There have been quite a few suggestions recently to increase the cash portion of the PP for those wanting to reduce volatility. Conversely, one could decide to also reduce the gold portion (not that I am advocating this).
The question is then what should the rebalancing bands be. If cash is 35% of the port and gold is 15%, should the bands be 25%-45% for cash and 5%-25% for gold, or should cash have larger bands and gold smaller ones than -/+ 10%? I understand this is not traditional PP any longer but wonder what would make the most sense.
Rebalancing bands in non-symmetric PP
Moderator: Global Moderator
-
williswine
- Associate Member
- Posts: 49
- Joined: Mon Sep 13, 2010 6:23 pm
Re: Rebalancing bands in non-symmetric PP
One approach is to simply increase the cash allocation to the point at which you're comfortable and then treat a portion as a traditional 4x25 PP. For example, with cash at 40% of the total, the remaining assets are each 20%. Treat this as a 4x25 of 80% of your assets (with traditional rebalancing bands), i.e. rebalance if any single asset (excluding cash) is more than 35% of 80% (i.e. 28%) or less than 15% of 80% (i.e. 12%) of the total. With cash at 70% (and each of the other 3 assets at 10%) the 35/15 rebalance bands are 35% of 40% (14%) and 15% of 40% (6%).
From 1972 to 2011 a 70/10/10/10 mix (rebalanced annually) had a CAGR of 8.53% (with a std dev of 4.86%) as opposed to the 4x25's CAGR of 9.79% and std dev of 8.42%. Even at 85/5/5/5 the CAGR is 8.09% (with a std dev of 4.34%). Drawdowns of twice the long term std dev shouldn't be unexpected. If you want to have a reasonable assurance you'll never have a drawdown of more than 10% boost your cash to 70% (but over long periods of time you won't keep up with a 100% stock portfolio, and you'll be truly screwed in certain SHTF scenarios).
The question boils down to "what are more afraid of" - losing 20% in the short term or losing 70% or more (in a SHTF scenario where dollars become worthless)?
From 1972 to 2011 a 70/10/10/10 mix (rebalanced annually) had a CAGR of 8.53% (with a std dev of 4.86%) as opposed to the 4x25's CAGR of 9.79% and std dev of 8.42%. Even at 85/5/5/5 the CAGR is 8.09% (with a std dev of 4.34%). Drawdowns of twice the long term std dev shouldn't be unexpected. If you want to have a reasonable assurance you'll never have a drawdown of more than 10% boost your cash to 70% (but over long periods of time you won't keep up with a 100% stock portfolio, and you'll be truly screwed in certain SHTF scenarios).
The question boils down to "what are more afraid of" - losing 20% in the short term or losing 70% or more (in a SHTF scenario where dollars become worthless)?
-
williswine
- Associate Member
- Posts: 49
- Joined: Mon Sep 13, 2010 6:23 pm
Re: Rebalancing bands in non-symmetric PP
I see but I'm not sure to follow your 1st paragraph example: "With cash at 70% (and each of the other 3 assets at 10%)"...
Probably the easiest way to put it is to take 60% and 140% of the desired percentage for a given asset class to obtain the lower-band-adjusted and upper-band-adjusted percentages, respectively.
In other words, using my original example, stock and bonds would first be at 25% going down to 25%*60%=15% or up to 25%*140%=35% while gold would first be at 15% going down to 15%*60%=9% or up to 15%*140%=21% and cash would first be at 35% going down to 35%*60%=21% or up to 35%*140%=49%. Following common wisdom on this forum, I believe once any of these lower or upper percentages is reached, the portfolio would have to be rebalanced to 25%, 25%, 15% and 35%, respectively.
With that said and recognizing the above is not a traditional 4x25 PP, HB in my 1987 first edition of "Why the best-laid investment plans usually go wrong" wrote that "You need to make an adjustment only once each year" or in case something big happens (he gives the example of a 50% gain - or loss - in an asset class). That is contrary to what I recall reading here that you adjust only in the latter case, not necessarily each year. Perhaps HB changed his mind after this first edition. Don't know if that's the case...
Probably the easiest way to put it is to take 60% and 140% of the desired percentage for a given asset class to obtain the lower-band-adjusted and upper-band-adjusted percentages, respectively.
In other words, using my original example, stock and bonds would first be at 25% going down to 25%*60%=15% or up to 25%*140%=35% while gold would first be at 15% going down to 15%*60%=9% or up to 15%*140%=21% and cash would first be at 35% going down to 35%*60%=21% or up to 35%*140%=49%. Following common wisdom on this forum, I believe once any of these lower or upper percentages is reached, the portfolio would have to be rebalanced to 25%, 25%, 15% and 35%, respectively.
With that said and recognizing the above is not a traditional 4x25 PP, HB in my 1987 first edition of "Why the best-laid investment plans usually go wrong" wrote that "You need to make an adjustment only once each year" or in case something big happens (he gives the example of a 50% gain - or loss - in an asset class). That is contrary to what I recall reading here that you adjust only in the latter case, not necessarily each year. Perhaps HB changed his mind after this first edition. Don't know if that's the case...
Re: Rebalancing bands in non-symmetric PP
In my case I have a spreadsheet that shows my 4-assets and the % ratio of each. Right now I'm around 37% cash with gold in the teens.
Then I track the 3-assets without cash. This shows the volatile assets amongst each other in a 3x33 style Permanent Portfolio. I choose 27/40 rebalancing bands for that.
Right now I've got:
Gold 30%
Stocks 37%
Treasuries 33%
I'm tempted to rebalance soon.
Then I track the 3-assets without cash. This shows the volatile assets amongst each other in a 3x33 style Permanent Portfolio. I choose 27/40 rebalancing bands for that.
Right now I've got:
Gold 30%
Stocks 37%
Treasuries 33%
I'm tempted to rebalance soon.