The All Weather Portfolio Guide, for everyone, for free

I have been fascinated by Risk Parity since the day I read about it. Combining uncorrelated assets, where each contributes in the same way to the portfolio’s overall volatility, is the best way to approach the capital allocator problem, at least to me.

Nice theory bruv, how do you implement it?

Here comes the pain. The two key elements to solve the above problem are the definition of each asset volatility and their covariance matrix or, as Wiki puts it, how we define assets into classes, risk, allocate risk within asset classes, forecast future risk and implement exposure to risk.

Do we use historical volatility? If so, how long in the past should we go? Or it is better an estimated value? Do we want volatility to be a static parameter or do we want our portfolio to reflect the current market situation? Now repeat the same process for the covariance matrix. As you realize, everyone is talking about Risk Parity as a well-defined strategy whereas Risk Parity is in the eye of the beholder.

In June 2019, Wilshire launched some Risk Parity Indexes. Guess when those indexes experienced the worst drawdowns?

This is just to say that solving for those parameters is not that straightforward and the reason why I am always on the lookout for possible solutions.

The other day I was watching listening to Mr.RIP and someone asked him “hey, did you watch Pietro Michelangeli’s video about the All Weather Portfolio?”. My brain reacted like this:

In my defense, 1) I listen to RIP at 1.5x while working 2) both strategies got famous because of Ray Dalio.

[little aside on RIP, it is really a pity that he stopped producing English content…at least for you, not for me, ’cause I understand Italian. The guy generates an absurd quantity of great content, and I am not even considering anything finance/markets-related because I normally skip those parts (I think you can understand why)]

Pietro Michelangeli is another Italian YouTuber who talks about finance 101. I know him because months ago he did a video with RIP, and that was it: I watched one of his videos, realized it was not for me and moved on.

At this point of the story, in my mind I am still convinced Pietro is going to talk about Risk Parity; and since his audience is pretty basic, I was curious to see if he managed to ‘crack the code’ and found a simple way to parametrize volatility and covariance. So I went and watched the video.

Unfortunately, it was just a short teaser for a pdf report he wrote on how to build a different version of the All Weather Portfolio. But wait a minute…

Would you buy a tutorial from a famous chef on how to cook plain vanilla spaghetti?

I would not be able to sell a bottle of water to a person that is dying from thirst even if my life depends on it. I cannot. This is why I am fascinated by people that manage to sell even the unsellable. The All Weather Portfolio is one of the most simple portfolios out there and the rules are public: why someone would spend €79 to get the same information you can find with a free Google search?

Yes but Pietro is selling an “updated” version of the portfolio targeted at Italian investors

This part is interesting and it is the first reason why I wanted to write this post.

From now on, the underlying assumption for all considerations you will find below is that you are an investor with a long time horizon. The optimal solution for investors that need their capital in 5 years or less, if they can be called ‘investors’, requires too many hypotheses on their specific circumstances.

The EUR based investor buying USD denominated assets

This is one of my financial pet peeves and Pietro’s video pushed me to finally put some numbers behind my thesis. Even if I lived in Switzerland and now in the UK, I always considered myself a EUR-based investor, for the simple reason that I will most likely spend my retirement in a country where the EUR is the legal tender (if there will still be the EUR by then…so far, so good). My portfolio is heavily invested in USD denominated assets and, not only I never had an issue with this, it was for the most part a great choice (i.e. I got higher returns compared to the alternatives: investing only in EUR denominated assets or overlaying an FX hedge on my portfolio).

Let’s take the S&P500 as an example:

The graph has three lines:

• the return of the S&P500 for a USD-based investor, the classic image you find everywhere on internet.
• the return of the same investment but for a EUR-based investor, meaning at the end of each year the investor has gained/lost also the performance of the EUR/USD in the same period (if you are asking yourself “hey, the EUR was introduced in 1999, why your graph goes back to 1973?” congrats, you won a candy. I started my analysis in 1973 because Pietro used that time frame and Nick Maggiulli as well, more on this later. Bloomberg built a ‘retrofitted’ EUR based on the performance of its constituents in the past, mainly the Deutsche Mark).
• the return of the S&P500 hedged in EUR; this follows the unhedged version until 1999 and then is hedged. Why? Because I needed the interest rate in EUR to calculate the hedging costs and those started to trade only in 1999.

