a-tree-on-the-horizon

World’s Simplest Trading System


[NB. OK, it's possible that this isn't absolutely the simplest trading system imaginable, but apart from buy and hold it is unlikely there are many systems much simpler than this one!]

The following chart illustrates such a portfolio for the FTSE 100 Index since 1995. The diamond markers indicate the decisions made at the end of each month whether to be in the market (green diamond) or in cash (red diamond).

Roughly, one can see that the system kept the portfolio in the market in uptrends and out of the market (in cash) when the market fell. This trading system is well-known in the US, what we will look at here is:

  1. if the trading system can be profitably applied to the FTSE 100 Index.
  2. Whether 10 months is the optimum parameter for the moving average (or would a 5-month, or 15-month, moving average produce superior results)?

Terminology: we will use SMATS(10) to refer to the 10-month simple moving average trading system. And SMATS(5) for the trading system using the 5-month simple moving average etc. Below we will analyse the trading system for 14 different parameters of the simple moving average, i.e. from SMATS(4) to SMATS(16).

Performance Analysis 

First, let’s look at the overall profitability of SMATS.

Proftiability 

The following chart plots the values of the SMATS portfolios for the 14 different simple moving averages (i.e. 4-month to 16-month). As a benchmark the FTSE 100 is added (i.e. this is the value of a buy and hold FTSE 100 portfolio). All values were re-based to start at 100.

Some observations:

  1. By the end of the 20-year period all the SMATS portfolios had out-performed the FTSE 100 – except SMATS(5).
  2. By the end of the period, SMATS(10) had the highest value; although it can be seen that it wasn’t consistently the most profitable throughout the whole period.
  3. For the first six years (up to August 2001) all SMATS under-performed the FTSE 100. This was caused by the market volatility in 1998 and 2001, which caused the portfolios to be whipsawed in and out of the market.
The following chart summarizes the final portfolio values in 2016 after running the trading system from 1995.

By 2016 the STATS(10) portfolio had the highest value of all portfolios at 269; the FTSE 100 buy and hold portfolio a value of 199.

Risk 

We’ve looked at profitability, let’s now consider the risk incurred by each portfolio. We’ll use volatility as a (fairly standard) proxy for risk.

The following chart shows the volatility of the portfolios over the 20-year period.

Not surprisingly the FTSE 100 had the highest volatility. The volatility of the SMATS portfolios was less due to the fact they were in cash for part of the time; broadly their volatility increased as the moving average month parameter increased.

The Sharpe Ratio combines returns with volatility to provide a comparative measure of profitability per unit of risk incurred. The ratio’s purpose is to answer questions of the form: is the profitability of a strategy justified by the risk incurred, compared to another strategy?

The following chart plots the Sharpe Ratio for the 14 portfolios. (The benchmark for the Sharpe Ratio calculation was the FTSE 100 Index.)

SMATS(10) had the highest (i.e. the best) Sharpe Ratio, although close behind were SMATS(14) and SMATS(15).

Max Drawdown 

Maximum Drawdown decribes the maximum loss a portfolio suffered from a previous high value. For example, in this test SMATS(10) had a max drawdown value of 22.8%. This means that over the 20-year test period the portfolio was at most 22.8% under water (from a previous high).

Frankly, max drawdown has more significance for strategies that employ leveraged products (e.g. futures), as drawdowns incur realised losses as margins have to be paid. By contrast in the case of unleveraged equities or ETFs, drawdowns incur unrealised losses. Having said that, unrealised losses can still be uncomfortable and can have a major adverse psychological impact on the investor or trader.

The following chart shows the max drawdown values for the 14 SMATS portfolios and the FTSE 100 Index.

Here the SMATS(10) portfolio only had a middling relative score. The best portfolios (i.e. those with the lowest max drawdowns) were: SMATS(7), SMATS(14), SMATS(15), and SMATS(16).

Trade Frequency

The following chart shows the average number of trades for the year for each portfolio. For example, over the 20-year test period SMATS(10) portfolio traded 36 times, which is an average of 1.7 times a year.

As would be expected the number of trades decreases as the length of the moving average month parameter increases. In other words, systems get whipsawed less with longer moving averages. 

The profitability figures above did not include transaction costs, but with the systems averaging under 2 trades per year the transaction costs would not be significant.

Summary Of Analysis

The following table summarises the above analysis. The values are colour-coded with green being the best value through to red being the worst for each respective analysis.

Conclusion

  1. This simple moving average trading system did work for the FTSE 100 (i.e. it out-performed the FTSE 100 Index) over the 20-year period.
  2. The best performing portfolio was indeed SMATS(10), i.e. the trading using the 10-month simple moving average. It had the highest absolute profitability and also the highest Sharpe Ratio. After SMATS(10), the best portfolio was the SMATS(14), followed by SMATS(15).

 -- By  Almanacist from Stock Market Almanac

  • Matt haines says:

    Nice article! Simple systems can be useful in a portfolio mix.

    You have one typo that makes things a little confusing.

    “By 2016 the STATS(10) portfolio had the highest value of all portfolios at 269; the FTSE 100 buy and hold portfolio a value of 1999.”

    That last number should be 199. 🙂

  • […] Källa: World’s Simplest Trading System – System Trader Success […]

  • Dan says:

    Excellent introduction to system development!

    Would you please advise which software do you use for testing?

    Thanks in advance

    • Howard says:

      TradeStation all the way for me, I have used it since the late 90s. To have your systems coded and fine tuned fire me an email.

  • Darth trader says:

    It looks to me like as 3.3% CAGR for the FTSE100 over 20 years with the 10 month average system adding 3.3% CAGR (from table); so a 6% overall return. Not a very profitable system IMHO.

    • twicetrader says:

      It is a poor yield for sure (I would not write an article on this) but the worst factor is in CAGR/MaxDD. The rule of thumb for a robust trading system is that CAGR is greater than MaxDD. That is, you could expect to recover max loss in a year, not over 6yrs.
      I am astonished that we are still talking about moving averages and all this garbage called indicators. They are always lagging!
      You want to know the most simple TS in the world?! CAGR 9%, worst year – 6%. These metrics are for last 50 years! not 20.
      4 etf, stocks, bonds, gold and cash and you beat 95% of money managers and this kind of trading system.

      • Brian says:

        What trading system is this?

  • Sean says:

    Good article. I aways enjoy your material. Thanks for all you do and contribute!

  • >