2005-2011: Effect of initial allocation

[Greg]

Interesting work here. Couple of questions:

1.) What program software did you use for this? Simba?

2.) In your upper plot you have Max values as 27.8%, 28.8%, 31.2%, 24.6% versus in the lowest plot you have the max as being 30%/26/20/24. Are these supposed to be the same values for comparing the difference between the max CAGR vs. min CAGR values?

3.) We had the giant run-up in gold during the last few years, would these percentages then have to all be taken with a grain of salt? We could say what the best percentage (or worst) in the past for returns, but do these results help us to change the initial allocations of 25% across the board? Also if this is only since 2005, that might not be a huge window of data to look at.

4.) I'm wondering if this window of data was expanded, to see what ideal percentages would look like for perhaps going back to the 1970s or 60s to get at least 50 years of data. Again though, any of these percentage values are only truly useful for previous data and not for the future. It's difficult to state which percentage allocations are ideal for the future except for stating we believe there are more times of prosperity than there are recessions so weighting a little more towards stocks and less towards cash. This can go into the VP-territory however.

[Greg]

As a potential alternative I recently looked at how often each asset was the best performing asset each year and came up with figures of around 45% for stocks, 20% for each of precious metals and long dated treasury's, 15% for cash.

I was thinking about something like this for my VP. It would seem that you are investing based upon the probabilities of the various assets appreciating more than the others. Even though stocks are quite volatile, if they were the best performers 45% of the time (as a sidenote, maybe I missed it but what time frame was this for?), that it'd make sense to weight stocks to 45%. I'm still young and interested in taking on more risk so something like this might be of interest for me. I'd have a regular PP at around 15% in each category and then the VP would be 40% made up of 5% LT, 5% gold/silver, and 30% Stocks (probably a bit of VTI, VWO, and perhaps VEA/others)

[Greg]

As a potential alternative I recently looked at how often each asset was the best performing asset each year and came up with figures of around 45% for stocks, 20% for each of precious metals and long dated treasury's, 15% for cash.

I was thinking about something like this for my VP. It would seem that you are investing based upon the probabilities of the various assets appreciating more than the others. Even though stocks are quite volatile, if they were the best performers 45% of the time (as a sidenote, maybe I missed it but what time frame was this for?), that it'd make sense to weight stocks to 45%. I'm still young and interested in taking on more risk so something like this might be of interest for me. I'd have a regular PP at around 15% in each category and then the VP would be 40% made up of 5% LT, 5% gold/silver, and 30% Stocks (probably a bit of VTI, VWO, and perhaps VEA/others)

UK yearly data since 1800 for stocks, silver, long dated gilt and cash (which I assumed to equal inflation)

Thanks Clive. Did you use simba or some other software to determine this? I'd be interested in doing this for US data for the past 100 year or so.

[Greg]

Clive,

Thanks for the data. It looks like you are correct in your counting. I'd be interested in seeing up until 2011 data as well but not sure if you have that or not. Otherwise I could probably find it elsewhere.

From those 80 years:

Stocks: 41 years at top -->    41/80 = 51.25%
Silver: 17 years at top -->      17/80 = 21.25%
LT Bonds: 12 years at top --> 12/80 = 15.00%
ST Bonds: 6 years at top -->    6/80 =  7.50%
Inflation: 4 years at top -->    4/80 =  5.00%

This is slowly looking like a Bogelhead portfolio with precious metals added in of 50% in Stocks, ~25% in Intermediate Bonds, ~25% in Gold/Silver for Inflation "protection".

Does this seem to make sense?

I guess the only other problem with this method is it purely relies on history and doesn't necessary state anything about if you had this type of portfolio how it would do in the future. We could have sustained deflation in the future which would be great for LT bonds. Ideally these values would stay the same over time but I suppose we have no idea. Wondering if it is worth having a regular PP and leaning towards something like these percentages with the VP.

[Greg]

I guess the only other problem with this method is it purely relies on history and doesn't necessary state anything about if you had this type of portfolio how it would do in the future

If you'd started in 1945 based on averages since 1926 (20 year history), and updated the averages each year (21 year history, 22 year....etc), it would have looked something like



Excel spreadsheet -> http://www.jfholdings.pwp.blueyonder.co.uk/1nv35t0r.xls

average of all of those were 56% stocks, 16% LTT, 11% STT, 17% Silver
with 3.6%, 1.5%, 2.4% and 5.1% respective standard deviations - so relatively consistent

Those were some very interesting plots. Thanks for the excel sheet Clive. So looks like overtime that stocks are more volatile but still do provide the greatest amount of return over time.

So this is looking like what I was saying before of 50% stocks, 20% LT bonds, 10% ST Bonds, and 20% Gold/Silver. It's banking on prosperity (which would be nice if that happens in my lifetime). I'll ponder this and perhaps state my VP portion of my portfolio is all stocks (emerging markets, small cap, and international developed) and perhaps a bit of silver.