Improving the Dalio/Robbins All-Seasons Portfolio

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
sandramjet
Posts: 442
Joined: Thu Oct 23, 2014 11:28 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by sandramjet »

This is an interesting discussion, to be sure. One question that I have not seen addressed is what is your timeline and how confident are you that you will stick to this portfolio for this timeframe? I realize you are just formulating the portfolio, but you have indicated several revisions to your positions and choices just during this conversation. At some point, you need to "set it in stone" and then stick to it for the next 30-60 years. How confident are you that you'll be able to do that, without changing it? Backtesting is fine, but it is the future (unknown) performance that ultimately counts. I'm definitely interested to see posts 15, 30 years from now that show your results!
BrianJ0101
Posts: 2
Joined: Sun Mar 05, 2017 1:59 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by BrianJ0101 »

First, thank you! This is an amazing thread and I loved all of the discussion.

I'm looking at implementing it in my accounts, but I have about 10% of my assets in a taxable account. Are any of these ETFs or funds better to run out of that account than others, or does it simply not matter? I'm not planning to add substantial new funds to this taxable account, if that should factor into it. I'm in the highest tax bracket, if that's relevant. I also have some previously taken losses that I can use to offset taxable gains, so I'm left a bit unsure as to what that picture leaves me in terms of what assets I should favor holding in the taxable account.
daffyd
Posts: 199
Joined: Sun Sep 29, 2013 11:51 pm
Location: Australia

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by daffyd »

BrianJ0101 wrote:First, thank you! This is an amazing thread and I loved all of the discussion.

I'm looking at implementing it in my accounts, but I have about 10% of my assets in a taxable account. Are any of these ETFs or funds better to run out of that account than others, or does it simply not matter? I'm not planning to add substantial new funds to this taxable account, if that should factor into it. I'm in the highest tax bracket, if that's relevant. I also have some previously taken losses that I can use to offset taxable gains, so I'm left a bit unsure as to what that picture leaves me in terms of what assets I should favor holding in the taxable account.
The index version of the equity slices, i.e. in ETF form VSS, VOE or similar, should both be reasonably tax efficient. Perhaps do a search for relevant topics, I seem to recall there's a spreadsheet floating around. The rest of the components should be held in tax advantaged space.
staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 8:40 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by staythecourse »

azanon wrote:So any reason why GSG might be optimal?
GSG follows the GSCI index. It is market weighted so energy,i.e. oil is about 75% of the index. USCI is a made up index from that fabaulous pro commodity paper from the ?Yale authors. SO USCI backtracking is just made up numbers on a hypothetical index. USCI is more active and GSG is pure passive long.

THE BIGGEST difference which I have not seen ANYONE notice except for me is that GSG invests in CERFs (which are ?5 year contracts) that pasively roll over. So that means in taxable they are not considered market to market as taxation goes. What many folks don't realize is CCF are extremely tax inefficient for the high income worker. At the end of every year tax season all (if any) gains are taxed as short term gains. So if you are at 40% federal then if they lose money you lose 100% of that money and if they win you gain only 60% of the gains.

Sort of the same as gold for the high income worker. If they lose money you lose 100%, but if they gain when it comes to selling you get taxed at collectibles (28%).

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle
staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 8:40 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by staythecourse »

Not sure if this was discussed as I did not read all the posts, but Harry Browne was asked when he was alive why not extended duration treasuries. His answer was they are TOO volatile in his equal protection portfolio.

I think the role of EDV would be if one has limited tax deferred accounts then the can hold less % of assets protecting against deflation by holding EDV vs. TLT.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle
gtwhitegold
Posts: 673
Joined: Fri Sep 21, 2012 1:55 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by gtwhitegold »

I personally am doing something similar to this with my portfolio, but in place of commodities I'm using AQR's QMHIX Managed Futures High Volatility Fund and QSPIX Style Premia Fund. I agree with Larry Swedroe that commodity futures are in contango too much to be profitable now and am using other sources of diversification. I hold them in my Fidelity IRA accounts.

http://www.etf.com/sections/index-inves ... nopaging=1

I would also consider using Vanguard's internal quant funds vice their externally managed active funds. Probably lower costs and less idiosyncratic risk.

viewtopic.php?f=10&t=206890
llamallama
Posts: 1
Joined: Mon May 01, 2017 8:03 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by llamallama »

azanon wrote:Another portfolio revision (inspired by Tag :mrgreen: )

15% Vanguard Selected Value Fund (VASVX) 0.39%
15% Vanguard International Explorer Fund (VINEX) 0.42%
10% Vanguard Emerging Markets Government Bond ETF (VWOB) 0.34%
15% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
30% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% PowerShares Optimum Yield Diversified Commodity Strategy (PDBC) 0.60%
7.5% iShares Gold Trust (IAU) 0.25%

> Swapped LTPZ and EDV for VAIPX and VLGSX, but dropped total bond exposure by 5% (to use on commodities). Despite the 5% drop, that's much longer composite duration exposure compared to what I had.

> Dropped VGPMX per tag's comments, and reinserted commodities (same 7.5% mix of commodities and gold as Dalio gave Robbins). I need more raw commodities (15% vs. 10%) since VGPMX is quasi leveraged. Using PDBC instead of USCI for lower cost, and some very high early reviews/endorsements of PDBC. Again, I imagine it really is a crapshoot which commodities ETF is best though. Just pick one.
@azanon Thanks for the great thread. Have you further tweaked your modified "all weather" portfolio? I see that the main difference between this allocation and the one in Tony Robbins book (Master the Money Game) is in the bonds portion. In the book, it's 40% 30-year bonds, 15% 10-year bonds. Am I right in thinking your allocation has some added leveraging effect? Also I'm curious why there's no real estate in the all-weather portfolio?
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

llamallama wrote:
azanon wrote:Another portfolio revision (inspired by Tag :mrgreen: )

15% Vanguard Selected Value Fund (VASVX) 0.39%
15% Vanguard International Explorer Fund (VINEX) 0.42%
10% Vanguard Emerging Markets Government Bond ETF (VWOB) 0.34%
15% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
30% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% PowerShares Optimum Yield Diversified Commodity Strategy (PDBC) 0.60%
7.5% iShares Gold Trust (IAU) 0.25%

> Swapped LTPZ and EDV for VAIPX and VLGSX, but dropped total bond exposure by 5% (to use on commodities). Despite the 5% drop, that's much longer composite duration exposure compared to what I had.

> Dropped VGPMX per tag's comments, and reinserted commodities (same 7.5% mix of commodities and gold as Dalio gave Robbins). I need more raw commodities (15% vs. 10%) since VGPMX is quasi leveraged. Using PDBC instead of USCI for lower cost, and some very high early reviews/endorsements of PDBC. Again, I imagine it really is a crapshoot which commodities ETF is best though. Just pick one.
@azanon Thanks for the great thread. Have you further tweaked your modified "all weather" portfolio? I see that the main difference between this allocation and the one in Tony Robbins book (Master the Money Game) is in the bonds portion. In the book, it's 40% 30-year bonds, 15% 10-year bonds. Am I right in thinking your allocation has some added leveraging effect? Also I'm curious why there's no real estate in the all-weather portfolio?
> Thanks, I had a lot of fun working on this, and really feel that something like this is closer to a Bridgewater do-it-yourself All-Weather "buy-and-hold" than the Robbins portfolio.

> I think you could add REITs if you want to, and not really harm the portfolio too much (Appian Road uses a 2% position in REITs). If I were to add REITs, I'd probably go with a global, low-cost REIT ETF like REET and, more importantly, make room for it from the equities. So maybe just a 5% position in REET, and drop the two equity funds to 12.5% each. I think capping equities at 30% is really important to the Dalio style portfolio.

> The only improvement I'd have is I'd now probably opt for BCI for commodities instead of PDBC. BCI was recently launched at a super-low cost of 0.29% ER. It's very small at the moment (only about 2.5 million), but the trading spread is relatively small anyway.

> Yes, EDV is allowing a pseudo leveraging effect, which effectively frees up some portfolio percentage to use on other asset classes mentioned in the Bridgewater All-Weather paper, such as LT TIPS. So a smaller percentage of EDV can substitute for a larger percentage of, say, Vanguard LT treasuries ETF or fund.