Few considerations:

• the orange line and the blue line follow themselves closely. Sometimes the blue line is higher, sometimes the orange line is. This is because we are talking about the EUR and the USD, the two biggest currencies out there (the RMB is not convertible, meaning it cannot be traded outside China, so it cannot be part of the same ranking). Their relative value has oscillated in a big range but, so far, it stayed like that, a range. The currency trading world is a world of relatives, not absolutes; as bad as your opinion can be about the USD, so far all the alternatives have proven to be at least equally bad.
• adding the EUR/USD performance to the S&P500 does not make the blue line more volatile, it just makes it different from the orange line: different tops and bottoms, roughly the same volatility (a standard deviation of 16 compared to 19 for the EUR-based investor). Look at the S&P500 negative years and the EUR/USD performance in %:

• out of 9 years, in 6 you got a better return combining the two and in 3 a…disaster. The alternative?
• if you hedge, as represented by the grey line, then you are (almost) sure to underperform both. Why? Because the cost of hedging is a function of the difference in interest rates between Europe/ECB and the US. Since the IR in the US have been almost always higher than in Europe, the hedged position performance was bogged down by this additional cost, on average 64bps/year.

Having to take the currency risk is not great but it is not a disaster either. There is a simple reason why I invest in USD-denominated assets: it is because I think they are great. And it is difficult to picture a scenario where those assets would remain great, in aggregate, but the US dollar would go to 0. We are not talking about a single stock here, like for example Samsung in South Korea.

Here is how the All Weather Portfolio would have performed for a USD-based investor and a EUR one:

I prefer another type of currency hedge, which brings me to:

Home Country Bias

The AWP is composed of only USD-denominated assets. Jack Bogle famously said that US investors should not bother investing outside the US because:

• US companies have exposure to other countries (i.e. Apple has revenues from all over the world)
• the US is the greatest country in the world

While I agree with the former, I am a bit skeptical of the latter. Credit Suisse publishes every year their Global Investment Returns Yearbook, where you can see that the US has not always been ‘on top of the world’:

It can be that in the future these slices will change again. This is why on this blog, and in my portfolio, I always praise a World Index exposure, both in stocks and bonds. It is true that bond indexes are built in a way that makes even less sense than market cap for stock indexes; yet, I still have to find a manager that reliably beat those indexes so, they are still the best we can get.

I do not know what type of change of assets Pietro is suggesting in his paper. But if the change is to move from the US to World, then again, the message is out there, has been there for a while and it’s free. I hope he is not suggesting to invest more in EUR-denominated assets just because…well, home country bias but maybe presented under a different angle. The World Index approach offers good diversification (risk management) with some momentum towards the best performers via the market cap rule. If a currency persistently over-performs, the index will automatically allocate more.

In the bond space, currency hedging has more supporters because bonds tend to have low volatility (do not look at this year 😉 ): adverse FX movements can easily wipe out one or more years of income generated by a bond portfolio. Again, if your horizon is long enough, those movements will mean revert and you save the hedging costs. I am in favor of tilting your bond portfolio towards your home currency if a) you live in a G10 country and b) the interest offered on your home currency is higher than the G10 average (see my previous post about the FX carry strategy).

So far in my analysis, I used a 15% allocation to gold instead of a 7.5% split between gold and commodities. The first reason is a practical one: the is no commodity index that goes back to the ’70s. The second is a financial one: all commodity indexes are quite bad. Storing commodities outside precious metals is expensive and some of them even have an expiry date. It is really difficult to get exposure to the ‘spot price’ of a commodity, the price you normally hear referenced in the news; this means indexes have to take an exposure with futures, where embedded storing costs can be even higher depending if that commodity price curve is in contango or backwardation. It is difficult to design a rule-based strategy, the index any passive ETF replicates, because of all these possibilities. Recently, active ETFs like $COM have provided better solutions…but they are available only to US investors.

