I made some graphs that help one get a better understanding of the PP from a modern portfolio theory perspective. We all know MPT has its flaws, but it is one of the few tools we have.
In each graph, start by looking at the colored dot which represents one of the PP asset classes. Each gray dot represents a movement closer to the PP by adding the other asset classes. Gray dots past the balanced PP (black dot) are increasingly underweight the colored asset class, past the balance of the PP.
What I find most interesting is that this analysis vindicates the equal weighting of gold, stocks, and Treasuries. You can see that the PP is on a relatively ideal part of the curve. However, the analysis also highlights that there is no clear "best" weighting of cash. There is a relatively linear risk/reward relationship for including more or less cash. This leads me to conclude that if one is considering tweaking the PP to meet their risk/return objectives (whether it be more or less risk), cash should be the first place to look. I am happy with the 25% weighting of cash personally, and I probably will not remove it from my portfolio because I like protection from Fed-induced recessions (Volcker and 1981).
everything comes from somewhere and everything goes somewhere
It seems to me theres also a wee bit too much gold and bonds.
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes
Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet. I should not be considered as legally permitted to render such advice!
MachineGhost wrote:
It seems to me theres also a wee bit too much gold and bonds.
The question then becomes where does that part of the portfolio go? Also, does a decrease in these two more volatile components change how one rebalances?
“Let every man divide his money into three parts, and invest a third in land, a third in business and a third let him keep by him in reserve.� ~Talmud
I have no problem with MPT when it is used to analyze the past as long as the assumptions are well known. (For example, using std deviation on portfolio balance as the measure of risk.)
Today I have 39.9% of my assets (everything liquid but the checking account) in the PP. Call it 40%. Therefore stocks, gold, cash, lt t's should each be 10% of my assets. However, including the other 60% (the variable portfolio) results in a total composition of:
(hopefully that adds up to 100% +/- reasonable rounding error)
The st-t's are actually 'SHY' in my PP account as part of the cash, so counting those as cash means I'm overall about 22.7% cash, or darn close the PP recommended amount.
If I interpret your graphs correctly, I could boost my returns by adding more lt-t's. Ugh. Tough pill to swallow. I don't think I can do it. (I will be adding a bit more if they drop much more... that will trigger a rebalance.)
And perhaps I'll be as surprised by lt-t's as I was last year... But how can they go up and for how long?
I can live with a bit higher volatility. I actually haven't much changed my portfolio composition in about 18 months now. It's been doing great, even with the selloff in PMs a year ago. And now...
In my mind's eye I've tried to imagine a PP performance plot as a surface. I guess portfolios without cash could be shown as surface over a triangle with each corner being 100% stocks or 100% LTT or 100% bonds and all the points on the square would be some blend between them. The surface would have peaks and valleys representing CAGR and it would be coloured from blue to red according to std deviation. I guess it would be awkward to fit four assets into such a plot
Last edited by stone on Sat Sep 29, 2012 7:00 am, edited 1 time in total.
"Good judgment comes from experience. Experience comes from bad judgment." - Mulla Nasrudin
I think the graphs are great. However, you maybe drawing conclusions for the wrong reasons.
Yes, MPT with all its flaws is a good mathematical tool. The major problem is that the inputs need to be about the future(i.e. expected risk and return in the future). Having said that let me comment on the last graph, with allocations only to Gold, Stock and Treasuries. To take advantage of that you need to draw a tangent to the curve from the point of the CASH return and 0 STD. While even CASH may have some risk, for the most part and for the purposes of the graph, treating cash as having 0 STD should be adequate. It is the tangent that prescribes where you should be on allocation. This point may be before(lower STD) or after(higher STD) than the PP allocation point shown. It may even be possible to add leverage(although inadvisable in most cases) to get a higher return than the PP with the same STD. It all depends on the slope of the tangent.
I believe then that the actual allocation to Cash depends on your needs(incl emergency money etc), your time horizon and the degree of risk you want to take or conversely the amount of return you need while staying on the tangent. If you do that you may find that the allocation is actually pretty close to the PP allocation unless you have excessive cash needs or few cash needs if you are excessively wealthy. The Cash allocation will be different for different people depending on their personal needs.
The Classic (Markowitz) Modern Portfolio Theory assumes that the individual assets come from all possible stock investments and the Market Portfolio is the Total Stock Market return (e.g. VTI)
I'm leaning towards proposing a more generalized version where all individual assets are all reasonably liquid stocks, bonds, commodities, and real assets and the Market Portfolio = Permanent Portfolio. And someone with a high risk tolerance could (in theory) move up the tangent line and leverage the Market (Permanent) Portfolio.
