutions.
Reducing Volatility and Drawdowns and Hedging Against Inflation and Rising Rates
It's unlikely that any of the following will improve the total return of the portfolio, and whether or not they'll improve risk-adjusted return is up for debate, but those concerned about inflation, rising rates, volatility, drawdowns, etc. and/or the future ability of long bonds to adequately serve as an insurance parachute may want to diversify a bit further with some of the following options:
- LTPZ – long term TIPS (inflation-linked bonds).
- SCHP – intermediate TIPS.
- VTIP – short TIPS.
- FNCL – financials (banks tend to do well when interest rates rise because they make more money on loans).
- FAS – 3x financials.
- VWO – emerging markets (diversify outside the U.S.).
- EDC – 3x emerging markets.
- EURL – 3x Europe.
- MCHI – China (lowly correlated to the U.S.).
- YINN – 3x China.
- VNQ – REITs (arguable diversification benefit from “real assets”).
- DRN – 3x REITs.
- EDV – U.S. Treasury STRIPS.
- TYD – 3x intermediate treasuries – less interest rate risk.
- UDOW – 3x the Dow – greater loading on Value and Profitability factors than UPRO.
- TNA – 3x Russell 2000 – small caps for the Size factor.
- TAIL – OTM put options ladder to hedge tail risk. Mostly intermediate treasury bonds and TIPS.
I may explore some new backtests with some of these as added diversifiers in the near future.
What we can say about the portfolio's heavy use of bonds is that rising rates would cause lower returns for this portfolio going forward compared to what it's seen in the past, even if the bonds still help reduce volatility and risk. Investors can mitigate that inflation risk by using TIPS, international stocks, and perhaps a nominal bond barbell using long bonds and short bonds (instead of the prescribed long and intermediate), as short bonds are a decent inflation “hedge” since new bonds at higher rates can be rolled faster.
As I've noted, while gold and broad commodities would almost certainly help weather an inflationary period “better” than just stocks and nominal bonds, I don't have as much confidence in them as a true inflation “hedge,” and I'd much rather use other assets like TIPS and short bonds. In fairness, TIPS weren't even around yet when Dalio first proposed the All Weather Portfolio's components.
Optimal Rebalance Interval
Regular rebalancing is necessary with any of these leveraged portfolios, as the leveraged funds will likely quickly stray from their target allocations. Below I've compared the historical metrics from different rebalance intervals (annually, semi-annually, quarterly, and monthly) for the 3x version using Utilities for the period 1987-2019:
Interval | CAGR | St. Dev. | Max Drawdown | Sharpe |
---|---|---|---|---|
Annually | 22.53% | 22.71% | -39.72% | 0.88 |
Semi-Annually | 22.23% | 21.90% | -38.23% | 0.89 |
Quarterly | 23.17% | 21.69% | -40.37% | 0.93 |
Monthly | 22.06% | 21.52% | -44.34% | 0.90 |
As I suspected, quarterly rebalancing performed the best historically by a decent margin, and would probably be the best choice going forward.
Conclusion
Were I to implement any of these leveraged versions, I'd probably go with the “regular” 3x using Utilities, as the risk parity now with UGL doesn't seem worth the trade-off to me, and rebalance quarterly, using M1 Finance. M1 features dynamic rebalancing so it will automatically direct new deposits to maintain your target allocations until the point at which your portfolio value becomes so high that your regular deposits can't keep things in balance (a great problem to have).
Canadians can find the above ETFs on Questrade or Interactive Brokers. Investors outside North America can use eToro or possibly Interactive Brokers.
That doesn't mean you should blindly copy it. I'm relatively young with a long investing horizon and a high risk tolerance. As I said earlier, assess your own personal factors before diving into leveraged funds. The assumption of more risk, especially with the use of leverage, gives you the potential for more reward, but also the potential for greater losses.
In any case, one should deleverage as time passes. If you have a shorter time horizon, if you don't have the stomach for leverage (most don't), or if you simply don't care for any of this more complicated leverage stuff, the “normal,” original All Weather Portfolio up at the top is perfectly fine. As I noted, I'd still swap out the commodities for utilities due to my contempt for the former.
Hopefully it goes without saying that all of the above leveraged options are not ideal for a taxable account due to the need for regular rebalancing.
Which one would you choose? Does the recent market turmoil have you reevaluating your true risk tolerance and considering an “all-weather” approach? Let me know what you think in the comments.
