melveyr wrote:
What if the PP dropped 50% in the middle of a year and happened to move back up to break even by Jan 1?
That 50% could have just as easily lined up with Jan 1. I wouldn't place too much emphasis on the Julian calendar. I think looking at intra year peformance is a great way to kick the tires. The PP is riskier than most people here think because they just look at the year end values.
The specific rules of the HBPP are not as important as the consistency of applying the rules. The HBPP is nothing more than a mechanical trading system based on a 25/25/25/25 allocation of stocks/LTT/cash/gold with 15/35 rebalancing rules. The fact that the HBPP only has to be traded every two or three years on average is just a property of the system.
By trying to "optimize" the performance of the HBPP over the last 40 years of fiat-currency market data, one runs the risk of curve fitting. A trading system that starts out fitting the past data like a mitten is turned into a trading system that ends up fitting the past data like a glove. Since the future relative performance of the four asset classes may not be the same as their past relative performance, optimization to the past might result in subperformance in the future.
A book I like that explains trading systems is
The Universal Principles of Successful Trading: Essential Knowledge for All Traders in All Markets by Brent Penfold (
http://www.amazon.com/gp/product/0470825804).
Applying the principles of this book to the HBPP would mean to choose the "factory default values" of 25/25/25/25 and 15/35 (or whatever you decide to use) and then backtest this system. If it works OK on past data and you feel confident of trying the system for the future (in other words, it either works or it doesn't work with its factory default values), then you assume the next trade will be twice as bad as the worst past trade. You have to be prepared for this type of surprise event taking place.
Only a philosopher can answer why you would assume future performance is only twice as bad rather than three or four times as bad as the worst past performance, but the idea is that markets change with time (usually). As an investor and trader, people have to be prepared to handle these kinds of surprise possibilities.
One can rebalance yearly according to the Julian calendar, monthly, daily (which may be difficult if the gold component is in coins or bullion in one's safety deposit box), every 500 days (which is arbitrarily chosen to make a point), or whatever. The idea is to do the rebalancing consistently according to a predetermine plan and to test it first to make sure it works on past data (with the proviso that the worst future performance might be a lot worse than the worst past performance). Be prepared to handle this possibility.
One of my approaches to managing this possible risk is to use strategy diversification. I will put no more than perhaps 40% of my investible funds in the HBPP on grounds that even though the HBPP worked OK in the past, it might stop working for some unknown reason in the future.
I also think of the HBPP as a form of index investing. The stock, LTT, and gold components used with the HBPP are really index types of assets. Since I believe in John Bogle and the advantages of index investing for a large percentage of my portfolio, I consider the HBPP to be a less volatile form of index investing relative to pure stock index investing.
