Asset Volatility Question

General Discussion on the Permanent Portfolio Strategy

Moderator: Global Moderator

Post Reply
barrett
Executive Member
Executive Member
Posts: 1982
Joined: Sat Jan 04, 2014 2:54 pm

Asset Volatility Question

Post by barrett » Fri Nov 07, 2014 6:59 am

I thought someone out there might have the relative volatility numbers on gold, stocks and LTTs. In my short time in the PP (since the beginning of this year more or less), it seems to me that both gold and LTTs have been more volatile than stocks... that they both have considerably more days than stocks with price variations of more that one percent, for example. Can someone either confirm this or show me otherwise? Where 60/40 BH-type investors have days here and there that they hear the phrase "markets plunge!", I feel like that's a super common thing in the PP. I know the usual warnings about not looking at assets in allocation. I'm just curious if my brain is sifting through the numbers and telling me that there is more volatility than there actually is. Thanks.
User avatar
Gosso
Executive Member
Executive Member
Posts: 1052
Joined: Fri Jan 06, 2012 8:22 am
Location: Canada

Re: Asset Volatility Question

Post by Gosso » Fri Nov 07, 2014 11:24 am

The gain to pain ratio shows that the PP has had a tough couple of years, while the 60/40 has enjoyed much happiness.  The chart below shows the gain to pain ratio measured over two year rolling periods (the higher the G/P the better).  It's pretty interesting - I bet it would be a leading indicator for fund flows into various mutual funds or ETFs for the PP and 60/40.

[img width=500]http://i58.tinypic.com/2e2m3pv.png[/img]
User avatar
sophie
Executive Member
Executive Member
Posts: 1959
Joined: Mon Apr 23, 2012 7:15 pm

Re: Asset Volatility Question

Post by sophie » Fri Nov 07, 2014 4:01 pm

Great chart Gosso!

Just one thing...how is the pain/gain ratio calculated?  Interestingly, the 60/40 appears to be very close to the PP or better for almost all the time except some periods in the 1970s.  I imagine that's the eras when gold was skyrocketing.  That's a bit counterintuitive given the performance charts showing the smoother, steadier performance over time of the PP.
"Democracy is two wolves and a lamb voting on what to have for lunch." -- Benjamin Franklin
User avatar
Mark Leavy
Executive Member
Executive Member
Posts: 1950
Joined: Thu Mar 01, 2012 10:20 pm
Location: US Citizen, Permanent Traveler

Re: Asset Volatility Question

Post by Mark Leavy » Fri Nov 07, 2014 4:35 pm

I've really been enjoying the recent posts on this board.  Some very thoughtful and serious queries and comments.

Barrett:

I looked through my recent simulation data and I don't really have anything to answer your question.  For what it is worth, this is what I can offer.

For gold and long term treasuries.  (using daily data over the last 10 years)
GLD has had a maximum draw down of 38% lasting 1046 days.
TLT has had a maximum draw down of 27% lasting 960 days.

These numbers are not inflation adjusted.

If interested, I could, with a small amount of work, come up with draw down numbers (for the last 10 years) that included equities and/or a vanilla PP.

I tend to look at draw down instead of standard deviation or variance as the metric of "volatility", however I am quite intrigued by Gosso's pain metric.  It makes a lot of sense to me and I think it is worth looking at.

The once and future melveyr (hallowed be his name) once talked about allocating portfolio positions inversely proportional to their volatility.  Seriously brilliant.  Except volatility is difficult to define.  I ran a few simulations using 10 year maximum draw downs as a proxy for volatility and the portfolios performed quite well.  Regardless of asset class.
barrett
Executive Member
Executive Member
Posts: 1982
Joined: Sat Jan 04, 2014 2:54 pm

Re: Asset Volatility Question

Post by barrett » Fri Nov 07, 2014 8:52 pm

Mark,

I was actually asking a much simpler question but I am happy to have it taken in a more interesting direction. My question was merely if gold and LTTs have been more volatile than stocks this year. I asked it at about 8:00am today and, sure enough, LTTs were up over 1% and gold went up almost 3% today. A three percent rise in stocks would produce headlines like "S&P Storms Ahead!" But these big moves in bonds and gold just seem to be happening a lot more frequently than they do with stocks. I realize with bonds that big moves are more common when interest rates are this low, so I guess that's the answer to part of my query.

Yeah, draw downs are what really matters as you have pointed out Mark (ditto with Sophie in another thread today).

By the way, what ever happened to melveyr? It seems he was gone from this forum before my time. Maybe we can tempt him back to weigh in on Gosso's Pain to Gain Ratio.... or Curve Fitted Back Testing (I too thought it had something to do with a spinal issue). Reminds me of Pointedstick's Carless Living thread that a bunch of us mistook for Careless Living.
User avatar
Mark Leavy
Executive Member
Executive Member
Posts: 1950
Joined: Thu Mar 01, 2012 10:20 pm
Location: US Citizen, Permanent Traveler

Re: Asset Volatility Question

Post by Mark Leavy » Fri Nov 07, 2014 9:18 pm

Apropos of nothing, I ran some numbers.  These don't really mean anything, but I thought I would calc them just the same.
(All numbers current to today - no inflation adjustments)

Using 10 year daily data, dividends reinvested (purple verticals) I came up with these numbers:

GLD: 9.72% CAGR, 40.52% draw down, 1173 days
TLT: 11.71% CAGR, 25.03% draw down, 762 days
VTI: 11.66% CAGR, 55.96% draw down, 1226 days

PP (15/35 bands) : 9.98% CAGR, 12.67% draw down, 512 days

You can clearly see that all of the three volatile assets are indeed volatile - and that the PP does a fine job of smoothing them all out.
It may not be optimum, and it is certainly not prescient, but it is a good bet.