M1 currently has a promotion for up to $500 when initially funding an investment account:
Are you nearing or in retirement? Use my link here to get a free holistic financial plan from fiduciary advisors at Retirable to manage your savings, spend smarter, and navigate key decisions.
Don't want to do all this investing stuff yourself or feel overwhelmed? Check out my flat-fee-only fiduciary friends over at Advisor.com.
Disclosures: I am long VWO, TMF, UPRO, and UTSL in my own portfolio.
Interested in more Lazy Portfolios? See the full list here.
Disclaimer: While I love diving into investing-related data and playing around with backtests, this is not financial advice, investing advice, or tax advice. The information on this website is for informational, educational, and entertainment purposes only. Investment products discussed (ETFs, mutual funds, etc.) are for illustrative purposes only. It is not a recommendation to buy, sell, or otherwise transact in any of the products mentioned. I always attempt to ensure the accuracy of information presented but that accuracy cannot be guaranteed. Do your own due diligence. I mention M1 Finance a lot around here. M1 does not provide investment advice, and this is not an offer or solicitation of an offer, or advice to buy or sell any security, and you are encouraged to consult your personal investment, legal, and tax advisors. All examples above are hypothetical, do not reflect any specific investments, are for informational purposes only, and should not be considered an offer to buy or sell any products. All investing involves risk, including the risk of losing the money you invest. Past performance does not guarantee future results. Opinions are my own and do not represent those of other parties mentioned. Read my lengthier disclaimer here.
Are you nearing or in retirement? Use my link here to get a free holistic financial plan from fiduciary advisors at Retirable to manage your savings, spend smarter, and navigate key decisions.
Don't want to do all this investing stuff yourself or feel overwhelmed? Check out my flat-fee-only fiduciary friends over at Advisor.com.
Justin says
Awesome writeup!
Quick question, on the 2x AWP – can you post a Portfolio Vis screenshot of the etfs/funds you used including your negative cashx position on the 1987 backtest? I can get to 2005, but can’t find the necessarily backfills on Portfolio Visualizer to get to that 1987 timeframe.
Thanks!
John Williamson says
Thanks Justin! Glad you found the content useful.
If you mean this backtest, I used my own simulation data I created. But these funds should get you a longer timeframe for the negative cash position method.
Justin says
Thanks,
Keep up the good work.
Karl Garz says
Can we get a risk parity 3x leverage portfolio comparison?
John Williamson says
What do you mean? There’s a comparison here of 3x “normal” vs. risk parity.
Karl Garz says
oops apologies. i missed that. thanks. this page is awesome
John Williamson says
Glad you like it!
Will says
Thank you! Great article!
John Williamson says
Thanks Will!
Rishi says
This is great stuff. I ran a backtest between the All Weathered Util (AWU) and a 40(VTI)-60(TLT) portfolio. The CAGR is slightly higher for 40-60 and drawdown is a bit more worse for 40-60 relative to AWU. I am wondering if AWU would perform differently over 40-60 in any market conditions or they work the same. In other words I am trying to see why I should use AWU in the first place.
John Williamson says
Thanks for the kind words, Rishi! Glad it provoked some thought and experimentation.
I think those results are about what we’d expect. Keep in mind the purpose of the AWP is to minimize risk and volatility, so its inclusion of gold and broad commodities smooths out that ride a little more compared to a portfolio of just stocks and bonds, but it usually sacrifices some returns in doing so.
More importantly though, remember we’ve been in a bond bull market for about the past 40 years, so there’s no guarantee bonds will continue to perform as they have. Moreover, I’d be a little hesitant to put that 60% entirely in long bonds.
So then it comes down to risk tolerance – how much volatility and risk you can stomach. It sounds like something like 60/40 or even 80/20 may fit your interests more, which would have higher expected returns than 40/60. For both of those I’d be more comfortable using long bonds.
Rishi says
Thank John. Your reply brought up an interesting thought about AWP, which is tilted to about 55% bonds. For all intents and purposes, interest rates can do one of two things from their current levels – stay about the same or go up. If they stay the same, the return on bonds funds will be about the same as the interest rate on the bonds – currently less than 1% for 10-Year US Treasury bonds and less than 1.5% for 30-Year US Treasury bonds. If interest rates go up, the market value of the bonds will go down and returns will be even lower. The exceptions could be when the market experiences flight to safety.
