Difference between PP and Risk Parity

[Hal]

Am I correct in saying Risk Parity and the PP are completely different approaches?

PP - focuses on assets the do well in different economic climates

Risk Parity - attempts to have equal volatility in each asset class that do well in different economic climates.

For instance, on this actuarial site: http://actuary-info.blogspot.com.au/201 ... escue.html
It is suggested that the minimum standard deviation is 15% each shares/gold and 70% 10 year bonds

While a "Lemonade PP": https://finpage.blog/2014/09/11/harry-b ... portfolio/
25% shares/gold 50% Intermediate Bonds

So while the actual percentages are close, the actual allocation is done by completely different methodologies?

For instance, if 50 year bonds were available, would the PP hold 25% of them as they would out-perform in a depression....

Looking forward to your thoughts,
Hal

[DragonJoey3]

Hal,

Yes, you are correct in noting that the PP is not a Risk-Parity approach. In fact for a good while there Gold was barely moving at all (very low volatility) and the PP would still allocate 25% to it.

Risk parity ascribes a level of risk per asset based on that assets volatility, and attempts to hold all constant. The PP assumes equal levels of risk for all economic conditions and assigns equal weight to all (even though perhaps not all are equally likely).

For example if the volatility of gold were to change dramatically over the next decade or so a risk-parity portfolio would hold more gold since the risk from holding it has now decreased and it can hold a larger segment in the portfolio (wishing to hold risks constant). The PP recognizes that gold is an inflation hedge (regardless of volatility), and that inflation is one of the four core economic conditions. Since the PP was created on the basis of not assuming which economic conditions the future holds we assign the risk of inflation over a given economic period to be one in four (25%) and hold gold accordingly.

With regards to your statement about 50 year treasuries, I suspect the PP would hold them if they were available since that would provide the best anchor against deflation and falling interest rates moving forward. In a sense we want the highest volatility in all sectors so as to help nullify losses in the remaining three sectors resulting from whatever economic condition the world is in at the moment.

~DragonJoey3

[Mr Vacuum]

Ooh, good one. The way I look at it, the underlying assets and economic principles are the same but the two portfolios optimize for different things. Risk parity optimizes for risk management and is willing to mess with picking a volatility look back period and changing the allocation over time, saying we can do a little better and it's worth the effort. The PP emphasizes Permanent, I.e. not touching the thing, saying 4x25 is close enough (and near optimal, to boot, long term) and it's not worth it to worry about changes.

4x25 HBPP is so simple that it almost must be a fundamental law of the universe. Hopefully people who have read all of Harry's books will chime in, but I think looking at the earlier more complicated allocations and the fact that the PRPFX fund Harry advised uses a different allocation to this day indicates that there are different ways to go and Harry really wanted to back up and simplify to something easier for people to implement but still true to the principles.

Awesome question. I look forward to more responses.

[buddtholomew]

Calling MG anyone?
Machine Ghost did some extensive work on risk parity and duration.

[jhogue]

I like the idea of 50 year Treasury bonds.

Several recent threads have expressed worries over the flattening of yield curve. Some have suggested retreating to shorter maturities-- a sort of "half barbell" response.

However, I think that 50 (or even 100) year T bonds would capture more volatility from the far end of the yield curve and require fewer sales of T bonds at 20 years for PP investors going forward. Such ultra long securities would become more valuable in case of a long-term Japanese-style deflationary period, let alone an actual depression.

[Jack Jones]

I like the idea of 50 year Treasury bonds.

Several recent threads have expressed worries over the flattening of yield curve. Some have suggested retreating to shorter maturities-- a sort of "half barbell" response.

However, I think that 50 (or even 100) year T bonds would capture more volatility from the far end of the yield curve and require fewer sales of T bonds at 20 years for PP investors going forward. Such ultra long securities would become more valuable in case of a long-term Japanese-style deflationary period, let alone an actual depression.

I think the best you can do now in that regard is Vanguard's Extended Duration Treasury ETF (EDV).

[Kbg]

Small technical point...RP also incorporates asset correlation as an explicit factor. Practical point, RP almost always uses leverage in portfolio construction or basically you end up with a big wad of cash/bonds.

[Tyler]

I personally view the PP as a simplified risk parity approach without the active management. Read anything by Ray Dalio about his All-Weather risk parity fund, and a lot of it could have been written by Harry Browne. The big names tend to make it sound a lot more complicated than it is to justify their big fees, but the basic idea is really not that different from the core PP concept.

[Hal]

Thanks for your comments - most interesting.

Tyler - I see you consider the PP as a simplified risk parity approach. After much reading of Bridgewater papers, I am not sure if HB meant the permanent portfolio to be a risk parity approach, rather he might have just came up with a similar allocation using a different methodology.