> Just for kicks, i calculated a new weighted cost of the portfolio with a 5% position in REET (dropping the other 2 equity funds to 12.5%), and changing the commodities to BCI. Comes out to 0.25% That's pretty cheap, considering it has some active funds and commodities!
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

15% Vanguard Mid-Cap Growth ETF (VOT) 0.07%
15% Vanguard International Explorer Fund (VINEX) 0.42% or VSS 0.13%
10% Vanguard Emerging Markets Government Bond ETF (VWOB) 0.34%
15% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
30% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% ETFS Bloomberg All Commodity Strategy (BCI) 0.29%
7.5% iShares Gold Trust (IAU) 0.25%

> Wanted to update the portfolio with latest proposed changes

> I changed Mid-Cap Value to Mid-Cap Growth. It dawned on me from the All-Weather Paper, that that segment of investing climate is for when there's higher growth than expected, or a growing economy. So as I thought about it more, I believe "growth stocks" should contrast better, on average, with the "defensive" portions of this portfolio. As an example here, I"m thinking of the 2000-2003 climate. So since the goal of the portfolio is to improve upon the robbins all-seasons portfolio, I think it's more important to stick to that approach than to, say, prefer value because there is research showing that value can outperform growth over all climates. It's about how it performs in this particular portfolio, not how they perform in isolation vs. each other.

>I added VSS or VINEX - Vanguard labels both as small/mid growth, so there's an active or index choice depending on which you prefer.

>BCI added for commodities (replacing PDBC) cause it is SO much cheaper than the alternatives.
mfng
Posts: 23
Joined: Sat Apr 29, 2017 9:06 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by mfng »

Thanks to everyone for the really interesting discussion on this thread. Can anyone walk me through how we would arrive at these weights using actual returns from previous years? It seems to me that volatility of the various asset classes couldn't possibly be constant over time (although there are good theoretical reasons why stocks might always have higher vol than bonds, for example). Does that mean that the ideal implementation of this strategy would involve recalibrating these weights from time to time, in contrast to thinking of the weights as fixed based on a single multi-decade window of average vol?
MoneyRider
Posts: 1
Joined: Tue Jun 06, 2017 2:48 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by MoneyRider »

What a fantastic discussion! Thanks all.

I stumbled on this topic from a Google search and wanted to ask what may be a naive question. In researching this strategy (and reading the white paper), it doesn't seem like much weighting is given to the probabilities of each market situation.

For example, it is more probable that we will be in an inflationary environment than a deflationary one. This isn't a terribly bold statement but I understand that in the purist form it's not guaranteed. It is possible that in the next 50 years we'll have more deflation than inflation but not probable. The fed target of 2%, inflationary pressures of government spending, yadda yadda yadda. Therefore, shouldn't there be an adjustment of risk compensation based on the probability of each of the four market conditions stated in the white paper?

Everyone will have their own interpretation of what the relative probability is, but shouldn't it be baked into the cake?
dore
Posts: 107
Joined: Tue Jan 08, 2008 5:27 pm
Location: Boise, ID

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by dore »

MoneyRider wrote:What a fantastic discussion! Thanks all.

I stumbled on this topic from a Google search and wanted to ask what may be a naive question. In researching this strategy (and reading the white paper), it doesn't seem like much weighting is given to the probabilities of each market situation.

For example, it is more probable that we will be in an inflationary environment than a deflationary one. This isn't a terribly bold statement but I understand that in the purist form it's not guaranteed. It is possible that in the next 50 years we'll have more deflation than inflation but not probable. The fed target of 2%, inflationary pressures of government spending, yadda yadda yadda. Therefore, shouldn't there be an adjustment of risk compensation based on the probability of each of the four market conditions stated in the white paper?

Everyone will have their own interpretation of what the relative probability is, but shouldn't it be baked into the cake?
Japan has been in a near deflationary environment for 25 years. Ray Dalio's all-weather risk parity portfolio doesn't make any assumptions regarding the probability of future economic conditions, which is why it's called the all-weather portfolio.
BrianJ0101
Posts: 2
Joined: Sun Mar 05, 2017 1:59 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by BrianJ0101 »

azanon wrote: Fri May 05, 2017 12:35 pm
llamallama wrote:
azanon wrote:Another portfolio revision (inspired by Tag :mrgreen: )

15% Vanguard Selected Value Fund (VASVX) 0.39%
15% Vanguard International Explorer Fund (VINEX) 0.42%
10% Vanguard Emerging Markets Government Bond ETF (VWOB) 0.34%
15% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
30% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% PowerShares Optimum Yield Diversified Commodity Strategy (PDBC) 0.60%
7.5% iShares Gold Trust (IAU) 0.25%

> Swapped LTPZ and EDV for VAIPX and VLGSX, but dropped total bond exposure by 5% (to use on commodities). Despite the 5% drop, that's much longer composite duration exposure compared to what I had.

> Dropped VGPMX per tag's comments, and reinserted commodities (same 7.5% mix of commodities and gold as Dalio gave Robbins). I need more raw commodities (15% vs. 10%) since VGPMX is quasi leveraged. Using PDBC instead of USCI for lower cost, and some very high early reviews/endorsements of PDBC. Again, I imagine it really is a crapshoot which commodities ETF is best though. Just pick one.
@azanon Thanks for the great thread. Have you further tweaked your modified "all weather" portfolio? I see that the main difference between this allocation and the one in Tony Robbins book (Master the Money Game) is in the bonds portion. In the book, it's 40% 30-year bonds, 15% 10-year bonds. Am I right in thinking your allocation has some added leveraging effect? Also I'm curious why there's no real estate in the all-weather portfolio?
> Thanks, I had a lot of fun working on this, and really feel that something like this is closer to a Bridgewater do-it-yourself All-Weather "buy-and-hold" than the Robbins portfolio.

> I think you could add REITs if you want to, and not really harm the portfolio too much (Appian Road uses a 2% position in REITs). If I were to add REITs, I'd probably go with a global, low-cost REIT ETF like REET and, more importantly, make room for it from the equities. So maybe just a 5% position in REET, and drop the two equity funds to 12.5% each. I think capping equities at 30% is really important to the Dalio style portfolio.

> The only improvement I'd have is I'd now probably opt for BCI for commodities instead of PDBC. BCI was recently launched at a super-low cost of 0.29% ER. It's very small at the moment (only about 2.5 million), but the trading spread is relatively small anyway.

> Yes, EDV is allowing a pseudo leveraging effect, which effectively frees up some portfolio percentage to use on other asset classes mentioned in the Bridgewater All-Weather paper, such as LT TIPS. So a smaller percentage of EDV can substitute for a larger percentage of, say, Vanguard LT treasuries ETF or fund.

> Just for kicks, i calculated a new weighted cost of the portfolio with a 5% position in REET (dropping the other 2 equity funds to 12.5%), and changing the commodities to BCI. Comes out to 0.25% That's pretty cheap, considering it has some active funds and commodities!
The one thing that stopped me from previously looking at adding a REIT as part of this portfolio was that I feel like owning a home already gives me a very large exposure to real estate. Granted, it's residential and extremely narrow, and not at all income/cash flow producing, but from an exposure perspective, it makes me not want to get additional portion of my assets in real estate.

Does that logic make sense? It was only in my most recent review of my total investments that I started thinking about my home in the equation. I had previously owned and liked REITs, and gotten good returns using them, but the All Seasons approach really forces me to think more defensively about balance AND returns, rather than solely chasing returns (and sometimes missing the mark).
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

BrianJ0101 wrote: Tue Aug 29, 2017 5:56 pm
azanon wrote: Fri May 05, 2017 12:35 pm
llamallama wrote:
azanon wrote:Another portfolio revision (inspired by Tag :mrgreen: )

15% Vanguard Selected Value Fund (VASVX) 0.39%
15% Vanguard International Explorer Fund (VINEX) 0.42%
10% Vanguard Emerging Markets Government Bond ETF (VWOB) 0.34%
15% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
30% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% PowerShares Optimum Yield Diversified Commodity Strategy (PDBC) 0.60%
7.5% iShares Gold Trust (IAU) 0.25%

> Swapped LTPZ and EDV for VAIPX and VLGSX, but dropped total bond exposure by 5% (to use on commodities). Despite the 5% drop, that's much longer composite duration exposure compared to what I had.