The backtest timelines used here are during a bond bull market. The times that are closest to where we are today might be the 1950’s, coming off of almost zero interest rates. It would be interesting to see how these portfolios, which are heavy on bonds, would function in say 1950-1980 time periods.
Erwin says
Great article, thank you.
Have you looked at RPAR ETF? It is based on Dalio’s all weather portfolio with some deviations.
Any comments, specially related to the bond allocation given today’s zero rate environment?
John Williamson says
Thanks Erwin, glad you enjoyed it!
Admittedly I haven’t looked too far into the specifics of RPAR, but here are a few thoughts I’ve had from several glances at the fund:
– It’s a relatively expensive way to hold a variation of Dalio’s AWP.
– Its target allocations for commodities and TIPS seem too high IMHO and cause the fund to have lower expected returns.
– It was created by some smart people, and would be a decent way for someone wanting a single fund to take up a volatility-minimizing approach if for some reason they didn’t want to take the time to buy a few more funds.
– I’d be more likely to just use a group of the cheaper ETFs I listed above to build the AWP.
Unrelated to RPAR, I know Dalio mentioned recently that he’s bullish on China as a way to hedge against U.S. stocks if bonds don’t provide sufficient downside protection going forward. KWEB might be a good choice for a diversifier if the investor is specifically looking to invest in Chinese stocks.
I’m not a fan of the Golden Butterfly Portfolio’s relatively large allocation to gold, but the modified version here might be a good alternative for someone looking for an all-seasons portfolio that’s not quite so bond-heavy.
John Williamson says
Hey Erwin, I finally got around to reviewing the RPAR ETF here.
SenorB says
John,
Great content on this blog, keep it up.
You present a solid case for the 3x Leveraged AWP Using Utilities. There’s something I don’t understand. When I backtest 3x SPY, in isolation in order to mimic UPRO, covering the 2008-2009 crash, it breaks, and I get an error message saying that the backtest didn’t survive the drawdown of that period. I understand that as a whole, the AWP is supposed to survive a significant drawdown, but why does that individual ETF fail? Is it a limitation of the backtest algorithm? Or maybe I’m doing it incorrectly. My concern is that in real life UPRO wouldn’t survive a decline such as we saw in 2008-2009, or worse. Is this a valid concern?
What am I missing?
Thanks.
John Williamson says
Thanks for the kind words!
Your backtest failed because a 100% 3x SPY position would have been wiped out in the 2008 crash. Note that the behavior of 3x SPY is slightly different from UPRO since leveraged ETFs reset daily. Because of this, a leveraged ETF can’t actually go to zero, because the market has circuit breakers that halt trading at 7%, 13%, and 20%. We saw these circuit breakers in action in the recent COVID crash.
As you’ve seen, for a single ETF like UPRO in isolation, a steep drop would indeed be concerning because it would take longer to recover, but this is precisely why we’re diversifying across uncorrelated asset classes in this case to protect the downside and reduce volatility and risk, thereby attempting to replicate the strategy of the “regular” All Weather Portfolio while juicing returns.
A 100% UPRO position – or 100% in any 3x leveraged ETF – would never be a smart move in my opinion.
Neil says
Hi John,
Thanks so much for your article.
How do you think about the purpose/function of intermediate bonds, vs doing 100% TLT or the long-term equivalent?
-Neil
John Williamson says
Glad you found it useful, Neil!
I’m of the mind that, even though bonds shouldn’t be feared at low, zero, or negative rates, a bond-heavy portfolio should probably not be entirely long-term bonds, especially for those with shorter time horizons. We’d be exposed to more interest rate risk and we’d be betting that stocks are more likely to go down.
With an asset allocation like 90/10 or 80/20, I’d gladly go all in with long-term bonds on the fixed income side, but here with the AWP we’ve got 55% bonds.
All that being said, going 100% long-term bonds is probably not going to make or break the portfolio. Going back to 1992, making that swap resulted in a slightly higher return but at the cost of greater volatility and larger drawdown and thus a lower risk-adjusted return as measured by Sharpe. Keep in mind though that for the most part this entire time period was a bond bull market with falling interest rates.
Neil says
Got it, super helpful.
I’m apprehensive about bonds because I do think we’ve had a secular bull market in bonds that makes further progress harder.