I have always used $CRB and looking at the graph now, my sense of frustration can be explained here:

I would not blame you to stick to the OG index with 7.5% allocated to commodities, just pay attention to the index composition you choose, different providers might go very different ways.

The AWP uses commodities as a hedge against inflation. In this sense, the FX risk component should be taken care by the commodity price itself. Think about it: the value of 1gr of gold should be… 1gr of gold. It’s the EUR and USD values that move around because of their different inflation rates. Anyway, here is the price of gold in USD (white line) and EUR (blue line):

If you are based in the EU and want the white line-associated returns, you have to hedge the FX component (and pay the hedging costs); otherwise, you have to be happy with the blue line-associated returns. The recent blue line spike is due to EUR weakness against the USD, more than the gold performance itself.

The All Weather Portfolio 2.0

Is the AWP built with the best available assets to navigate the ‘four quadrants’, the four possible combinations of economic growth/decline and inflation up/down? No. And how it could be, Ray created it in 1996. Pietro is not the only one that adventured down this road. 8 months ago I wrote a post about the Cockroach Portfolio, an idea championed by Jason Buck of Mutiny Funds.

Do you remember the world eight months ago? ARKK was valued at $120, the S&P500 was comfortably trending higher and NFTs were great investments. GMO was the only shop in town talking about inflation-linked bonds, as they have been for years.

My spider-sense gets activated when I see Pietro pushing his €79 report because he added inflation-linked bonds to his version of the AWP. Great market timing as a selling point; as an investor P&L point…maybe it would have been useful a year ago?

I write about ‘permanent portfolios’ all the time here. I created TheItalianLeatherSofa Model Portfolio just a month ago, maybe Pietro does the same? It was just an un-timely discovery on my side? A coincidence?

wut??

“one of the few portfolios that outperformed the S&P500 in the long run”? Did I read it correctly?

Wait a sec, the AWP DOUBLED the S&P500? Did the world start to spin the other way around and I did not notice? How can a portfolio that allocates 55% to bonds do that?

After I climbed back to my chair, I did a quick Google and Nick Maggiulli restored my confidence in internet scammers:

Not only the AWP did not beat the S&P500, but it also did not even beat the 60/40. To be 100% sure, I re-run the numbers myself, here is the difference in nominal terms:

Did Pietro forget the S&P500 pays dividends? For sure the AWP backtest includes them so….I do not know?

(if you are wondering why my pic is so different from Nick’s, that’s the motherfu**ing inflation baby: my S&P500 compounds at 11.68% while Nick’s at 6.4%)

Conclusion

I make mistakes all the time. After I click “Publish” on a post, I immediately start to panic and think ‘OMG, what if this calc is wrong, what if I missed a point, what if I am so stupid that everything I wrote is wrong?”. I checked that video 20 times. And I am sorry (weeeeeeeell, am I?) but it has written SCAM all over it. I even feel stupid because I guess if someone 30 years ago found a way to throw 5 ETFs together and double the return of the S&P500 with half the volatility…people would be already talking about it?

My employer pays my Bloomberg but we live in a world of very good, and free, financial tools. I should set “F4 = PortfolioCharts.com” because I write it so many times. Portfolio Visualizer, Koyfin, YCharts, Composer: choose your fighter. The All Weather Portfolio looks outdated to you? Great, go out and test your own version. Are you afraid of FX risk? Test it. Do you want to tilt your portfolio to a factor or a region? Do your bloody homework.

It is not about the €79 or the Lambo Pietro will buy with them, this is the only way to build the necessary confidence to sustain periods like the one we are in right now. You can invest in the perfect portfolio but if you panic and sell at the wrong time, it was like not having it. And maybe it will help you to identify the next Pietro you will find along the road 😉

What I am reading now:

Follow me on Twitter @nprotasoni