> Dropped VGPMX per tag's comments, and reinserted commodities (same 7.5% mix of commodities and gold as Dalio gave Robbins). I need more raw commodities (15% vs. 10%) since VGPMX is quasi leveraged. Using PDBC instead of USCI for lower cost, and some very high early reviews/endorsements of PDBC. Again, I imagine it really is a crapshoot which commodities ETF is best though. Just pick one.
@azanon Thanks for the great thread. Have you further tweaked your modified "all weather" portfolio? I see that the main difference between this allocation and the one in Tony Robbins book (Master the Money Game) is in the bonds portion. In the book, it's 40% 30-year bonds, 15% 10-year bonds. Am I right in thinking your allocation has some added leveraging effect? Also I'm curious why there's no real estate in the all-weather portfolio?
> Thanks, I had a lot of fun working on this, and really feel that something like this is closer to a Bridgewater do-it-yourself All-Weather "buy-and-hold" than the Robbins portfolio.

> I think you could add REITs if you want to, and not really harm the portfolio too much (Appian Road uses a 2% position in REITs). If I were to add REITs, I'd probably go with a global, low-cost REIT ETF like REET and, more importantly, make room for it from the equities. So maybe just a 5% position in REET, and drop the two equity funds to 12.5% each. I think capping equities at 30% is really important to the Dalio style portfolio.

> The only improvement I'd have is I'd now probably opt for BCI for commodities instead of PDBC. BCI was recently launched at a super-low cost of 0.29% ER. It's very small at the moment (only about 2.5 million), but the trading spread is relatively small anyway.

> Yes, EDV is allowing a pseudo leveraging effect, which effectively frees up some portfolio percentage to use on other asset classes mentioned in the Bridgewater All-Weather paper, such as LT TIPS. So a smaller percentage of EDV can substitute for a larger percentage of, say, Vanguard LT treasuries ETF or fund.

> Just for kicks, i calculated a new weighted cost of the portfolio with a 5% position in REET (dropping the other 2 equity funds to 12.5%), and changing the commodities to BCI. Comes out to 0.25% That's pretty cheap, considering it has some active funds and commodities!
The one thing that stopped me from previously looking at adding a REIT as part of this portfolio was that I feel like owning a home already gives me a very large exposure to real estate. Granted, it's residential and extremely narrow, and not at all income/cash flow producing, but from an exposure perspective, it makes me not want to get additional portion of my assets in real estate.

Does that logic make sense? It was only in my most recent review of my total investments that I started thinking about my home in the equation. I had previously owned and liked REITs, and gotten good returns using them, but the All Seasons approach really forces me to think more defensively about balance AND returns, rather than solely chasing returns (and sometimes missing the mark).
> Yes, that makes sense. I wouldn't recommend trying to insert REITs into the formula, but just made a suggestion how to do that if someone really wanted to. I'd rather just use half "aggressive growth" US, and half "growth/aggressive growth" International (including EM).

.........

> I propose one other change to my continuously revised portfolio; I think using local currency EM bonds instead of currency hedged is more appropriate for this portfolio. I was originally swayed go to just go with VWOB (which is currency hedged) because that's what appianroad.com was using, but I got to digging a little deeper into why Bridgewater was picking this seemingly odd asset class. Since I last posted, I've found several articles (such as this one: https://www.researchaffiliates.com/docu ... rrency.pdf ) which address why local currency EM has proven to respond well to unexpected inflation spikes here in the US. So I'm guessing that's why that specific asset class, including local currency, made it into the all-weather design. Also, since there's only 15% foreign currency exposure without it, this will bring that up a bit more for better global balance.

The most well established, and reasonably priced, local currency EM bond fund appears to be EMLC. So unless I'm missing something here, that appears to be a better choice for this portfolio's purpose than VWOB.

Revised portfolio:
25% Vanguard Mid-Cap Value ETF (VOE) 0.07%
10% Market Vectors Emerging Mkts Local ETF (EMLC) 0.44%
20% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
30% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% ETFS Bloomberg All Commodity Strategy (BCI) 0.29%
7.5% iShares Gold Trust (IAU) 0.25%

.............

(9/11/2017) - Made one other edit to the portfolio, and actually bought this for my non-TSP portfolio. I dropped the stock percentage to 25% (from 30%), and raised EDV to 20% (from 15%). I realized I was using the original 30% equities in the Robbins portfolio, but was substituting far more aggressive stocks with higher beta, but wasn't really offsetting that on the bond side. Since I'm aiming for a portfolio that's fully agnostic to the investing climate, I think this is a better balance. I'm also dropping VSS since it's a vanilla International Small Cap, but Explorer is intentionally a small/mid-cap international growth fund. So I think I'm done. I certainly prefer this version over the one Dalio gave Robbins.

(9/16/17) - Another revision; Dropping VINEX, and switching to Mid-cap Value for entire stock position. Several reasons: 1. It's cheaper 2. Per portfolio visualizer, most all foreign stock has much closer correlation to other parts of this portfolio for reasons I won't go into, so US stock seems to work better 3. Based on my analysis of risk balancing using the "all weather story", the stock position should approximate the weight-adjusted volatility of the TIPS position. Again, using portfolio visualizer, VOE matches the closest. 4. To my surprise, using portfolio visualizer, value stocks seem to have a lower composite correlation to the other portfolio components than growth stock. So I'm swapping back to value. 5. The local currency EM bond will still give some foreign currency exposure.

Now the back-tested risk/return of this portfolio is incredible (you have to substitute for BCI and EMLC). Also the correlation to the US stock market is a very low 0.18, so it's almost just outright uncorrelated to the US market.
Last edited by azanon on Sat Sep 16, 2017 9:34 am, edited 2 times in total.
ThrustVectoring
Posts: 771
Joined: Wed Jul 12, 2017 2:51 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by ThrustVectoring »

dore wrote: Tue Jun 06, 2017 8:49 pm
MoneyRider wrote:What a fantastic discussion! Thanks all.

I stumbled on this topic from a Google search and wanted to ask what may be a naive question. In researching this strategy (and reading the white paper), it doesn't seem like much weighting is given to the probabilities of each market situation.

For example, it is more probable that we will be in an inflationary environment than a deflationary one. This isn't a terribly bold statement but I understand that in the purist form it's not guaranteed. It is possible that in the next 50 years we'll have more deflation than inflation but not probable. The fed target of 2%, inflationary pressures of government spending, yadda yadda yadda. Therefore, shouldn't there be an adjustment of risk compensation based on the probability of each of the four market conditions stated in the white paper?

Everyone will have their own interpretation of what the relative probability is, but shouldn't it be baked into the cake?
Japan has been in a near deflationary environment for 25 years. Ray Dalio's all-weather risk parity portfolio doesn't make any assumptions regarding the probability of future economic conditions, which is why it's called the all-weather portfolio.
You can't just not make any assumptions. Like, not extrapolating past inflation environments into your future risk model is an assumption - specifically, that the market expectation for future inflation is wrong, and that protecting against deflation is cheaper than the market is pricing it.
Current portfolio: 60% VTI / 40% VXUS
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

ThrustVectoring wrote: Thu Aug 31, 2017 5:38 pm
dore wrote: Tue Jun 06, 2017 8:49 pm
MoneyRider wrote:What a fantastic discussion! Thanks all.

I stumbled on this topic from a Google search and wanted to ask what may be a naive question. In researching this strategy (and reading the white paper), it doesn't seem like much weighting is given to the probabilities of each market situation.

For example, it is more probable that we will be in an inflationary environment than a deflationary one. This isn't a terribly bold statement but I understand that in the purist form it's not guaranteed. It is possible that in the next 50 years we'll have more deflation than inflation but not probable. The fed target of 2%, inflationary pressures of government spending, yadda yadda yadda. Therefore, shouldn't there be an adjustment of risk compensation based on the probability of each of the four market conditions stated in the white paper?

Everyone will have their own interpretation of what the relative probability is, but shouldn't it be baked into the cake?
Japan has been in a near deflationary environment for 25 years. Ray Dalio's all-weather risk parity portfolio doesn't make any assumptions regarding the probability of future economic conditions, which is why it's called the all-weather portfolio.
You can't just not make any assumptions. Like, not extrapolating past inflation environments into your future risk model is an assumption - specifically, that the market expectation for future inflation is wrong, and that protecting against deflation is cheaper than the market is pricing it.
For better or worse, the all-weather strategy design puts equal risk weightings to each of the 4 possible scenarios discussed. So, if one knew for certain one of the 4 scenarios were more likely, then the design is flawed to the extent that they are not equal.

I think the strategy focuses less on obsessing over which scenario is more likely, and much more on simply ensuring that there are at least some positions in the portfolio that should theoretically do well in any investing climate, and also that all of the positions, over the very long term, should have a higher expected return than cash. Even gold should match inflation, which is as we know quite a bit higher than the 0% nominal return of cash.

So, in theory, this portfolio should have outstanding Sharpe/Sortino rations, with returns hopefully meeting or exceeding a traditional 60/40.

This discussion reminds me of one of William Bernstein's short books, Deep Risk. I believe in that book, he did attempt to weight the likelihood of each scenario and resultantly advised overweighting whatever asset classes provided protection from what he saw as the more likely scenarios. So if you find that approach more sensible, then this portfolio probably wouldn't be for you. Me? I don't know which is more likely, so I tend to be more comfortable with the all-weather.
CDNnewbie
Posts: 1
Joined: Mon Sep 18, 2017 9:26 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by CDNnewbie »

Thank you for this! I really enjoyed the read. Are there any Canadians on the board who have found a way to successfully implement the portfolio? I am struggling.

If I switch currencies I get slaughtered right off the top, and can't predict if currencies would be in my favour when I want to switch back. So, I am trying to use as many Canadian funds to mimic the portfolio that I can, which is not very hard for the US stock ETFs - not as much choice so less possibility for nuance but still very possible - but I don't have or can't find options for the TIPS portion.

About 8% of my portfolio is currently in USD, so I can use that to buy the TIPS, biasing towards the LT bonds. Based on this board I don't think that is nearly enough. So, do settle for only 8% TIPS, switch more currency into USD to buy more, or do I try and make up the balance with the Canadian version - Real Return Bonds? There are only 2 choices that I can find, and I am wondering if that is problematic.

Options are ZRR or XRB for Real Return CAD long-term bonds - XRB seems on average to hold longer bonds by about 3 years but MER is higher.

15% VUS CAD hedged US Total Market MER .16 (or VSP US S&P? MER .08)
7.5% VEE CAD hedged Emerging Markets MER .24
7.5% VDU CAD hedged Developed World Excluding US MER .21
7.5% ZMT CAD Base Metals commodity MER .63
7.5% CGL CAD Gold MER .55

Option 1 of bond portion
33% ZTL CAD hedged US long-term bonds MER .23
14% VBU CAD hedged US intermediate bonds MER .22
7% LTPZ USD TIPS long-term bonds MER .20
1% SCHP USD TIPS intermediate bonds MER .05

Option 2 of bond portion
10% ZTL CAD hedged US longterm bonds MER .23
15% VBU CAD hedged US intermediate bonds MER .22
8% LTPZ USD TIPS long-term bonds MER .2
22% XRB CAD Real Return long-term Canadian bonds

I would be grateful to anyone who is interested, and if there are any Canadians on this board that have had success, I would love to know what you did. Thank you so much in advance! I am having trouble pulling the trigger ...
David Scubadiver
Posts: 522
Joined: Thu Mar 24, 2016 8:40 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by David Scubadiver »

unclescrooge wrote: Sun Dec 25, 2016 12:56 pm
azanon wrote:
josh1130 wrote:
livesoft wrote:If you work your way up to 31 funds, then you can call it the Baskins/Robbins All-Flavors Portfolio.
8-)
Jokes aside, I bet we're no more than 5-10 years away from being able to "construct" our own custom portfolio with infinite amounts of very small allocations (fractional shares) to many asset classes, by using roboadvisor-like tools to do it with. So it'd work like a buffet of, say, 100 different ETFs and you could add whatever percentages of each to your own personal mix at, say, Betterment (for a small, nominal fee, of course).

Seriously, mark it down. This is coming. 10 years max.
It's already here. You just have to pay $30/month for it.
Where can you do this for $30/month? If that we included rebalancing and any sort of tax management (like minimizing gains when a sale is made) that would be great. M1 was the only service I was aware of (granted I haven’t been searching) and it charged 0.15% AUM for portfolios over 100k.
User avatar
unclescrooge
Posts: 6265
Joined: Thu Jun 07, 2012 7:00 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by unclescrooge »

David Scubadiver wrote: Fri Oct 13, 2017 7:04 am
unclescrooge wrote: Sun Dec 25, 2016 12:56 pm
azanon wrote:
josh1130 wrote:
livesoft wrote:If you work your way up to 31 funds, then you can call it the Baskins/Robbins All-Flavors Portfolio.
8-)
Jokes aside, I bet we're no more than 5-10 years away from being able to "construct" our own custom portfolio with infinite amounts of very small allocations (fractional shares) to many asset classes, by using roboadvisor-like tools to do it with. So it'd work like a buffet of, say, 100 different ETFs and you could add whatever percentages of each to your own personal mix at, say, Betterment (for a small, nominal fee, of course).

Seriously, mark it down. This is coming. 10 years max.
It's already here. You just have to pay $30/month for it.
Where can you do this for $30/month? If that we included rebalancing and any sort of tax management (like minimizing gains when a sale is made) that would be great. M1 was the only service I was aware of (granted I haven’t been searching) and it charged 0.15% AUM for portfolios over 100k.
Open a retail account at folio.
thejimmysmith
Posts: 4
Joined: Sun Oct 22, 2017 3:02 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by thejimmysmith »

Great post. I really came into investing after I read Robbins first book and learned about All Seasons then started doing an investing deep dive. I'm taking my time going through this thread and I apologize if this has already been answered but why no real estate investment trusts in this portfolio?
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

thejimmysmith wrote: Mon Oct 23, 2017 7:32 am Great post. I really came into investing after I read Robbins first book and learned about All Seasons then started doing an investing deep dive. I'm taking my time going through this thread and I apologize if this has already been answered but why no real estate investment trusts in this portfolio?
Without re-reading the thread, a REITs question similar to that did come up. But in short and summary, I tried to reconstruct/estimate a static "all-weather" portfolio based upon the actual paper that's publicly available at Bridgewater, linked in the original post of this thread which I believe is more accurate than the Robbins one. That paper was very revealing as to what asset classes to use and gave strong clues as to how to balance those. Discussion of REITs simply doesn't come up in that paper so I wouldn't have had any basis for their inclusion in my portfolio.
dabears5496
Posts: 11
Joined: Mon Feb 13, 2017 10:35 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by dabears5496 »

@azanon

I'm trying to replicate your updated All Weather Portfolio. I've been trying to replicate this portfolio and you all did amazing work. So I'll use yours!

Revised portfolio:
25% Vanguard Mid-Cap Value ETF (VOE) 0.07%
10% Market Vectors Emerging Mkts Local ETF (EMLC) 0.44%
20% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
30% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% ETFS Bloomberg All Commodity Strategy (BCI) 0.29%
7.5% iShares Gold Trust (IAU) 0.25%

Here is my situation:

Roth IRA: $50k currently. Probably won't be adding more capital to this as I now contribute to a Traditional IRA (current earnings at $120k/yr).
Traditional IRA: $10k
Brokerage: $70k
Work 401k: $10k in Roth. $2k in Traditional. I will be contributing 15-18% of earnings into the Traditional portion moving forward.

