I have tried modeling this portfolio using numbers from peaktotrough.com. The numbers don't match the Simba numbers exactly, but it's good to get a second opinion.
Rebalancing with the top two performers (50:50:0) from the previous year gave a CAGR, over the 42 years from 1/1/1973 to 12/31/2014, of 12.3%, with six down years; over the last 10 years it has been 11.6% (one down year). For fun, I considered a version where 10% of the worst-performing asset was also added (45:45:10) for hedging and to make it more PP-like: here, the CAGRs over 42 years (five negative years) and 10 years (no negative years) have been 11.0 and 10.7%, respectively.
azmat9, it looks to me like you have averaged your annual returns, rather than compounded them. I think that's why you are getting such a large CAGR of 14.2%.
Thanks. I did do the average annual return and mistakenly used this for CAGR. Overall, this seems like a good strategy, but the potential drawdowns in 1981 may be too much to handle.
Perhaps this is a stupid question but I am curious...how did you come up with a 32.1% performance for 2009 for this strategy? If you had picked the best two assets from 2008, you'd have picked gold and LTTs. In 2009, gold returned a little less than 25% IIRC and LTTs (the ETF TLT) returned -21.8%. Averaged out (since this strategy would use 50/50 of each), that comes out to roughly 1.7%. That's nowhere near 32.1%...it's off by more than an order of magnitude.
No, you are correct. Some of my values after 2000 were still off and I think it was happening from the copy and paste into here. I have manually corrected them as below. The annual return is at 14.1% w/ the new values -
So it looks like azmat's numbers are slowly getting corrected and maybe we can start drawing some basic conclusions. What jumps out for me is that following the "Azmat II" strategy, an investor is largely in step with long-term bull markets. I agree with his original thinking that the first few years should probably not be considered (people for all practical purposes were not free to buy and sell gold prior to 1972, so we don't really know what the free-market starting point might have been). If we start with the 1981 data, it looks like one would have been in stocks 17 of 20 years between 1981 and 2000, and then in gold for 12 of 13 years between 2001 and 2013. Put another way, this strategy would seem to take advantage of long-term momentum while also putting an investor at risk of having 50% of their assets in a winner when it at last comes to a screeching halt (see gold in 2013).
Considering how crucial sequence of returns is for someone close to or in retirement, it's not an old person's game, but might it have merit for someone younger?
A big question for me is whether or not the future can be counted on to produce long-term bull markets. I think not but would be curious to hear what others have to say.
So it's a momentum based trading system that features infrequent asset-switching/rebalancing and only holds two assets at a time. These types of trading systems suffer from significant trading start date performance variability. For instance, if you started trading on February 1st using trailing 12-month returns instead of January 1st, the selected assets can be entirely different and this effect would ripple forward throughout the years completely altering the future performance. And since you are holding only two volatile assets for an entire year, that can lead to massive differences. The only way to fix this is to trade more frequently in an attempt to catch the momentum swings as they occur.
The PP suffers from this to some extent but not nearly as much because there's no asset switching. Everyone hits their rebalance bands at different times. Think about a person lucky enough to hit the 35% gold rebalance band at the 2011 peak vs. someone only hitting 34% and riding it down to where it is today. That can cause some fairly significant portfolio performance variability amongst PP holders but it tends to balance out over time because everyone is holding the same assets all of the time.
Last edited by iwealth on Wed Feb 11, 2015 9:02 am, edited 1 time in total.
This was a fun exercise to see what the results would be. The final CAGR is 13.1%. I believe the sharpe ratio would be slightly lower than that of the HBPP. I do think this strategy was lucky in certain years, especially in 2008 (if stocks outperforms bonds by a mere 3% in 2007, we have a loss of 18% instead of a gain of 13%). Like most people, what drew me to the HBPP was the high annual returns coupled with the low drawdowns. This strategy is nice, but the drawdowns could be high.
Kbg wrote:
There are variations on this as well. For example, you could have a core PP of say 15 or 20% per asset and then use momentum to over and underweight.
Have you made any progress on backtesting this yet? Still on the backburner over here.
"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!
Backtesting using an 80% HBPP and 20% Top 2 strategy yields the following results. Essentially, a 1% CAGR increase for the HBPP with the same STDEV increase.
Kbg wrote:
There are variations on this as well. For example, you could have a core PP of say 15 or 20% per asset and then use momentum to over and underweight.
Have you made any progress on backtesting this yet? Still on the backburner over here.
Interesting, but rolling my eyes at the two hucksters. 1992 isn't a backtest, its a joke.
I'm determined that a proper backtest will take all variables into consideration, including tax rates on the increased transanction costs of any modification to the vanilla.
"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!
I had a real good run but only say 10 years, with a low Sharpe ratio. I didn't have such a well defined system. Mine was to buy beaten down stock market sectors, and hold them till they went up enough. Rinse and repeat. It seems like today, sectors are more correlated, so I don't know if I could make it work again.