GLD
[img width=600]http://i61.tinypic.com/346auu0.png[/img]

TLT
[img width=600]http://i57.tinypic.com/2vdjpe8.png[/img]

VTI
[img width=600]http://i60.tinypic.com/qxqtfr.jpg[/img]

Permanent Portfolio (15/35)
[img width=600]http://i58.tinypic.com/2wd7ue0.png[/img]
User avatar
Pointedstick
Executive Member
Executive Member
Posts: 8864
Joined: Tue Apr 17, 2012 9:21 pm
Contact:

Re: Asset Volatility Question

Post by Pointedstick » Fri Nov 07, 2014 9:55 pm

This comes up from time to time. The TL;DR version is that long bonds and their funds like TLT lag the other two major asset classes in volatility. Most don't see this as a huge problem, but for those who do, there's are a few solutions: zero-coupon bonds or one of their funds, like EDV or ZROZ, for instance.
Human behavior is economic behavior. The particulars may vary, but competition for limited resources remains a constant.
- CEO Nwabudike Morgan
rickb
Executive Member
Executive Member
Posts: 762
Joined: Mon Apr 26, 2010 12:12 am

Re: Asset Volatility Question

Post by rickb » Fri Nov 07, 2014 11:08 pm

Mark Leavy wrote: The once and future melveyr (hallowed be his name) once talked about allocating portfolio positions inversely proportional to their volatility.  Seriously brilliant.  Except volatility is difficult to define.  I ran a few simulations using 10 year maximum draw downs as a proxy for volatility and the portfolios performed quite well.  Regardless of asset class.
Another poster (long gone, along with his posts), Clive, was all over this.  Here's a thread with a quote of one his (self-deleted) posts:

http://gyroscopicinvesting.com/forum/pe ... -weighted/

Last I heard, he still posts (same username) at BH.

I don't think it's any big secret that melveyr is Ryan Melvey, who blogs (only occasionally at this point) at stableinvesting.com.  As it turns out (also not any big secret) he's really, really young (maybe 23 now).  My guess is he'll eventually end up at some hedge fund, like maybe Bridgewater.  If I were Ray Dalio, I'd make him an offer he couldn't refuse.
User avatar
dualstow
Executive Member
Executive Member
Posts: 14225
Joined: Wed Oct 27, 2010 10:18 am
Location: synagogue of Satan
Contact:

Re: Asset Volatility Question

Post by dualstow » Sat Nov 08, 2014 7:28 am

rickb wrote: I don't think it's any big secret that melveyr is Ryan Melvey, who blogs (only occasionally at this point) at stableinvesting.com.  As it turns out (also not any big secret) he's really, really young (maybe 23 now).  My guess is he'll eventually end up at some hedge fund, like maybe Bridgewater.
...
Could be. I think he's presently at Amazon. I miss his stableinvesting posts.
RIP Marcello Gandini
User avatar
Gosso
Executive Member
Executive Member
Posts: 1052
Joined: Fri Jan 06, 2012 8:22 am
Location: Canada

Re: Asset Volatility Question

Post by Gosso » Sat Nov 08, 2014 8:37 am

Fun thread.
sophie wrote: Great chart Gosso!

Just one thing...how is the pain/gain ratio calculated?  Interestingly, the 60/40 appears to be very close to the PP or better for almost all the time except some periods in the 1970s.  I imagine that's the eras when gold was skyrocketing.  That's a bit counterintuitive given the performance charts showing the smoother, steadier performance over time of the PP.
Thanks!  The gain to pain ratio is the sum of the returns from positive months divided by the sum of the returns from negative months (can be used on daily data as well).  I thought it was neat that the ratio showed a large divergence between the PP and 60/40 over the past few years, and might be applicable to barrett's questions.

I think the two year rolling was a bit too noisy, so I extended the G/P ratio to five year rolling, and it looks much more clear.  I don't think it tells us anything new, but still interesting to see.

[img width=500]http://i62.tinypic.com/33krf5h.png[/img]
Last edited by Gosso on Sun Nov 09, 2014 8:54 am, edited 1 time in total.
Kbg
Executive Member
Executive Member
Posts: 2815
Joined: Fri May 23, 2014 4:18 pm

Re: Asset Volatility Question

Post by Kbg » Tue Nov 11, 2014 9:08 pm

Mark Leavy wrote: Apropos of nothing, I ran some numbers.  These don't really mean anything, but I thought I would calc them just the same.
(All numbers current to today - no inflation adjustments)

Using 10 year daily data, dividends reinvested (purple verticals) I came up with these numbers:

GLD: 9.72% CAGR, 40.52% draw down, 1173 days
TLT: 11.71% CAGR, 25.03% draw down, 762 days
VTI: 11.66% CAGR, 55.96% draw down, 1226 days

PP (15/35 bands) : 9.98% CAGR, 12.67% draw down, 512 days

You can clearly see that all of the three volatile assets are indeed volatile - and that the PP does a fine job of smoothing them all out.
It may not be optimum, and it is certainly not prescient, but it is a good bet.

GLD
[img width=600]http://i61.tinypic.com/346auu0.png[/img]

TLT
[img width=600]http://i57.tinypic.com/2vdjpe8.png[/img]

VTI
[img width=600]http://i60.tinypic.com/qxqtfr.jpg[/img]

Permanent Portfolio (15/35)
[img width=600]http://i58.tinypic.com/2wd7ue0.png[/img]
When the endless gold and long bonds suck conversation fires up again posters should study Mark's GLD vs. TLT charts.  I'll go out on a ledge here...just about the time long bonds go off the cliff folks who have been accumulating gold as it has cratered are going to be happy.
Post Reply