To your point on utilities as having some of the characteristics of bonds (inflation protected, always need, regulated so legally mandated returns on capital, etc) how do you think a 2x levered AWP portfolio where you ditch bonds, in whole or in part, for utilities would perform?
And, last question… if you have a porfolio that lets you use cheap (1-2%) actual margin, is there a benefit to implementing the 2x AWP by levering up the individual component ETFs to a 2x weight, instead of using a 2x levered ETF? Would imagine you’d get less vol drag, lower fees, and retained dividends, but in momentum times lose out on the “vol drag” when irt benefits you.
John Williamson says
Impossible to know the future so I don’t know how one would perform. You might consider diversifying internationally with the stocks for some extra risk reduction. You could also just use the 3x at a lower allocation to achieve the leverage ratio you want and save some on fees over the 2x funds.
As for margin vs. ETFs, depends on the leverage ratio you’re going for. You’d still probably come out ahead on fees with the ETFs if you’re not using them for your entire portfolio. But yes you’re right, less vol. drag with margin. Also consider the relative tax-inefficiency of these leveraged funds in a taxable account, and that margin isn’t available for retirement accounts.
Koldeep says
correct if I’m wrong, but wouldn’t margin calls wipe one out at larger positions if doesn’t keep a wad of cash sitting as collateral? (Max drawdown was close to 50% for 3x). In recession, if you’re leveraged 3x and there is a 12% drop in your portfolio, the NAV would drop 36%; new position is 1.92x (don’t meet margin call) instead of 2.64x (meet margin call).
Am I losing the math here?
John Williamson says
The strategies here use leveraged ETFs, so no margin calls. But yes, if you were using margin, you may incur margin calls depending on drawdowns and the broker’s maintenance requirements.
Akshay says
For all the above portfolios: How frequently did you rebalance them?
John Williamson says
Annually.
Strategist GJ says
Quick question, for your “risk parity” leveraged version at the end, are you effectively allocating 20% risk to each of the 5 leveraged positions? If so, shouldn’t it be 12.5% for both Gold and Utilities seeing that those are supposed to, together, form your inflation protection, rather than separately? I suppose you may have already considered that, and I also might be wrong, but I figure if you hadn’t, that might be a great way to “split the difference” between your two options compared in the bottom chart?
John Williamson says
Risk parity of the assets means all the assets have the same volatility, contributing the same amount of “risk” to the portfolio. It’s not just 20% to each asset.
Tony says
Hi John,
I noticed that the average return of the 3xAWP (~22%) is not 3 times the return of the AWP (~29%), whereas this relationship holds for the volatilities as expected. What explains this difference in performance from the expected performance of the 3xAWP?
Thanks,
Tony
John Williamson says
Hey Tony, where are you getting those numbers?
Tony says
Hi John,
I mixed up the numbers. 3xAWP from backtest comes at 19.82% whereas the AWP comes at 7.93%, hence 3 times that is 23.79%. All the above are for the Dalio’s AWP. If you use Util in place of Commodities the 3xAWP strategy comes at 22.98% whereas the AWP comes at 9.08%, hence 3 times that is 27.24%. In both cases there is a gap between the expected 3xAWP and the realized 3xAWP average return. What explains this difference? In both cases it is around 4% per year.
Thanks,
Tony
John Williamson says
CAGR is not average return. We also wouldn’t expect the CAGR’s or average returns of these portfolios to be perfect multiples of one another. Remember these leveraged products aim to deliver 3x the daily return of the underlying index; it says nothing about their long-term behavior. The leveraged ETFs also have much higher fees, and beta slippage from resetting daily. Lastly, remember that a 10% drop requires an 11% gain to recover, but a 20% drop requires a 25% gain to recover.
Josh says
Hey John,
really grateful about the work you’ve done here. I’m a big fan of the All Weather Portfolio and a big fan of Ray Dalio. It is really interesting to see the portfolio’s returns juiced up through leverage.
I was just wondering, since I also never invested or traded a leverage ETFs, if it is really possible to hold for example UPRO for a longer period of time? The link you’ve provided about “The Long Term Behaviour of Leveraged ETFs” was very interesting to me, but they’ve ended their article with: “Leveraged ETFs can drop to zero if the market drops enough in one day. You can lose all your money.”
Therefore, I’ve also read the underlying Paper by Tony Cooper and tried to follow it as much as my mathematics would allow me. Does that mean, through rebalancing the leverage on a daily basis, a drawdown (not in 1 day) of let’s say 35% of the S&P 500 does not necessarily mean a total loss of my 3x leverage S&P500 position, but a rather big decline of the S&P 500 in a single day could?
