A Unique Sector Rotation Strategy

0 Posted by - October 30, 2016 - Uncategorized

The investment strategy of sector rotation is not new. The economy moves in cycles and certain types of businesses should perform better than others at different points in time.

economic-cycle-smallA new and booming economic cycle might have jewelry companies doing better whereas electric utilities might out-perform during a recession. You get the idea.

Below are the sector exchange-traded funds that we will be using for this strategy as well as the sector they represent.

    1. XLB – Materials
    2. XLE – Energy
    3. XLF – Financial
    4. XLFS – Financial Services (new)
    5. XLI – Industrial
    6. XLK – Technology
    7. XLP – Consumer Staples
    8. XLRE – Real Estate (new)
    9. XLU – Utilities
    10. XLV – Healthcare
    11. XLY – Consumer Discretionary

For the sake of historical backtesting, we will exclude the 2 newcomers. If we were to hold the 9 main sector funds equal-weight since 2002, this would be our return.

equal-weight-sector

The red line represents our static sector fund strategy. It beats the market by a little because some sectors are not equal weight in major market indices. For instance, less than 3% of the S&P 500 index is made up of Telecom stocks while Healthcare makes up almost 15%.

But what about a different type of sector rotation strategy?

The creation of sector ETFs allows a unique glimpse into the investor mindset. Because each sector is made up with a different amount of stocks of varying sizes, it makes comparison difficult. But when we create these pre-packaged sector funds, we can analyze these funds side by side – apples to apples.

I have noticed the following anomalies as regards sector funds:

  • Lesser traded sector funds have higher annual returns
  • Lesser traded sector funds have shallower crashes

A simple strategy where we hold the 3 sector funds with the least amount of trading activity results in the following returns.

top3sectorfunds

An additional 2.7% annual return and portfolio drawdown of -41% vs -53% of the market…this is impressive! Why might this be the case?

The short answer is, I don’t really know. My guess is that the sectors which trade the least are not very exciting and defensive in nature. When oil is rocketing or plummeting, investors will aggressively trade the Energy sector ETF. I believe that the more speculation there is in a sector, the more volatile it will become and the less long-term performance you will receive. Just stick to the boring old workhorse that plods along and you are often rewarded handsomely for  it.

Another simple, and I believe related, factor to screen for is Beta. Beta refers to how close the fund trades to the market. Basically, if the fund goes up 10% when the market goes up 5% you have high Beta. If the fund drops 10% when the market goes up 5% you would have negative Beta.

Holding the 3 lowest Beta sector funds would have the following results:

lowbetafunds

The performance is similar to the low turnover strategy but with even better bear market performance.

The third rule that I use which works well with the other two is to look for decreasing short interest. Short interest is where investors short a stock with the expectation that prices will drop. But when investors change their sentiment and buy back shares (close their short positions), we can conclude that opinion on this sector is changing for the better.

When we combine all three rules into a ranking system and hold the top 3 sector funds according to this strategy, we get the following results.

sector-ranking-strategy

If you want to use, copy, alter and save this ETF sector rotation strategy use the link provided. Of course, you need to be signed up to Portfolio123 first. If you have not already done so, please check out my first Backtest Investing post that discusses some essentials before you sign up. And if you decide not to use Portfolio123, you can still use these simple techniques to trade with whatever strategy you prefer.

No comments

Leave a Reply

%d bloggers like this: