Testing Six Investment Strategies in 2015
It’s not a good idea to switch between portfolio strategies all the time, as I already said in this post. But it can’t hurt to paper-trade strategies, i.e. to just look at how a strategy would perform without committing any actual money to it.
Since 2015 is upon us, now’s exactly the time to start tracking a few strategies. By the end of the year, we’ll have a full 365 day look at the strategies. While 1 year is certainly not enough to say that any strategy is better than another, it’s a great duration to become familiar with new strategies and maybe even decide whether or not they’re worth adding to your portfolio.
The portfolios I am going to track for the next 12 months are the following:
- the permanent portfolio (yearly)
- a risk parity portfolio (quarterly)
- a dual momentum portfolio (monthly)
- a trending value portfolio (yearly)
- an adaptive asset allocation portfolio (monthly)
- a simple 60/40 stock/bond portfolio (yearly)
I’ll be posting performance updates every month, including a performance and risk ranking (% gained and Sharpe), and portfolio charts.
Here’s a quick explanation of what the idea of each of these portfolios is:
– a permanent portfolio:
Based on Harry Browne’s book “Fail-Safe Investing: Lifelong Financial Security in 30 Minutes”, this portfolio is a composition of 4 asset classes, each weighed at 25% of the total portfolio: stocks / long-term bonds / cash / precious metals (gold)
Here is my setup for this portfolio (10000 EUR for every one of the 4 positions):
I’ve used European options except for GLD because US funds handle dividends differently (the US taxes foreign investors) so this setup would be better for a European like me than US ETFs would be.
– a risk parity portfolio:
Instead of an allocation based on value (i.e. 25% of the total portfolio invested in asset A, 25% in asset B) like the permanent portfolio, the risk parity portfolio tries to budget risk by having each asset class contribute the same amount of risk / volatility. That’s what this looks like:
This year’s version uses 3 asset classes (european bonds, gold, world stocks). Here’s the actual portfolio calculation. The weights of each position were calculated with a spreadsheet I found online. There are lots of them so it should be easy enough to find them.
I personally do not like allocating 20% of the portfolio to gold, but I wanted to reflect the current 3 major asset classes that Europeans have in their portfolio: some stocks, some bonds and gold.
– a dual momentum portfolio:
This is based on the Dual Momentum paper by Gary Antonacci. The basic idea is to define asset classes (4 in this case: equity, bonds, real estate, commodities) and look at a few ETFs in each class and then buy the one with biggest 12 month momentum if it has performed better than the risk-free reference asset. My risk-free ETF equivalent to US short-term treasuries is ERNE (iShares Euro Ultrashort Bond UCITS ETF). The ETFs I use are:
- EQUITY: CSUSS, DJSC, IJPE
- BONDS: IBGL, CSGBE7-EUR, HYLD
- REAL ESTATE: IUSP, IUKP, IWDP, IPRP
- COMMODITIES: CSGOLD, DBZN
The portfolio at the beginning of January looks like this:
– a trending value portfolio:
Trending value is a quantitative stock picking system described by O’Shaughnessy in his book “What Works on Wall Street”. The basic idea is to screen stocks for low prices first (using Prise-to-Earnings, Price-to-Sales, etc.) and then buy the 25 or 50 stocks with the highest 6 month price appreciation. In my own implementation, I’ve added a Piotroski filter to keep out stocks of low quality. Since I already use this method in my own portfolio, I’ll just use my own stock portfolio here instead of setting up a new portfolio for this. The current positions as of January 3rd are:
– an adaptive asset allocation portfolio:
Described in this paper, adaptive asset allocation uses 10 asset classes (I will use European ETFs), takes the Top 5 of those by 6-month return and constructs a minimum variance portfolio that is rebalanced on a monthly basis.
The 10 ETFs I will use are here:
- CSUSS – iShares MSCI USA Small Cap UCITS ETF
- DJSC – iShares EURO STOXX Small UCITS ETF
- IJPE – iShares MSCI Japan EUR Hedged UCITS ETF
- IEEM – iShares MSCI Emerging Markets UCITS ETF (Dist)
- IWDP – iShares Developed Markets Property Yield UCITS ETF
- IUSP – iShares US Property Yield UCITS ETF
- CSGOLD – iShares Gold (CH)
- CSBGE7 iShares Euro Government Bond 3-7yr UCITS ETF (Acc)
- IBGL iShares Euro Government Bond 15-30yr UCITS ETF
- DBZN – DB Commodity Booster Bloomberg UCITS ETF
I use the Google Finance Portfolio feature to find out which are the Top 5 by momentum. The current Top 5, weighted for a minimum variance portfolio, look like this.
Weighted to get the maximum Sharpe Ratio, they look like this:
Close enough. I’ll use the Sharpe-optimised portfolio. I used a spreadsheet found online and adjusted it to fit my purpose (a 5 ETF portfolio with a minimum weight of 10% and a max of 30% per position). Here it is: 5ETF portfolio minimum variance calculation. The mathematics, while initially daunting, aren’t so bad after all. If you don’t understand any of the concepts, look them up online. (I don’t guarantee that all the calculations are correct, I’m not a mathematician or statistician.) So here’s the portfolio:
– a simple 60/40 stock and bond portfolio:
Pretty self-explanatory, 60% is allocated to stocks, 40% to long-term bonds. Again, European ETFs are used here for the same reason as before.
That’s it! Now let’s see where we end up by December 2015! If you have any questions, ask away in the comments!