Thank you so much again for your work and looking forward to your answer.
Beste regards,
Josh
John Williamson says
Josh, thanks so much for the kind words and for your comment.
It is definitely possible to hold leveraged ETFs long-term; I do it myself.
In theory, a 3x ETF would be wiped out if the underlying index drops by 33.4% in one day, as you’ve noted. In actuality, this can’t happen, because the market has circuit breakers that halt trading at 7%, 13%, and 20%, and these leveraged products reset daily. We saw these circuit breakers in action in the recent COVID crash. So while these leveraged ETFs can indeed experience steep drops, they can’t actually go to zero.
For a single ETF like UPRO in isolation, a steep drop would indeed be concerning because it would take longer to recover, but this is precisely why we’re diversifying across uncorrelated asset classes in this case to protect the downside and reduce volatility and risk, thereby attempting to replicate the strategy of the “regular” All Weather Portfolio while juicing returns.
Josh says
Thank you so much for the super fast reply and clearing things up! All the best.
Karl Garz says
Any role (and any availability) of leveraged developed and developing world etfs, small and mid cap, and international diversified treasury etfs? I know the US and the dollar yada yada, but the rise of crypto, the rise of china (an opaque unfair system), recent trump debacle, the polarization, the civil war talk, the system is more fragile than it lets on. i’m not sure US treasuries can be considered safe from a 200 year experience, especially that 200 years is an average life span of human made system. I’m ranting, but is their diversification outside US treasuries and securities with leverage/ without leverage
John Williamson says
Hey Karl, good question. Direxion actually recently closed many of their 3x international (ex-US) funds including DZK which was 3x Developed Markets, but it had severed tracking error anyway. I’ll see if I can hunt down any leveraged Emerging Markets funds, since they’d offer a superior diversification benefit over Developed Markets anyway, as Developed Markets are highly correlated with the U.S. Emerging Markets have also historically paid a significant risk premium.
A good option may be a dash of YINN, which is 3x China, since it’s lowly correlated to the U.S. market, and I think Dalio himself has mentioned he’s bullish on China.
Some other attractive diversifiers may be DRN (3x REITs) and LTPZ (long term TIPS; not leveraged). You could also obviously simply park a large sum of money in a “regular” Emerging Markets fund like VWO from Vanguard.
Karl Garz says
Thank you for the suggestions.
Not sure one can trust china. No history and no present of independent institutions to safeguard private property.
Karl Garz says
the circuit breakers can’t be said of US treasuries and Gold 3x ETFs right? though unlikely for treasuries to drop 33% in a day, but idk, can any scenario cause this? default? government/congress conflict on budget or something like we saw recently?
also, gold is now 2x which is safer but what scenario has previously crashed the ugld?
John Williamson says
Right, but if those are going down, I’d hope stocks are going up. Rising rates and/or hyperinflation are scenarios where that may happen, but I don’t think it’d happen in a single day. Hard to say what causes gold to crash. It crashed in 2008 and late 2012. It’s usually simply used as a diversifier due to its low correlation to both stocks and bonds, not as a primary asset. Focus on the portfolio as a whole, not on each asset in isolation.
Karl Garz says
oh yes, sure, i know. my point is to mitigate the possibility of a wipeout of 3x leveraged etf in a single day freak accident or black swan.
John Williamson says
Indeed. That’s why we’re diversifying across assets in this case – to reduce risk and protect against black swan events. It would obviously be much more of a concern with a single asset type, e.g. 100% UPRO, which would probably be a foolish investment.
Mirco says
Hello John.
THX for that great informations!
I am new to the stock market and try to get as much ( high quality ) information as possible to build a ” lazy portfolio”. I read the book money from Tony Robbins and since than the All Weather Portfolio took my interest most.
I found your very interestingly report and the idea to put leverage to this portfolio is very very interesting for my portfolio which I want to build soon.
After thinking a lot about it ( I am 43 and want to let the portfolio work 15-20 years ; and I am ok with the risk of it ) I probably will go with the 3X AWP with utilities. I think it has the best risk to benefit balance.
But in your last words you write :
“Disclosures: I am long TMF, UPRO, UTSL, and VTI.”
what do you mean with this and what percentages would you go with each of them ?