My issue here is that I only have select options in my 401k. Here are the funds in my Vanguard 401k.
Vanguard 500 Index Adm (VFIAX) B 0%
Vanguard Balanced Index Fund Adm (VBIAX) B
Vanguard Developed Mkts Index Adm (VTMGX) B
Vanguard Dividend Apprec Idx Adm (VDADX) B
Vanguard Dividend Growth Inv (VDIGX) B
Vanguard Emerging Mkt Stk Index Adm (VEMAX) B
Vanguard Growth Index Adm (VIGAX) B
Vanguard High Dividend Yield Index Inv (VHDYX) B
Vanguard Inst'l Target Retirement 2015 (VITVX) B
Vanguard Inst'l Target Retirement 2020 (VITWX) B
Vanguard Inst'l Target Retirement 2025 (VRIVX) B
Vanguard Inst'l Target Retirement 2030 (VTTWX) B
Vanguard Inst'l Target Retirement 2035 (VITFX) B
Vanguard Inst'l Target Retirement 2040 (VIRSX) B
Vanguard Inst'l Target Retirement 2045 (VITLX) B
Vanguard Inst'l Target Retirement 2050 (VTRLX) B
Vanguard Inst'l Target Retirement 2055 (VIVLX) B
Vanguard Inst'l Target Retirement 2060 (VILVX) B
Vanguard Inst'l Target Retirement 2065 (VSXFX) B
Vanguard Inst'l Target Retirement Income (VITRX) B
Vanguard Mid-Cap Value Index Adm (VMVAX) B
Vanguard Prime Money Market Inv (VMMXX)
Vanguard REIT Index Adm (VGSLX) B
Vanguard Shrt-Term Infl-Prot Sec Idx Adm (VTAPX) B
Vanguard Small Cap Index Adm (VSMAX) B
Vanguard Small-Cap Growth Index Adm (VSGAX) B
Vanguard Small-Cap Value Index Adm (VSIAX) B
Vanguard Strategic Small-Cap Equity Inv (VSTCX) B
Vanguard Total Bond Market Index Adm (VBTLX) B
Vanguard Total Intl Stock Index Adm (VTIAX) B
Vanguard Value Index Adm (VVIAX) B
Vanguard Wellesley Income Adm (VWIAX) B

Any recommendations on how I can use these options to accomplish your portfolio allocation? Or does it make sense to look at my 401k and my Roth IRA as a whole and invest in some items in the 401k while doing the others in the IRA? My issue with that is that if I do it all under one account, I can automate the distribution from my bi-monthly paychecks. FYI - my IRA's are with Fidelity so i don't have the restrictions I do for the 401k. In saying this, I do get free trades for the iShares ETF's so i'd love to mirror your portfolio with iShares items if i can.

Thanks for the help!
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

dabears,

I personally find it too complicated and confusing to try to make one portfolio by combining my 401(k) holdings (it's Thrift Savings Plan in my case) with my external IRA/taxable holdings. So from personal experience, I'd recommend just making a selection for the 401(k), then separately implementing the all-seasons portfolio in your IRAs/taxables.

From that list, Wellesley would probably operate most similarly to an all-seasons portfolio in that it historical has had very impressive return and low volatility at the same time. It's at least risk parity in the sense that the bond holdings are twice that of the stock holdings, to at least partially compensate for their much lower volatility. In my personal opinion, it is Vanguard's best fund. If they ever add the Global version of Wellesley, I'd switch to that.

Azanon
PChang94303
Posts: 2
Joined: Mon Nov 06, 2017 3:22 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by PChang94303 »

Kibitzing, might Motif be the way to do 'Baskin Robbins' portfolios like this?
sergtitov
Posts: 1
Joined: Wed Dec 13, 2017 1:53 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by sergtitov »

azanon wrote: Tue Dec 12, 2017 8:35 am dabears,

I personally find it too complicated and confusing to try to make one portfolio by combining my 401(k) holdings (it's Thrift Savings Plan in my case) with my external IRA/taxable holdings. So from personal experience, I'd recommend just making a selection for the 401(k), then separately implementing the all-seasons portfolio in your IRAs/taxables.

From that list, Wellesley would probably operate most similarly to an all-seasons portfolio in that it historical has had very impressive return and low volatility at the same time. It's at least risk parity in the sense that the bond holdings are twice that of the stock holdings, to at least partially compensate for their much lower volatility. In my personal opinion, it is Vanguard's best fund. If they ever add the Global version of Wellesley, I'd switch to that.

Azanon
They do have global Wellesley and Wellington, just launched in November. Please see https://personal.vanguard.com/us/funds/ ... irect=true
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

sergtitov wrote: Wed Dec 13, 2017 1:55 am
azanon wrote: Tue Dec 12, 2017 8:35 am dabears,

I personally find it too complicated and confusing to try to make one portfolio by combining my 401(k) holdings (it's Thrift Savings Plan in my case) with my external IRA/taxable holdings. So from personal experience, I'd recommend just making a selection for the 401(k), then separately implementing the all-seasons portfolio in your IRAs/taxables.

From that list, Wellesley would probably operate most similarly to an all-seasons portfolio in that it historical has had very impressive return and low volatility at the same time. It's at least risk parity in the sense that the bond holdings are twice that of the stock holdings, to at least partially compensate for their much lower volatility. In my personal opinion, it is Vanguard's best fund. If they ever add the Global version of Wellesley, I'd switch to that.

Azanon
They do have global Wellesley and Wellington, just launched in November. Please see https://personal.vanguard.com/us/funds/ ... irect=true
I meant, if dabears ever has Global Wellesley added to his 401(k) list, then he should get that instead.

Although, now that I think about it a bit more, VG Target Retirement Income is probably an even closer proxy for a "poor-man's" risk parity, since it includes TIPS. So I guess I change my mind here - VTINX is the closest fit in his 401(k). And it has the same % equity as the Robbins portfolio.
dabears5496
Posts: 11
Joined: Mon Feb 13, 2017 10:35 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by dabears5496 »

thank you. And the VITRX is the same as the VTINX? Looks very similar.
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

dabears5496 wrote: Wed Dec 13, 2017 11:33 am thank you. And the VITRX is the same as the VTINX? Looks very similar.
Right - it's just the institutional version of the same fund.
jddphd
Posts: 42
Joined: Mon Oct 01, 2007 11:16 am

40% of investable funds in taxable account... what to do?

Post by jddphd »

Hello all,

I've read this thread with interest (and appreciation for the back-and-forth).

I am contemplating adopting this portfolio but I have a tax conundrum.

40% of my investable funds are in a taxable account. I believe I can put the equities and commodities in the taxable account. That would bring me to 32.5% But then what? TIPS, EDV, EMLC, or IAU?

One good thing: I live overseas and that foreign income tax credit is useful.

Thanks in advance,
- JD
overthought
Posts: 274
Joined: Tue Oct 17, 2017 3:44 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by overthought »

azanon wrote: Wed Dec 21, 2016 10:21 am
josh1130 wrote:
livesoft wrote:If you work your way up to 31 funds, then you can call it the Baskins/Robbins All-Flavors Portfolio.
8-)
Jokes aside, I bet we're no more than 5-10 years away from being able to "construct" our own custom portfolio with infinite amounts of very small allocations (fractional shares) to many asset classes, by using roboadvisor-like tools to do it with. So it'd work like a buffet of, say, 100 different ETFs and you could add whatever percentages of each to your own personal mix at, say, Betterment (for a small, nominal fee, of course).

Seriously, mark it down. This is coming. 10 years max.
I'm pretty sure that's exactly what M1 Finance is offering right now... and they recently dropped their fees to zero. Haven't tried it myself, but the same thought about robo-balanced portfolios did cross my mind when I stumbled across them a few days ago.
30sep16
Posts: 46
Joined: Fri Sep 30, 2016 8:52 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by 30sep16 »

Given that:

1. The top hedge fund investor (Dalio) recommends this portfolio,

2. The other top hedge fund investor (AQR/Asness) also endorsed the same basic AA in their "Commodities for the Long Run" study (looking back to the year 1877 they concluded 29% equities, 54% bonds and 17% commodities was ideal),

3. The popularity of the Tony Robbins book,

4. The fact that there's a fund available for just amount every purpose you can imagine...

Why on earth is there no fund available that implements such an AA for those who are not high-net worth individuals???

P.S. Thank you azanon for all the work you have done on this topic!
All Seasons
Posts: 227
Joined: Sun Dec 10, 2017 3:14 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by All Seasons »

""
Last edited by All Seasons on Sat Mar 31, 2018 10:55 am, edited 1 time in total.
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

30sep16 wrote: Sat Dec 30, 2017 7:50 pmWhy on earth is there no fund available that implements such an AA for those who are not high-net worth individuals???

P.S. Thank you azanon for all the work you have done on this topic!
You're welcome! I really appreciated the project, and also all of the help I received (and continue to receive) to refine the project, and the end result for me was a pretty exciting portfolio that I believe has some very nice "defense" built into the design, and one I currently use outside of my federal TSP.

