The simple yet powerful idea between all these relative strength strategies is to capitalize on the well documented momentum effect within and across asset classes.
In this short article, we will show an example of how this can be achieved efficiently.
Step 1. Determine your ETF universe
The definition of the universe is a key step in creating this kind of strategies. Indeed, this step of the process has to balance several requirements:
- The need to be comprehensive in the coverage of broad asset classes, while avoiding too many extremely correlated ETFs
- The need to limit the size of the universe, and carefully consider the volatility and behavior of the ETFs you include in your universe.
The universe we use covers the following asset classes:
- US Equities: 4 ETFs, including SPY, VB, and AMJ
- Advanced Markets Equities: 4 ETFs, including EFA, EWA and EWJ,
- Emerging Markets Equities: 6 ETFs (both regional and large countries ETFs)
- Real Estate Investment Trusts (REITs) and Quasi-Equity: VNQ and PFF
- Commodities: 4 ETFs covering the main sub-classes: DBA (Agricultural commodities), DBB (Base Metals), DBO (OIL), and GLD (GOLD)
- Fixed-Income: 4 ETFs, including SHY (CASH), TLT (Long Term Treasuries), and TIP (Inflation-Protected Treasuries).
- Inverse ETF: SH (Short S&P500)
Step 2. Determine a set of rules
Having a clear set of rules that have been tested over a long and diverse period of time has many benefits, the main one being that it removes emotion from the investment process.
In this article we will focus on a particular example, but more detailed model results are available here.
We will use the following rules:
- Twice a month (1st trading day and the 15th of the month), rank the ETFs in the universe based on their risk-adjusted returns (the latter is calculated as a combination of total returns and volatility over various timeframes ranging from 1 to 12 months)
- Invest in the Top 4 ETFs (25% each) that rank higher than the cash ETF. Note: For example, if the cash ETF ranks 3rd, then allocate 25% to the Top 1, 25% to the Top 2 and 50% to cash.
- Rinse and repeat every 15 days ( a monthly frequency lead to very similar results)
Below is the equity curve of the model over the past 10 years (initial investment: $100,000):
The table below gives the annual returns for the last 10 years:
This represents a CAGR (Compounded Annual Gross Return) over 27%, with a 17.6% volatility (to be compared to 20.6% for SPY over the same period)
Allocating a portion of capital to this kind of strategy can give a large boost to a portfolio's return and allow even the most conservative investors to participate in large trends as they develop and mature. You will also note that the underperformance relative to SPY in a handful of years (2012 being the most noticeable) is far more than offset by the large outperformance in almost every year. In this regard the 2008 return is particularly outstanding (+34.5%). We wrote an article a couple of years ago underscoring how this kind of strategy can mimic the performance of the hedge fund exposure that is not avaliable to the retail investor looking to replicate the asset allocation of big institutional endowments.
For the first two weeks of January 2013, the model held 4 positions, for a total gain of +1.64%. The 4 positions were:
On January 15, 2013, EWZ and EWA were dropped and replaced by the two new entries in the Top 4 (VB and ILF).
The outperformance of the strategy is quite consistent when the frequency, or the number of positions is modified.