PPs for Different Risk Tolerances

[melveyr]

Work has finally slowed down so I had some time to do some thinking and number crunching. I made a post here you guys might find interesting:
http://www.stableinvesting.com/2014/11/ ... s.html?m=1

I think that the PPs core drivers can be exploited at varying levels of risk, and I hope the post illustrates that. I am curious about everyone's thoughts on this

[MachineGhost]

The theoretical principle is sound but the implementation is problematic without using past-looking risk metrics.  I started to do work on exactly that concept last year ("Duration Hedged PP") but when it came to gold, I was stumped.  Forward-looking durations can also change interrim and necessitating a portfolio rebalancing to stay in parity.  So while its trivial to calculate for the S&P and TBonds, gold is the obstacle to successful implementation.  How does one calculate an imputed duration from gold???

So I'm not going to implement a PP based on either past-looking or forward-looking volatility because that is just an indirect ad hoc risk measurement for bankruptcy.  Duration is the theoretically correct way to measure risk and target risk.  Until then, equal-weight I remain.

More juice can be had from trend following (or adjusting the rebalancing bands for you purists) then fiddling with allocation.

[D1984]

Work has finally slowed down so I had some time to do some thinking and number crunching. I made a post here you guys might find interesting:
http://www.stableinvesting.com/2014/11/ ... s.html?m=1

I think that the PPs core drivers can be exploited at varying levels of risk, and I hope the post illustrates that. I am curious about everyone's thoughts on this

melveyr,

I'm just wondering...why didn't you include a PP made with TMF? It has a duration of just over 51 years.

See http://www.direxioninvestments.com/wp-c ... -Sheet.pdf

Also, why did you choose to amp up (or amp down) the bond duration instead of doing same with the level of stock risk? For instance, one could use more stocks (i.e. more than 25%) if they used SPLV or HDV, or they could use less stocks (less than 25%) if they used something like a mixture of QLD and RZV, or they could allocate even less % to stocks with something like UPRO and TNA.

Unfortunately, as MG has alluded to, it's virtually impossible to moderate/increase gold this way (well, it would actually be quite possible, but it would require the holding of gold plus the use of options--either on GLD or on gold futures--rather than just holding only gold itself).

[Mark Leavy]

Very good to hear from you again Melveyr.  You make me think every time you post.

I like your thinking here.  This is how I broke it down for my slow thinking self:

1) Treasuries from cash to long term bonds are a continuum.
2) The PP holds 50% of its assets in this continuum and the volatility of that 50% is directly proportional to the average duration of the bonds/cash.  This is just math.
3) We have a lot of options to slide that duration lever up or down.
4) Not so many options with the other assets - but we can change their allocation percentages.
5) In summary, we can change the duration of the bonds/cash and the allocation percentage of gold and equities to adjust the PP to whatever risk tolerance we desire - without violating the fundamental premises of the PP.

Well... I may have completely misinterpreted you - but again, welcome back.

[melveyr]

MG,
I'm not sure I understand the application of duration outside of bonds. Standard deviation has its flaws (and is a moving target) but I think it's the best we have.

D1984,
I am not a fan of the x3 (or x2) products. They are designed to be trading vehicles and the products' own prospectuses warn against holding over long periods of time. I didn't dial up or down the equity portion because it is less "mathy" than doing so for bonds because duration is such a clean concept. But, adjustments to the equity portion could certainly warrant a closer look.

Mark,
100% my thought process, especially the reasoning for why I only fiddled with the bond portion

[MachineGhost]

I'm not sure I understand the application of duration outside of bonds. Standard deviation has its flaws (and is a moving target) but I think it's the best we have.

All investments are a discounted stream of future cash flows.  The duration measures how long it would take to receive back 100% of your risk principal plus any interest and dividends, aka total return.  That point is the terminal date.  Within a retirement portfolio concept, to hedge the terminal date risk all assets must have equally matched durations.  Since gold is not an investment paying cash flows, I don't have the requisite skills to impute a duration from it.  I'm sure there are ways.

[buddtholomew]

Adding to or removing cash from the PP achieves the same result as you control weighted duration, correct? I am 20/20/60 with a mixture of TLT and CD's to created a weighted fixed income duration of 5.6 years. Its mental accounting as I get PP returns and then whatever my CD's are paying at the time.

[MachineGhost]

Adding to or removing cash from the PP achieves the same result as you control weighted duration, correct? I am 20/20/60 with a mixture of TLT and CD's to created a weighted fixed income duration of 5.6 years.

Are you comfortable with that against your stock portion being (20%?) whatever weight * 51 years duration?

[buddtholomew]

Adding to or removing cash from the PP achieves the same result as you control weighted duration, correct? I am 20/20/60 with a mixture of TLT and CD's to created a weighted fixed income duration of 5.6 years.

Are you comfortable with that against your stock portion being whatever weight * 51 years duration?

I'm not smart enough to challenge the 4x25 allocation and I prefer to hold additional cash to purchase equities, gold or treasuries should rebalance bands dictate.