Jun

8

 I have often walked down the moving average street, but I like to look at what number for the average elicits buying so as you get near it, you can hope for a nice change in the distribuion of subsequent changes. I like to stop and stare at the amount that the curent price is above moving averages of different length and look at the expectations that follow various amount above and below. The changes in direction of the moving average have also been of interest. And the first advisory service I ever bought in commodities was called 'the cumulative average'. 60 years ago I bought it.

Ed Stewart writes: 

In 2012 I applied a 10 - 20 moving average cross to VIX trading product as an example showing the propensity to trend downwards in those markets, do mostly to the massive contango effects that were even more severe at that time - I also noted that every single MA combination worked in a wide range. A guy has continued to track that simple MA cross in XIV (inverse VIX etn), and it has continued to work, often much better than "sophisticated" multi-factor systems. I have had a great deal of luck trading the VIX futures with a combination curve slope, moving averages, and my preference for getting a period after a (seemingly) failed breakout of elevated volatility.

My thought based on this is that if one has reason to believe a market has a great deal of trend persistence yet timing might still be an issue, the simple MA approach seems like a good or reasonable tool. It's not the tool or technique itself so much but the features of a market that count and define if an idea or tool might work.

On the distance from MA idea, I like to do a similar thing but use mid-point of an X period range or a point like open, close, or other specific time.

Gary Phillips writes: 

"It's not the tool or technique itself so much but the features of a market that count and define if an idea or tool might work."

Good point. Any technical information and inferences made from using this or related indicators reflect not a primary but a secondary process that involves compliance of the indicators with fundamentals and/or a cognitive bias. However, indicators that are derivatives of price, track price changes; and, if there is persistence (the future is like the past) they inevitably end up contributing to the myth that they are predictive.

Stefan Martinek writes:

I think the best tools/techniques "learn" from the market and use the data features in some way (e.g. market specific level of noise, noise "memory", etc.). This is why I never use MAs or anything that has MAs inside where we arbitrarily via parameter selection force our views on data. Good techniques are usually adaptive and ask data what parameters are preferred now.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search