Can a system be an OAK?

I've been student of trading approaches for stocks, mutual funds, and ETF's for a long time. It only takes one backtest to see the amazing results you could have experienced if you had only done X or Y consistently for 5 or 10 years. If an investor could earn 25% per year over 10 years, they would compound a total return of over 9 times the original amount invested ($10,000 would grow to $93, 132). Of course, the problem is that patterns in markets change and they can even change frequently.

Patterns and systems can be unpredictable. There are approaches, such as getting out of the market when the daily price drops below the average price of the market over the last 200 days (called "moving average") that can be amazingly helpful for one decade, then cost the investor loads of money over the next decade. Most investors just don't have much patience to wait out such a long period of underperformance.

That brings me to a key point with using a trading system: they don't remove the need to handle risk and loss. Using a system can be just as painful or even more so than buy and hold investing. When markets are falling and some system gets investors out they stand on the sidelines and jeer at the foolish buy and hold investors. They are quiet, however, about the previous 10 times they got out and the market rallied higher leading them to miss out on some big gains.

The most promising systems, in my opinion, are simple and don't rely on too much magic. For example, some types of investing rely on charts that have pictures, such as cups, handles, candlesticks, and so on. You are better off testing something that is purely quantitative.

Momentum investing is a better opportunity. You could create a very short list of investments such as ETF's that represent US large caps, international developed large caps, and emerging markets and buy the ETF that has the best 12-month return. Then you could re-evaluate the ETF's at the end of each month.

If you ran that backtest with the ETF's SPY, EFA, and EEM, you end up with a return of 9.17% per year from Jan 2005 to the end of May 2018. If you made a portfolio of all three in a fairly traditional allocation (60% SPY, 25% EFA, and 15% EEM) you would get a return of 8.05% from the same period. 1% per year doesn't sound like much, but over time it can give you a significant advantage.

You will have to endure some periods of underperformance when you use any system and risks will probably be higher than a buy-and-hold approach.