So, jumping in our time machine and having a chat to Harry.....

"We have these new 50 Year Bonds, what allocation would you suggest ? "

1. The Risk Parity PP Harry: "Well, they have twice the volatility, so around half of the normal TLT allocation ( 12.5 % )

2. The 4 Economic Climate PP Harry: "Use the longest Treasury Bond you can get and allocate 25%"

3. The Unknown Harry : you fill in the quote

For the people that have read all his books and listened to his talks, do you think he would describe the PP as being intentionally built around a risk parity (ie equal volatility ) approach?

Hal

[Tyler]

For the people that have read all his books and listened to his talks, do you think he would describe the PP as being intentionally built around a risk parity (ie equal volatility ) approach?

I haven't read all his stuff, but I don't think so. I think it's more likely that some of the big risk parity players intentionally built around the Permanent Portfolio economic conditions idea, layered on the risk parity concept as a further tweak to optimize it mathematically, and claimed that they had some brilliant original idea. Note that the PP works without complete risk parity, but risk parity does not work without properly distributing assets among the possible economic conditions (covering all risks).

[Hal]

Hmmm. It seems to me the whole risk parity approach relies on the asset class volatility remaining stable.

If volatility varies greatly from historical norms due to some shock, you could be left exposed with a large asset class holding.
Eg: PP long term treasuries = 25%, All seasons portfolio = 40% L.T. and 15% Intermediate Bonds.

Have to give it some more thought, but currently I think Harry's approach minimises large drawdowns better.
It has built in 25% "firewalls".

[Kbg]

There is an excellent set of articles on RP applied to the PP at GestaltU. It's a thought provoking piece.

[Hal]

Thanks Kbg,

A really good find, will have to read it carefully.
This diagram hit home....

http://www.gestaltu.com/wp-content/uplo ... erfall.png

[Kbg]

Thanks Kbg,

A really good find, will have to read it carefully.
This diagram hit home....

http://www.gestaltu.com/wp-content/uplo ... erfall.png

Yes sir, that's it in a conceptual nutshell. I've become somewhat of a believer in naive risk parity because there isn't much I haven't seen it improve.

[Mark Leavy]

Thanks Kbg,

A really good find, will have to read it carefully.
This diagram hit home....

http://www.gestaltu.com/wp-content/uplo ... erfall.png

Yes sir, that's it in a conceptual nutshell. I've become somewhat of a believer in naive risk parity because there isn't much I haven't seen it improve.

I'm also a strong believer in naive risk parity - with the exception that I prefer my measure of asset volatility to be the max historical drawdown for that asset over the longest period that I can get relevant data for.

It keeps things simple, seems a bit more relevant to my interests and the metrics don't change much over time.

[Kbg]

Thanks Kbg,

A really good find, will have to read it carefully.
This diagram hit home....

http://www.gestaltu.com/wp-content/uplo ... erfall.png

Yes sir, that's it in a conceptual nutshell. I've become somewhat of a believer in naive risk parity because there isn't much I haven't seen it improve.

I'm also a strong believer in naive risk parity - with the exception that I prefer my measure of asset volatility to be the max historical drawdown for that asset over the longest period that I can get relevant data for.

It keeps things simple, seems a bit more relevant to my interests and the metrics don't change much over time.

?

How do you apply that to day the PP? Sum the MaxDDs and then weight accordingly by the inverse?

[Mark Leavy]

How do you apply that to day the PP? Sum the MaxDDs and then weight accordingly by the inverse?

Exactly. Except for the last few years I've been applying it to my version of the 2X PP. (After being heavily influenced by you to delve deep into the mechanisms of XIV).

Allocation is (MaxDD)/sum(MaxDD)). Then normalize sum(allocations) to 1.

I run 45% Long Bonds, 35% Physical Gold, 15% XIV, 5% Cash (strictly for slop).
Rebalance when XIV hits 7.5% or 30%. Three times so far in 2.5 years.

And then outside of the portfolio, I maintain 3 years of living expenses in cash.

In rough numbers, this gives me a very good approximate 2X PP (both up and down) without running any real leverage - and still holding "safe" instruments like long bonds and physical gold.

I like using a double/halving of XIV as my rebalance trigger as it feels like I am just using Shannon's Daemon to make my bets. Roughly even odds of a double or a half - but over the long run, the doubles return more than the halves lose.

Rebalance. Ignore the portfolio. The "naive" Risk Parity keeps it relatively stable. Then when a rebalance band is hit, I check to see if my last roll of the "XIV dice" doubled or halved my bet.

[Kbg]

Cool. TMF/XIV makes for an interesting mix in a good way...