I don't know why it hasn't caught on all that much. I imagine the 8-year-long or so massive US bull market currently has just about everyone focused on equity-heavy portfolios, at least for the time being.
thejimmysmith
Posts: 4
Joined: Sun Oct 22, 2017 3:02 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by thejimmysmith »

Again, thank you for this thread. I've learned alot. I'll be honest I came into investing by first reading one of the Robbins financial books then deep diving for 6 months or so into everything I could get my hands on for investing. Which actually might be a good thing since my first knowledge of mutal funds and investing was risk parity and not just buying stock and bonds.

So my question, with the market the way it is today (Dow hitting 26K) , why wouldn't I want to build an All weather with 50% in stocks since that's where the quickest growth and most growth is going to come.

I understand that this is the All-Seasons and that large of a percentage into stocks defeats the whole purpose of using this portfolio but I understand the "bones" of All Seasons. So as someone who's about to investing his Roth IRA into a portfolio, why wouldn't I go with 50% in stocks for the next 6 months while the economy is booming?
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

thejimmysmith wrote: Tue Jan 16, 2018 8:25 amSo my question, with the market the way it is today (Dow hitting 26K) , why wouldn't I want to build an All weather with 50% in stocks since that's where the quickest growth and most growth is going to come.

I understand that this is the All-Seasons and that large of a percentage into stocks defeats the whole purpose of using this portfolio but I understand the "bones" of All Seasons. So as someone who's about to investing his Roth IRA into a portfolio, why wouldn't I go with 50% in stocks for the next 6 months while the economy is booming?
I think you answered your own question. If you opt for a portfolio with 50% stocks, you would be going with a portfolio that's something other than an all-seasons/all-weather portfolio because a 50% position in stocks would produce a portfolio that shows favoritism to an investing climate where growth ends up being above market expectations and inflation is below market expectations. In contrast, what we're doing here is trying to devise a portfolio that will perform roughly similar in any investing climate.

If you have reason to believe that the economy will be booming for the next 6 months, and you turn out being correct, then yes it is definitely true that a portfolio that favors equities will likely outperform a neutral biased all-seasons/all-weather portfolio. So I think at the very least, before you consider going with something like this, you have to acknowledge that you really don't know whether growth or inflation will be above or below expectations. If you actually do know that growth will be higher than expectations, I'd suggest 2 things; 1. Don't use this portfolio and 2. Please share with all of us Bogleheads the insight you have! :mrgreen:

Finally, let me add that I think I know at least one thing you're getting at. You might be suggesting that you believe certain investing climates are more likely than others. I recall William Bernstein in his short book "Deep Risk," talked about that issue, and advocated under-allocating to protecting against certain risks, and over-allocating to risks he perceived to be more likely. So I guess that would be a variation of what we're doing here, but that's all together something different though.
b19ckr9bb1t
Posts: 1
Joined: Mon Jan 22, 2018 3:28 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by b19ckr9bb1t »

Thank you for all the work you've done on this revised portfolio. Could I ask how it performed comparatively to the 3 fund portfolio for the last 50-60 years? Or if you could point me to the right direction of how to get a program / platform to perform these calculations. Thanks in advance.
kotsp
Posts: 28
Joined: Sun Jan 14, 2018 8:32 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by kotsp »

I am quite armature at AA. I was playing around with Azanon's portfolio at visualizer. Made some tweak as follows;

Total US Stock Market 25.00%
Emerging Markets 10.00%
10-year Treasury 20.00%
Long Term Treasury 30.00%
Gold 7.50%
Commodities 7.50%

Back testing result "appears to be" very promising, but not sure it is correct. Is it too good to be true? Am I missing anything here?

Here is the results;

CAGR Stdev Best Year Worst Year Sharpe Ratio Sortino Ratio
6.64% 7.67% 14.78% -6.74% 0.79 1.18

Compared above to @azanon's results:
CAGR=6.09%
Stdev=8.39%
BestYear = 20.47%
WorstYear = -13.85%
SharpeRatio=0.67
SortinoRatio=0.95

Any input is much appreciated.

Thanks.
Cantrip
Posts: 116
Joined: Fri Oct 16, 2015 7:21 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by Cantrip »

I agree with you that getting emerging market exposure through stocks rather than bonds/local currency is a cheaper/more efficient way doing it. Throwing out commodities and replacing with gold makes it simpler and the backtest (at portfoliovisualizer.com) looks even better: only a 3% max drawdown over the past 23 years.

Total US Stock Market 25.00%
Emerging Markets 10.00%
10-year Treasury 20.00%
Long Term Treasury 30.00%
Gold 15%

The correlation differences are really nice, but I'm worried (as many other posters have) about the interest rate risk lowering returns.
kotsp
Posts: 28
Joined: Sun Jan 14, 2018 8:32 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by kotsp »

Cantrip wrote: Tue Jan 23, 2018 6:50 am I agree with you that getting emerging market exposure through stocks rather than bonds/local currency is a cheaper/more efficient way doing it. Throwing out commodities and replacing with gold makes it simpler and the backtest (at portfoliovisualizer.com) looks even better: only a 3% max drawdown over the past 23 years.

Total US Stock Market 25.00%
Emerging Markets 10.00%
10-year Treasury 20.00%
Long Term Treasury 30.00%
Gold 15%

The correlation differences are really nice, but I'm worried (as many other posters have) about the interest rate risk lowering returns.
Thank you @cantrip for the good input. If I were to go all US, the standard deviation improves still lot (from 7.83% to 7.28%) while drawdown improves from -3.31 to whooping -1.74. But we all know EM were not that good back then. Question is do we bet on EM going forward.

Thanks.
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

Cantrip wrote: Tue Jan 23, 2018 6:50 am I agree with you that getting emerging market exposure through stocks rather than bonds/local currency is a cheaper/more efficient way doing it. Throwing out commodities and replacing with gold makes it simpler and the backtest (at portfoliovisualizer.com) looks even better: only a 3% max drawdown over the past 23 years.

Total US Stock Market 25.00%
Emerging Markets 10.00%
10-year Treasury 20.00%
Long Term Treasury 30.00%
Gold 15%

The correlation differences are really nice, but I'm worried (as many other posters have) about the interest rate risk lowering returns.
kotsp substituted EM stock for the bonds not because he thought it was "a cheaper/more efficient way of doing it", rather because portfoliovisualizer doesn't have EM bonds as an option so you have to proxy it.

If you want to use EM stock and 15% gold instead of EM bonds and 7.5% commodities/Gold, you of course can for your own personal portfolio. The objective here was to try to recreate Bridgwater's All-Weather portfolio as described in their paper. They explicitly use/mention EM Bonds in the paper. And Dalio himself recommended the commodities/gold combo for the portfolio (to Robbins).

My other comment would be that no one was trying to find the most impressive back-tested portfolio. It just so happened that the estimated All-Weather portfolio does very well from a risk-adjusted return perspective. So, again, if you want to tweak by seeing how high you can get the return in portfoliovisualizer, that's a completely different effort.

If you want to search for other asset classes to help with the project, you'd want to focus more on finding uncorrelated asset classes. It wouldn't take any deep thinking to realize, for instance, that EM bonds will be less correlated with US stocks, than EM stocks are. The (lack of) correlation is more important than the "efficiency" of the asset class in isolation. Treasury STRIPS aren't efficient by themselves (IT bonds are far more efficient in isolation). Neither are commodities or EM bonds. But put them all together and voila - because they, in layman's terms, play off each other and you make money/reduce risk with the rebalancing.
kotsp
Posts: 28
Joined: Sun Jan 14, 2018 8:32 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by kotsp »

azanon wrote: Tue Jan 23, 2018 9:48 am
If you want to search for other asset classes to help with the project, you'd want to focus more on finding uncorrelated asset classes. It wouldn't take any deep thinking to realize, for instance, that EM bonds will be less correlated with US stocks, than EM stocks are.
Not exactly. With substituting EM stocks for EM bonds (unhedged global bonds) (along with other changes discussed), the overall correlation has improved from 0.65 to 0.48. That too at the same time CAGR has improved from 5.68% to 7.60%, which I guess nobody can say no. More gain for less drawdown -3.31% vs. -8.84%
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

kotsp wrote: Tue Jan 23, 2018 10:48 am
azanon wrote: Tue Jan 23, 2018 9:48 am
If you want to search for other asset classes to help with the project, you'd want to focus more on finding uncorrelated asset classes. It wouldn't take any deep thinking to realize, for instance, that EM bonds will be less correlated with US stocks, than EM stocks are.
Not exactly. With substituting EM stocks for EM bonds (unhedged global bonds) (along with other changes discussed), the overall correlation has improved from 0.65 to 0.48. That too at the same time CAGR has improved from 5.68% to 7.60%, which I guess nobody can say no. More gain for less drawdown -3.31% vs. -8.84%
Unhedged global bonds are not EM bonds. And EM Stocks are stocks. Stocks at 35% will overweight 2 of the 4 economic scenarios vs. the other 2.

