The Risk Parity Portfolio Construction Journal

D

Daal

In this journal I'm looking to discuss, share ideas and debate different methods on how to construct a balanced multi-asset passive portfolio. These days people call this "risk parity".

Friends and family ask me sometimes about managing money for them, I always refuse simply because the amounts are not big enough to handle the hassle (and extra stress). Furthermore since I'm doing a lot of short-term trading with Portfolio Margin, I have no shortage of capital. At the same time, I want to be able to guide them in the right direction so they can run the portfolio themselves.

IME, most people who "decide to invest in the stock market" do it in ways that doom them to failure from day one. They typically pick a few stocks that they guess are good companies and hold them while they are going up (maybe even buy more), only them to puke out and sell it all when the market collapses. The standard deviation of their approach is so high that they would never invest on it if they knew what they could face.

The risk parity approach is perfect for those people. Its also ideal for traders who have extra cash laying around that they do not need as margin for short-term trading.
I'm not a fan of the idea of running a lot of risk with that cash because the capital devoted to short-term trading is already highly risky, the extra volatility of making risky investments with excess capital brings extra stress and this hurts short-term trading performance due "mental capital" drawdowns.
I rather follow Taleb's "barbell" strategy of having no risk with 90% of the capital and a lot of risk with 10%.

A conservative risk parity approach is not necessarily without risk but it reduces drawdowns dramatically

For instance, Ray Dalio (a big proponent of this approach) described a typical portfolio for the book "Money: Master the Game"
He suggests:
40% in 20-25y US Treasuries
15% in 7-10y US Treasuries
30% in stocks (not necessarily just one index)
7.5% in commodities
7.5% in gold

Over the last 40 years this portfolio returned 9.88% a year with a max drawdown of -~4% (avg loss ~1.5%). It was up 85% of the time at any given year end

Now, I do have some problems with this portfolio, specially for those who live outside the US. But that is the point of this journal, to discuss, make suggestions, debate and come up with different portfolios that might make more sense.

Here are some reading materials for those interested in participating
http://www.advisorperspectives.com/newsletters12/pdfs/Why_a_60-40_Portfolio_isnt_Diversified.pdf
http://www.elitetrader.com/et/index...rm-investment-who-wants-to-discuss-it.125840/
The last one includes several posts made by Cutten, so make sure you read all the pages

There is also some info from David Swensen in the book I mentioned earlier. He is more aggressive and owns a lot more stocks but I think overall his portfolio is good for those who can tolerate bigger drawdowns.

I will post occasionally new ideas on portfolios as I come across them and hopefully people will have some good inputs and we can come up with better developed ideas
 
Last edited:
Risk parity is nice because it closely approximates the optimal portfolio when the sharpe ratios of the underlying assets are similar. It removes much of the uncertainty of how to manage or direct assets.

Some critics point out it will leverage low risk premiums (i.e. do you really want to be long a ton of bonds right now?).
 
Harry Browne's well known Permanent Portfio is:

25% Stocks
25% Gold
25% Cash/Short Term Treasuries
25% Long Term Treasuries

I keep a pretty nice chunk of my cash in the Permanent Portfolio.

There are many articles and books on that portfolio as well.
 
If you are advising people on risk parity portfolios then I would start with the question of what the universe would be. Most studies as far as I remember, indicate no more than 20 or so are needed. Post above is going with four. Where would you start?
Then there is the question of what instruments you would choose to replicate the chosen universe. When one says stocks is that the cash SPY or something else? How much tracking error against that are you ok with?
How would you gain exposure to segments like private equity or maybe insurance or em debt etc which imo are oft ignored but crucial for low correlation universe construction.
Then as #2 hints at, how would you handle vol/leverage measurement problems for carry products with significant tail risks but low second moments. Would you linearly obtain 'parity' or tweak the weightings non linearly ? How, if so?
My 2 cents are that true rpp that can be on autopilot is very hard from the retail side. My toy setup would be to just use the best proxies for 'stocks', 'bonds', gold, dollar and cash and devise a mechanical/statistical way to dynamically regime switch between weightings of those. Obviously, this would leave you behind the ball (lagging) real time a bit but would achieve the respectable drawdown adjusted numbers you are looking for, and it would also be something you could potentially quickly package as a 'service' or 'guidance' for you friends and fam.

regards.
 
Interesting info here
http://bonnerandpartners.com/a-smarter-way-to-build-wealth/

"
CL: Your latest book is called Global Asset Allocation. I read it, and I found it fascinating. The core message is that the exact mix of asset classes in your portfolio matters less than the fees you pay for your portfolio. Can you tell me a bit more about that?

MF: Investors spend 90% of their time obsessing over their asset allocation – in other words, how much to invest in stocks, bonds, commodities, and cash.

So, I looked at asset allocation models of 15 of the world’s most successful money managers. Guys that manage hundreds of billions of dollars… like Rob Arnott at Research Affiliates… Ray Dalio at Bridgewater Associates… Mohamed El-Erian, the former co-chief investment officer of Pimco… and David Swensen, the chief investment officer at Yale University.

I also looked at asset allocation models that have been around for 20 or 30 years… such as Harry Browne’s Permanent Portfolio.

I tested each asset allocation all the way back to 1973 to see how they performed over the long term – both on a nominal and a real basis, meaning after accounting for inflation.

CL: And the results were surprising…

MF: They sure were. Although all the asset allocations were completely different from one another, what I found was that they all had incredibly similar performance over more than three decades… as long as they had some of the main ingredients: global stocks, global bonds, and real assets, such as real estate and commodities.

The ones that had more in real assets did better in the 1970s, because of the high levels of inflation. But those then did worse in the 1980s and 1990s, as inflation levels fell. The ones that had more equities, or equity-like investments, did great in the 1980s and 1990s.

But they were all very similar over the long run. If you exclude the Permanent Portfolio – which has 25% of the portfolio in cash – they had less than a one-percentage-point difference between returns.

CL: That’s fascinating. I would have expected a much bigger difference in returns.

MF: Me too…

Then, I did a simple thought experiment. What if you were able to go back to 1973 in a time machine and say, “All right, I’m going to invest in the best of these 15 portfolios.” What if you had perfect foresight, in other words, on which asset allocation will be the best performer?

Now, let’s say you implemented your asset allocation through mutual funds. And you paid the average annual fee for mutual funds of 1.15% a year. Well, that would knock the performance of the best-performing asset allocation almost all the way down to worst performer.

Then you say, “You know what? I’m going to hire a financial adviser to help me stay on track with my allocation.” Well, the average fee for a financial adviser is 1% a year. If you add the average 1% fee on top of the average 1.15% fee for mutual funds, that takes the best-performing asset allocation and makes it worse than the worst asset allocation.

Over the long run, the fees you pay on your portfolio are more important than your asset allocation."
 
I will say this, its tough to hold or invest in these portfolios now. Pretty much anything that you look at seems overpriced. Risk-premiums have been compressed to much that I decided to RISK missing out on future gains but waiting for a significant correction to get in.
I figure this is a better alternative than investing and then freaking out over corrections. This will hurt my performance in my main income generating activity (daytrading).
I rather sit on 3-4y US Treasuries and money market ETFs while I wait for some kind of value in stocks, bonds and REITs
 
People are blaming the market sell-off on Risk Parity. This seems to be a weird definition of RP. Or it is being implemented weirdly. The whole idea of Risk Parity is to have portfolio diversified so you don't get involved the market girations because your drawdowns are already limited. It makes little or no sense to say you are doing RP and then sell stocks like crazy beacuse the VIX is up
 
You must log in or register to reply here.
Back
Top