Method Pt 1 : The When

Many years ago I read "Secrets for Profiting in Bull and Bear Markets" an excellent book by Stan Weinstein.  It contained practically everything anyone would need to make a decent fist of investing or trading so naturally I gave it a quick try and then moved on to the next thing when a small correction delivered me a string of losses.  One of the things I realize now is that for many years I did this and in hindsight practically ANY of the various things I tried then dumped would have delivered me a much larger bank account right now if I had only persisted with them.

In his book Mr Weinstein uses the concept of having a 30 week moving average to determine if a market is in an up trend or otherwise.  Similarly Dave Landry uses a combination of 10, 20 and 30 day moving average  in his book "A Layman's Guide to Trading Stocks"  which is another excellent investment book.

Even though I understood the concept of why it might be useful to know this it wasn't until I read Nick Radge's "Unholy Grails" that I finally understood the power of it.  In "Unholy Grails" Nick Radge discusses eight different methods of trend trading.  The mechanics of each one is described in detail and then they are tested over around 13 years of real market data to see how they perform on the ASX All Ordinaries group of stocks.  Crucially, he then tests them again but uses a 75 day moving average of the All Ords as an on/off switch for the trading system. So the system turns on and buying is allowed when the All Ords is above it's 75 day moving average and turns off, meaning no new entries and going defensive on existing held stocks, when the all Ords is Below it's 75 day moving average.

Generally the overall returns when using the index filter are smaller than when you leave the system on all the time - so what's the point?  The filter makes more sense when you measure returns, not just on absolute terms, but compared to how large a draw-down you might incur on your way to the target, and, what the standard deviation of returns in the testing was.  So an unfiltered return might be 33.59% p.a. compared to a filtered return of 29.86% but the draw-down trading the unfiltered was nearly 44% vs 33% on the filtered.  That is a massive difference,  especially when you are sitting staring at your depleted bank balance wondering if you have the intestinal fortitude to keep trading a system that is actually working perfectly well.  Personally I am a fan of using a filter - I know that sometimes I will be sitting in cash for a year or so but I am perfectly comfortable with that.

So what should you use?  Again, it depends on your own risk appetite and personal likes and dislikes. Each method has pros and cons, generally the less whipsaws in the switch the later you are to join the party and run away when the fun stops.  As an example, this is the All Ords with a 30 week (red) and 75 day (green) moving average.  You can see clearly how the red line doesn't flip on and off as often but tends to be slower to react to the big trend changes by at least a few weeks.  Also, remember you need to be looking at the index that represents the actual stocks you are trading.

Chart courtesy of www.IncredibleCharts.com