Greetings from Germany
Mirco
John Williamson says
Mirco, thanks for the comment.
First, as I pointed out in the post, I definitely would not use one of these leveraged strategies with 100% of my portfolio.
Secondly, being new to the market and having that time horizon are even more reasons to not use leverage with the bulk of one’s portfolio in my opinion.
Maybe start out with a normal, unleveraged lazy portfolio and then use a modest amount of leverage after you get some experience and investing knowledge. Most investors don’t know their real risk tolerance until they go through a market crash.
That disclosure at the end just means I hold those tickers in my personal portfolio.
Niklas says
Hello Mirco,
I am from Germany aswell and wondering how you plan to build the Portfolio for yourself since due to the EU-regulations it is not possible to buy most of the american ETFs by using a regular Broker.
I am very interested how you are planning to build it and would be thankfull if you would share your solution!
Regards from Germany
Niklas
Graham George says
Great article! With the US printing so much money and the USD is in danger of losing its reserve currency status over a period of time and China becoming a dominant force, do you think these allocation ratios would work in the future especially, I am concerned about the US bonds part which is in a slippery slope with these mounting US debts and the country lacking productivity edge?
John Williamson says
Thanks Graham! And thanks for your comment.
I think these allocations would still be okay going forward. After all, this portfolio is specifically designed to eliminate the need for these types of worries.
If you don’t have faith in U.S. securities, you could diversify internationally across stocks and bonds or even specifically incorporate Chinese stocks, TIPS, etc.
Tommy Trousers says
Really loved the article! I wanted to ask how you backtested AW2x and AW3x back from 2006 or 2007 considering those ETFs were not around back then, especially the ones like UPRO, that would be down 95+% during the 2008 crash
John Williamson says
Thanks Tommy! I used “normal,” unleveraged versions of those funds at 2x and 3x allocations with a negative cash position to simulate constant leverage.
P B says
Hello
Thank you for the article.
Can you elaborate on the calculation you used for 2x and 3x (value of negative cash position) ?
I am concerned it may not be really reliable and representative. As you explained, with beta slippage, sometimes it has a negative effect and sometimes a positive effect. It may be more accurate to double / triple every daily return. Also, did you take into account the higher ETF fees ?
Thank you
John Williamson says
P B,
I was previously achieving that rough backtest using a negative cash position to simulate borrowing.
I had the same concerns though, so I actually updated the post using my own data series I created in an attempt to more accurately simulate how these leveraged ETFs – which are relatively new products – would have behaved historically. This uses the daily returns and accounts for the daily resetting, volatility decay, fees, and borrowing costs. This should be invariably more robust than the quick way I described previously.
In fairness, the new, accurate simulations really weren’t too far off from the quick and dirty method of using the negative cash position.
Mirco says
Hello John.
That is very interesting. It would be very nice if you can explain more precisely how to Do it.
I tried but i dont know where to put the 2 or 3x allocation and where to put the negative Cash Position.
Please explain ….
John Williamson says
Hey Mirco,
I actually just updated the post to use some new simulated data series I created to more accurately illustrate the historic behavior of the leveraged ETFs discussed. If you want a rough, quick, dirty backtest that simulates borrowing money and keeping that leverage ratio constant, your input in Portfolio Visualizer for 3x would look like this:
Mirco says
THX
John Williamson says
Hey Tommy,
Wanted to let you know I updated the post using my own data series I created in an attempt to more accurately simulate how these leveraged ETFs – which are relatively new products – would have behaved historically. This should be invariably more robust than the quick way I described previously to get a rough idea by simulating borrowing using the negative cash position.
Edric says
Great article you got there. I learnt a lot from this.
I did a backtest comparing the S&P500 PE and ETF VTI with excel. Here are some of my insight:
1) when S&P500 PE is below 24, increase the portion of stock ETF, decrease the portion of bond ETF
2) when S&P500 PE is above 24, decrease the portion of stock ETF, increase the portion of bond ETF
The typical AWF without leverage starting from 10,000 in 1 Jan 2005 got to 20,077 in 31 Dec 2019,
return 100.77%. Maximum drawdown -13.9%.
The adjusted portfolio, starting from 10,000 in 1 Jan 2005 got to 25,275 in 31 Dec 2019.
return 152.7%. Maximum drawdown -12.46%
In general, found that one could buy more stock ETF when stock is undervalue, and buy more bond ETF when stock is overvalue.