It's besides the more primary point though which is this is/was an attempt to estimate or approximate the portfolio as described by the "All Weather Story" posted at Bridgewater. EM Bonds were an asset class listed. EM Stocks were not. I was only trying to improve the Dalio/Robbins portfolio because it wasn't matching up with the Bridgwater Paper.

If you're curious about testing correlations at portfoliovisualizer, I'd recommend using actual ETF tickers (EMLC for instance) to get your correlation figures. You're not required to just use the pre-chosen asset classes that don't really fit what you're trying to test.
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

I will offer one tidbit though about EM Stock: Bridgewater's 13-f filings show EM stock (VWO and EEM) at 45.5% - so its their top holding as of 9/30/2017 13F Holdings Summary (source: https://whalewisdom.com/filer/bridgewat ... ciates-inc). That being said, I'll admit I don't know much about 13F filings. I see a 14.9Billion market value listed there, when Bridgewater is supposed to have a total of $150 billion. Also, for all I know, that EM stock might be held exclusively in their "Pure Alpha" portfolio.

In any event, I don't know how to use that unless someone with some inside information or better understanding of Bridgewater could interpret that for us. So lacking that, I'll just stick with the All Weather Story paper, and of course anything Dalio shares. Dalio's next book in ~ 1.5 years is supposed to cover investing, so maybe he'll reveal more then.
RookieGER
Posts: 3
Joined: Thu Feb 15, 2018 4:47 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by RookieGER »

Very interesting discussion/thread here.

I am currently also trying to implement my personal version of the All-Weather Portfolio. For this purpose I want to backtest some scenarios with portfoliovisualizer (not for optimizing return but to get an impression what to expect). However, I am facing some issues with lack of data and therefore time horizons to be tested.

@Azanon you posted that you were able to do backtesting until 2001. It would really help me if you could provide me with the exact tickers which you have used. I will be happy to share my resulting asset allocation as soon as I am done with my analysis.
42Morri
Posts: 1
Joined: Wed Nov 14, 2018 2:55 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by 42Morri »

Thank you for the continued updates! This was a exciting read for me. I am new here and learning. Thanks for the patience. My question is regarding the allocation breakdown. EMLC is a bond etf correct? I might be missing something but does this mean you have the allocation as:

20% Stocks
65% Bonds
15% Commodities

Thanks.
azanon wrote: Wed Dec 21, 2016 7:28 am Lately, I've been working on a project of improving the Ray Dalio All-Seasons portfolio, discussed by Bridgewater here: https://www.bridgewater.com/research-li ... -strategy/ but the brought more in the public eye by Tony Robbins recent book.

If you compare the paper to the portfolio that Dalio allegedly gave Robbins, it really doesn't match the white paper, and it also clearly is not maximizing the strategy as discussed in the paper. Specifically, it seems to over-allocate to defending against deflation, and under-allocate to defending against inflation The other issue is the Robbins portfolio isn't adequately designed to generate enough return, especially considering that it isn't leveraged.

Anyway, here's my version:

15% Vanguard Mid-Cap Value Index (mutual fund or ETF)
15% Vanguard FTSC All-World ex-US Small Cap ETF (VSS)
20% Vanguard Extended Duration Treasury ETF (EDV)
20% Vanguard TIPS (VAIPX)
10% VanEck Vectors JP Morgan EM (local currency) Bond ETF - (EMLC)
10% United States Commodities Index (USCI)
10% iShares Gold Trust (IAU)

Ok, so what do I have going on here? First of all, I was able to keep the low stock position at 30%, but amp up the expected beta (by a full 1% on the portfolio) simply by using a value and small tilt on the US, and a small tilt on the foreign, while still keeping the cost low on both.

For the long-duration treasuries, instead of using 40% LT government bonds that pay interest (such as VGLT), I'm substituting 25-yr zero-coupon Treasuries to create a maximize faux leverage which allows me to drop the total allocation to use elsewhere. And using calculators at portfoliovisualizer, I was able to estimate that the 30% position on equities that I'm using approximately match the volatility of a 20% position in the zeros (amazing isn't it, that it takes a 1.5:1 of stocks to match the volatility of those STRIPS).

I'm going with a 20% position in TIPS because the white-paper clearly addresses the importance of inflation-linked bonds, and how they would be favored in 2 of 4 investing climates. While I briefly considered short-term TIPS due to less correlation with EDV, and the stocks, I'm ultimately rejecting it, because independently and over a long-term, the regular TIPS would have a higher return, and the collective duration exposure of EDV and VAIPX would still be lower than the "Robbins/Dalio" portfolio. That being said, if you wanted a bit more stability, and could live with slightly less return, short-term TIPS would be a decent substitute IMO.

For additional inflation fighting (and really an important "enemy"), I'm including 10% commodities. For similar reasons, I'm adding in 10% Gold (and also for Dalio's strong statements on Gold). Note also that this is close to the Robbins/Dalio allocation.

I'm rounding the portfolio out with 10% in local currency EM debt because it's listed in the white-paper, because it gives additional currency diversification (only 15% without it in this portfolio), and because EM debt should have high independent return.

I believe the portfolio comes out to a weighted cost of 0.23%. That's cheap enough to not be an issue. Rebalance annually.

So anyway, that's it. If you like this sort of thing, I think my version is considerably better than the half-baked Robbins one. If you have suggestions for how to tweak it further, please let me know, but please limit them to ones that would be consistent with the strategy as described in the white paper.

Again, remember, the point of the strategy is to hold a portfolio that has neutral bias towards what the future holds. So, it avoids the bias that most people have in their portfolio, which is towards economic expansion/growth/prosperity (due to high stock allocations).

........................
Edit (9/20/2017)

I wanted to make sure my most up-to-date portfolio is listed up front. See the thread to see how that evolved with others help (Also, I actually own this portfolio for my non TSP holdings which is about 40%):

Revised portfolio:
25% Vanguard Mid-Cap Value ETF (VOE) 0.07%
10% Market Vectors Emerging Mkts Local ETF (EMLC) 0.44%
20% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
30% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% ETFS Bloomberg All Commodity Strategy (BCI) 0.29%
7.5% iShares Gold Trust (IAU) 0.25%

................
Edit (9/5/2018)

I took a closer look at estimating long-term standard deviation for each asset class (using Portfoliovisualizer), and the weight/loading to each of the 4 economic scenarios discussed in the Bridgewater "The All Weather Story", and realized that the overall weightings still weren't as optimized as they could be. I was able to drop the spread between the highest risk quadrant and the lowest risk quadrant by 67% by making one change: Dropping VOE by 5% to 20%, and raising LTPZ by that same 5% to 35%. This will, of course, lower overall portfolio expected return but, again, the aim here was to create an improved, non-leveraged risk parity portfolio. It's expected return "is what it is". Anyway, by my revised calculation, the spread in risk between the riskiest quadrant, and the least risky quadrant is now only 7%, or 7% more risky.

Revised portfolio:
20% Vanguard Mid-Cap Vaue ETF (VOE) 0.07%
10% Market Vectors Emerging Mkts Local ETF (EMLC) 0.30% (note the ER dropped since last year)
20% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
35% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% ETFS Bloomberg All Commodity Strategy (BCI) 0.29%
7.5% iShares Gold Trust (IAU) 0.25%
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

42Morri wrote: Wed Nov 14, 2018 3:02 pm Thank you for the continued updates! This was a exciting read for me. I am new here and learning. Thanks for the patience. My question is regarding the allocation breakdown. EMLC is a bond etf correct? I might be missing something but does this mean you have the allocation as:

20% Stocks
65% Bonds
15% Commodities

Thanks.
azanon wrote: Wed Dec 21, 2016 7:28 am Lately, I've been working on a project of improving the Ray Dalio All-Seasons portfolio, discussed by Bridgewater here: https://www.bridgewater.com/research-li ... -strategy/ but the brought more in the public eye by Tony Robbins recent book.

