3
$\begingroup$

Somewhere along the lines, Ray Dalio's all weather became known to be a portfolio with the following weights:

  • 30% Equities
  • 40% Long Term Treasuries
  • 15% Intermediate Term Treasuries
  • 7.5% Gold
  • 7.5% Commodities

When I run this with portfolio visualizer, and drill down to the metrics, I get a risk contribution that's not at all equal across different asset classes.

Risk Contribution The questions are:

  • 1) Why is that?
  • 2) How were the above weights derived?
  • 3) How are these weights static when risk parity talks about using leverage to increase exposure in low volatility assets compared to the higher volatility ones?
$\endgroup$
  • 1
    $\begingroup$ These weights come from an interview (see “Money: Master the Game”). The interviewer asked Dalio what strategic asset allocation he’d recommend to a non-institutional investor who can’t use leverage. Suffice to say that these are not what the actual All Weather portfolio uses at all. $\endgroup$ – Helin Sep 29 '19 at 10:35
  • $\begingroup$ And this is what confuses me. What do these weight allocations do for the average investor if they're far from what all weather really uses???? I don't understand and I see them quoted all over the internet as the de-facto "all-weather" strategy for the average joe $\endgroup$ – schone Sep 29 '19 at 14:06
2
$\begingroup$
  • As mentioned in the short comment above, this particular allocation comes from an interview documented in the book Money: Master the Game. Dalio explained that "in his All Weather strategy, they use very sophisticated investment instruments, and they also use leverage to maximize returns." The interviewer then followed up by requesting the weights that an "average person" can do, "without any leverage, to get the best returns with the least amount of risk." Dalio then provided these weights, emphasizing that it was a "sample portfolio" that "wouldn't be exact or perfect."

  • The actual All Weather portfolio is almost certainly dramatically different. At a high level, the portfolio is constructed such that it has (roughly) equal risk exposures to four environments – rising growth, falling growth, rising inflation, and falling inflation. Absolutely no claim is made that the contribution to total risk is equal across underlying asset classes/instruments.

  • The exact methodology to implement All Weather is not known, which is understandable. But the conceptual framework is published on this website and is very educational.

| improve this answer | |
$\endgroup$
  • $\begingroup$ Is there any way one can calculate the exposures utilizing these particular instruments to prove to one self that atleast we're close towards having equal risk in the four environments? $\endgroup$ – schone Sep 29 '19 at 21:35
  • 1
    $\begingroup$ @schone There's no precision to this, but my quick calculation suggests that the portfolio still has too much growth exposure and insufficient inflation exposure. It is definitely an improvement upon traditional 70/30 or 60/40 type of portfolios. $\endgroup$ – Helin Sep 30 '19 at 7:08
  • $\begingroup$ Can you share how you calculated that? and how does one stay on top of this? $\endgroup$ – schone Sep 30 '19 at 16:23
  • $\begingroup$ There are many ways to do this. I won't get into our exact methodology, but here's one example published by AQR: aqr.com/Insights/Research/Journal-Article/… $\endgroup$ – Helin Sep 30 '19 at 23:32
1
$\begingroup$

What horizon are you using to calculate your "equal risk contribution" weights? I am guessing it's 1 year, maybe 5 year... versus Dalio looking over decades. Plus, where are TIPS? When the Taper Tantrum hit in 2013, Bridgewater was on the record being 180% long of these! Gold obviously shares characteristics/correlation with TIPS; but BW holds truck loads of the latter, if you're looking into replication strategies.

The short answer to your question is that "risk parity" covers a multitude of sins (and graces). It's an approach, not a strategy. Equal Risk Contribution is but one way to try to achieve risk parity. And it's far from infallible. Build a portfolio with eg the S&P500, the NASDAQ, the Russell2000, and Treasuries. ERC across these four is still going very heavy on the stocks.

Dalio's original All-Weather approach was to diversify the portfolio to positive and/or negative growth and/or inflation shocks. Indifference to these simply requires an estimate of the long-term beta of each market to each flavour of shock. From which the (unlevered) portfolio composition becomes a simple simultaneous equation to solve.

hope this helps

| improve this answer | |
$\endgroup$
  • $\begingroup$ 10 years is what i'm doing as it's the limits of what portfolio visualizer has for some of the ETFs mentioned across the internet. Is there a place that has the ability to simulate a lengthier time frame? Perhaps even using futures instead of ETFs? $\endgroup$ – schone Sep 29 '19 at 14:04
0
$\begingroup$

Probably this does not answer your question at all, but as a portfolio manager I believe it is very much worth sharing: the number of successful funds not using asset allocation techniques is very high out there. This does not mean firms are unefficient, rather raises questions over the reliability of such optimization tools. Plenty of literature over their issues: instability, curse of dimensionality, non-representative, etc. To conclude, it does not surprise me seeing weights far from literature-backed allocations

| improve this answer | |
$\endgroup$
  • $\begingroup$ So which techniques do they use to allocate capital? $\endgroup$ – Lisa Ann Sep 29 '19 at 14:54
  • $\begingroup$ When you have a benchmark, which is not of the cases, you are market neutral and active some factors $\endgroup$ – Vitomir Sep 29 '19 at 14:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.