What do you think about it? Is it an good idea to adjust the portfolio % accordingly based on PE?
John Williamson says
Thanks Edric! Glad you could learn a thing or two. And thanks for your comment.
I don’t employ or advise market timing, and I definitely wouldn’t use P/E to try to time the market. The P/E ratio of an index doesn’t really tell the same information as the P/E of an individual stock.
The long-term average P/E of the market is essentially random. There’s no reason it has historically settled around 15. There has been no reversion to the mean. Thus there’s no reason why that average – historically 15 – can’t decrease or increase. P/E or P/S of an index may tell you something about expected future returns (comparatively), but it can’t tell you if it is “overvalued.”
What’s special about your use of the number 24 for the inflection point? Sounds like it may be a case of overfitting. It’s easy to construct timing models that outperformed the market historically, but they almost always completely flop upon implementing them in the real world with real dollars.
If I were to use market timing, it would be based on volatility – probably a target volatility or inverse volatility strategy.
Gavin says
Great post!
How did you model the last portfolio going back to 1992 when many of the leveraged ETFs you’re backtesting didn’t exist then? For example, UTSL started trading in 2017.
John Williamson says
Thanks Gavin! In PV I just used unleveraged versions of the ETFs (e.g. XLU) at 300% with an offsetting negative cash position to simulate constant leverage.
John Williamson says
Hey Gavin,
Wanted to let you know I updated the post using my own data series I created in an attempt to more accurately simulate how these leveraged ETFs – which are relatively new products – would have behaved historically. This should be invariably more robust than the quick way I described previously to get a rough idea by simulating borrowing using the negative cash position.
Antonio says
How viable is to do this strategy through Options on Etfs ?
John Williamson says
Probably not very viable. You could use LEAPs on SPY for the U.S. stocks position, but I doubt LEAPs would be available on all the assets. If they are, spreads would likely be wide. You could possibly continually roll options with short expiration dates, but that’d be cumbersome. But even short-term options on sectors like Utilities probably wouldn’t have great liquidity. And then I’d imagine options on bond ETFs would be tricky too. I’ve never looked into it. Futures would probably be a better route for derivatives.
Justin says
Hi John, thanks for the article!
May I know if the returns are before or after fees?
Would the return of leveraged funds drop because of the higher expense ratios?
Thanks,
Justin
John Williamson says
Hey Justin, returns shown are after fees.
Xavier says
Hi John, thank you for the great posts on this website! I have shared your content with a lot of friends who all think that your work is amazing.
I want to implement both this approach and the Hedgefundie tactic. I am however worried about the current pricing of TMF and the fact that both TMF and UPRO could take a solid dive at the same time in the near future… TMF did not retrace its gains from March but UPRO is back to ATH.
I get that you say don’t time the market but this situation doesn’t appear normal.
Any thoughts?
John Williamson says
Xavier, thanks so much for the kind words. Really glad you and your friends have found the content useful.
I would ask what makes you worried about the current pricing of TMF and what makes you think both TMF and UPRO could crash together in the near future? If you have these reservations, leverage – especially in the form of the Hedgefundie strategy – may not be for you. The leveraged All Weather Portfolio would definitely be the more “conservative” of the two leveraged strategies with its comparatively greater diversification. Keep in mind TMF is basically just crash insurance in the Hedgefundie strategy.
The market doesn’t really ever have a “normal” environment. For the most part, it is a memoryless random walk. This is why black swan events are named such, and is why Dalio created the All Weather Portfolio in the first place.
Antonio says
What would the version of AWx2 that you would use ? I mean,if you could specify the allocation,thanks
John Williamson says
Hey Antonio. I outlined a 2x version using Utilities here.
Antonio says
Thanks a lot. I’ve sent you a Linkedin messagge
PS: I suspect you’re the same Tyler from Portfoliocharts.Am I right ? 🙂
John Williamson says
No, that’s not me.
a man from Korea says
Thank you for your kind explanation of leverage all weather portfolio.
I have 2 questions.
1. For rebalancing, how about use band rebalancing?
because it is very easy to monitor our portfolio, so I can handle adjust portfolio weights if I know optimized band value. Or just re-balancing each quarter enough to adjust portfolio?
2. If I try to invest all country not only USA, How about DZK and EDC also?
Thanks in advance.