If you compare the paper to the portfolio that Dalio allegedly gave Robbins, it really doesn't match the white paper, and it also clearly is not maximizing the strategy as discussed in the paper. Specifically, it seems to over-allocate to defending against deflation, and under-allocate to defending against inflation The other issue is the Robbins portfolio isn't adequately designed to generate enough return, especially considering that it isn't leveraged.

Anyway, here's my version:

15% Vanguard Mid-Cap Value Index (mutual fund or ETF)
15% Vanguard FTSC All-World ex-US Small Cap ETF (VSS)
20% Vanguard Extended Duration Treasury ETF (EDV)
20% Vanguard TIPS (VAIPX)
10% VanEck Vectors JP Morgan EM (local currency) Bond ETF - (EMLC)
10% United States Commodities Index (USCI)
10% iShares Gold Trust (IAU)

Ok, so what do I have going on here? First of all, I was able to keep the low stock position at 30%, but amp up the expected beta (by a full 1% on the portfolio) simply by using a value and small tilt on the US, and a small tilt on the foreign, while still keeping the cost low on both.

For the long-duration treasuries, instead of using 40% LT government bonds that pay interest (such as VGLT), I'm substituting 25-yr zero-coupon Treasuries to create a maximize faux leverage which allows me to drop the total allocation to use elsewhere. And using calculators at portfoliovisualizer, I was able to estimate that the 30% position on equities that I'm using approximately match the volatility of a 20% position in the zeros (amazing isn't it, that it takes a 1.5:1 of stocks to match the volatility of those STRIPS).

I'm going with a 20% position in TIPS because the white-paper clearly addresses the importance of inflation-linked bonds, and how they would be favored in 2 of 4 investing climates. While I briefly considered short-term TIPS due to less correlation with EDV, and the stocks, I'm ultimately rejecting it, because independently and over a long-term, the regular TIPS would have a higher return, and the collective duration exposure of EDV and VAIPX would still be lower than the "Robbins/Dalio" portfolio. That being said, if you wanted a bit more stability, and could live with slightly less return, short-term TIPS would be a decent substitute IMO.

For additional inflation fighting (and really an important "enemy"), I'm including 10% commodities. For similar reasons, I'm adding in 10% Gold (and also for Dalio's strong statements on Gold). Note also that this is close to the Robbins/Dalio allocation.

I'm rounding the portfolio out with 10% in local currency EM debt because it's listed in the white-paper, because it gives additional currency diversification (only 15% without it in this portfolio), and because EM debt should have high independent return.

I believe the portfolio comes out to a weighted cost of 0.23%. That's cheap enough to not be an issue. Rebalance annually.

So anyway, that's it. If you like this sort of thing, I think my version is considerably better than the half-baked Robbins one. If you have suggestions for how to tweak it further, please let me know, but please limit them to ones that would be consistent with the strategy as described in the white paper.

Again, remember, the point of the strategy is to hold a portfolio that has neutral bias towards what the future holds. So, it avoids the bias that most people have in their portfolio, which is towards economic expansion/growth/prosperity (due to high stock allocations).

........................
Edit (9/20/2017)

I wanted to make sure my most up-to-date portfolio is listed up front. See the thread to see how that evolved with others help (Also, I actually own this portfolio for my non TSP holdings which is about 40%):

Revised portfolio:
25% Vanguard Mid-Cap Value ETF (VOE) 0.07%
10% Market Vectors Emerging Mkts Local ETF (EMLC) 0.44%
20% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
30% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% ETFS Bloomberg All Commodity Strategy (BCI) 0.29%
7.5% iShares Gold Trust (IAU) 0.25%

................
Edit (9/5/2018)

I took a closer look at estimating long-term standard deviation for each asset class (using Portfoliovisualizer), and the weight/loading to each of the 4 economic scenarios discussed in the Bridgewater "The All Weather Story", and realized that the overall weightings still weren't as optimized as they could be. I was able to drop the spread between the highest risk quadrant and the lowest risk quadrant by 67% by making one change: Dropping VOE by 5% to 20%, and raising LTPZ by that same 5% to 35%. This will, of course, lower overall portfolio expected return but, again, the aim here was to create an improved, non-leveraged risk parity portfolio. It's expected return "is what it is". Anyway, by my revised calculation, the spread in risk between the riskiest quadrant, and the least risky quadrant is now only 7%, or 7% more risky.

Revised portfolio:
20% Vanguard Mid-Cap Vaue ETF (VOE) 0.07%
10% Market Vectors Emerging Mkts Local ETF (EMLC) 0.30% (note the ER dropped since last year)
20% Vanguard Extended Duration Treasury ETF (EDV) 0.07%
35% PIMCO 15+ Year US TIPS ETF (LTPZ) 0.20%
7.5% ETFS Bloomberg All Commodity Strategy (BCI) 0.29%
7.5% iShares Gold Trust (IAU) 0.25%
Yes that is correct (the asset allocation of stocks, bonds, commodities). It's a best attempt at a risk-balanced portfolio, that still maximizes return without any real leverage. Despite only 20% equities, realize in practice it will be quite a bit more "supercharged" than that allocation implies (or should have a relatively high long-term expected return), because of the faux leverage with those bonds, along with EM local currencies. And of course commodities aren't for the faint of heart either.

Probably the most important thing for a holder of this portfolio to do, is to not constantly monitor each of the individual asset classes. As I've been running mine, it's easy to see that several of the individual holdings can be wildly volatile and almost always at least one of the assets will be doing poorly, but the overall portfolio changes far more modestly from day-to-day.

As with any portfolio, you can't just run it a few months or even a year or two and decide if it works. You have to be patient. It hasn't been the best year or two for risk parity of any type.
meshuggahnmcc
Posts: 10
Joined: Mon Nov 19, 2018 2:21 pm

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by meshuggahnmcc »

hey man just registered to say thank you for all your hard work! This thread was amazing. Woke up today searching for this info.

IM not seeming to get your market correlation results when I plug it all into portfolio vizulaizer. Any help in this regard? The generic all weather on the visualizer seem to have better metrics in terms of correlation and beta.

Not sure if you watched this(https://www.cnbc.com/2018/11/15/hedge-f ... arket.html) . Dalio speak about some interesting concepts.
jpiabrantes
Posts: 9
Joined: Wed Nov 28, 2018 8:56 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by jpiabrantes »

Excellent discussion!

This website shows the positions held by Bridgewater: https://www.nasdaq.com/quotes/instituti ... -lp-699510
maybe it could give some clues to what products they hold for their all-weather portfolio?
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

meshuggahnmcc wrote: Mon Nov 19, 2018 2:30 pm hey man just registered to say thank you for all your hard work! This thread was amazing. Woke up today searching for this info.

IM not seeming to get your market correlation results when I plug it all into portfolio vizulaizer. Any help in this regard? The generic all weather on the visualizer seem to have better metrics in terms of correlation and beta.

Not sure if you watched this(https://www.cnbc.com/2018/11/15/hedge-f ... arket.html) . Dalio speak about some interesting concepts.
Thanks for the compliment. Would you mind reminding me (by quoting me) where I made reference to market correlation results (so I don't have to re-read the thread)? I'm not exactly sure what you mean by the generic all weather having better metrics for correlation. Of course lower correlation to the US Market would be better (for an "all-weather" fund), but that metric is not the be-all-end all. I gave more deference to representing each of the 4 quadrants discussed in the "all weather story" than I did trying to obtain the lowest US market correlation.
Last edited by azanon on Thu Nov 29, 2018 7:35 am, edited 1 time in total.
Topic Author
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Improving the Dalio/Robbins All-Seasons Portfolio

Post by azanon »

jpiabrantes wrote: Wed Nov 28, 2018 9:50 am Excellent discussion!

This website shows the positions held by Bridgewater: https://www.nasdaq.com/quotes/instituti ... -lp-699510
maybe it could give some clues to what products they hold for their all-weather portfolio?
I forget the exact rules, but 13-f's have a lot of things left off of them. In any event, I believe their "Pure Alpha" fund is actually larger than the all-weather fund, so their overall 13-f is going to be of little use anyway. I'm only attempting to make a proxy all-weather fund without leverage.
Post Reply