John Williamson says
1. Rebalancing strategy is up to you. Bands wouldn’t be a bad idea. Potentially a bit more work.
2. International diversification on the stocks side would be a solid idea but apparently there are some issues with DZK.
a man from Korea says
Oh, thanks for rapid answers.
Do you have an idea of international diversification for leverage all weather portfolio?
John Williamson says
I haven’t explored international (ex-US) diversification with leveraged funds and probably won’t, considering the potential tracking and liquidity issues of those particular ETF products at this time. EDC may be an okay idea at a small allocation but is still fairly highly correlated with the US market.
Andy says
Hmm, now what about using the VIX as a proxy/canary to determine when to go long leverage vs not….. Or perhaps emerging markets? I wondering if this would markedly improve results.
John Williamson says
Hey Andy, that would probably be a fine idea. One of the protocols discussed in the Bogleheads thread regarding the Hedgefundie Adventure was to use volatility targeting with a 1-month lookback period to know when to move more into bonds. Similar play, though likely unsuitable for novice investors wanting a lazy portfolio who don’t employ market timing.
Andy says
Thanks!
Yeah, laziness is good – but…. a bit of effort can yield some different results.
For discussion purposes I could draw attention to this interesting article https://pdfs.semanticscholar.org/8b6e/9067c2cece212f8c902653a1d85f3eb0fde0.pdf
Conceivably, adding the VIX as a hedge would help.
Andy says
Also, from my trials, replacing VTI with QQQ yields superior results thanks to the texh sector. And removing VGIT to hold just VGLT can simplify things with similar results.
In fact, and adapted version would be
50% – QQQ
40% – TLT (or VGLT)
5% – VPU
5% – IAU (or GLD)
John Williamson says
Yea QQQ would definitely backtest better due to tech’s stellar run in recent years, as QQQ is basically a tech fund, but then these allocations completely stray away from an all-weather-type portfolio, and especially from Dalio’s prescribed All Weather Portfolio.
Hanhan Li says
Most retirement plans use a portfolio which is simply a combination of total stocks and total bonds. I wonder how does it compare with the Ray Dalio All Weather Portfolio? The main questions are
1. Do gold/commodities/utilities help at all?
2. Do non-treasury bonds (such as corporate bonds) help at all?
John Williamson says
Hanhan, thanks for your comment.
True, nothing wrong with a plain ol’ 60/40 portfolio for retirement. As we’d expect, the AWP will likely simply lower volatility and risk while sacrificing some return. I’ll let you backtest that comparison yourself.
1. In this post I discussed how these diversifiers do what I just described.
2. Corporate bonds won’t help in a diversified, long-term portfolio; I delved into that here. TIPS may be a suitable alternative for an inflation hedge, but there’s no leveraged TIPS product to my knowledge.
Adeel says
Hi John,
TMF inception date was 4/19/09 and UBT inception date was 1/19/10.
How did you backtesr the 2x and 3x AW on PV back to 2006?
Thanks
Adeel says
I just saw your comment below about using leveraged percentages on unleveraged funds, but PV is still wanting my allocations to add up to 100%.
John Williamson says
Negative cash position to offset.
John Williamson says
Hey Adeel,
Wanted to let you know I updated the post using my own data series I created in an attempt to more accurately simulate how these leveraged ETFs – which are relatively new products – would have behaved historically. This should be invariably more robust than the quick way I described previously to get a rough idea by simulating borrowing using the negative cash position.
Mike Emami says
Hi John,
Thanks for this article. I was so happy when I saw your article. Since it gave me validation to my thoughts replacing commodities with utility ETF (XLU). I have also, changed the VTI to half and half. That means I did 15% QQQ, 15% VTI. I am also, trying two different portfolio to see which one would be better for bonds, I have one with combinations of IEF, and TLT (VGLT) and one with BLV, and BIV. The only thing that I wasn’t in agreement is to call this portfolio Lazy. Since I think the best way to build wealth is over time with power of compounded interest.
John Williamson says
Sounds good, Mike.
Personally I wouldn’t want to overweight tech that much, since VTI is already roughly 1/4 tech, but I see many investors with tech tilts like that.
Good luck with your test on different bonds. I avoid corporate bonds and use treasury bonds only in my diversified accounts.
Also, the unleveraged All Weather Portfolio definitely fits the “lazy portfolio” designation, as the investor does not need to change allocations as time passes. The leveraged versions are arguably a different story, requiring more frequent rebalancing and monitoring.