Mark Shipman in his book "Big Money Little Effort," describes a system for deciding when to invest and when to sell in a long-term trading strategy. He uses a 30 week and 50 week moving average, and advises a sale when the longer term MA crosses above the longer term and a purchase when the opposite cross-over occurs. As I show above, the system would have worked quite well in the period since 1990, warning in quite good time of all the Bears and signalling the Bulls. It would have missed some short term fluctuations, some of which I have boxed in the chart. In my previous analysis, due to inexperience, I drew MAs based on Monthly rather than Weekly periods, which would have been far too slow to be of use. Despite working reasonably well, there is, nevertheless, a time-lag, and the system does not flag a reversal in advance. The comment of Hillelzacay on my previous post is pertinent to the present time.
Komentar
HermanBrummer
⋅
Hi mate,
I think this is good research that you are conducting. I started off the same way, and spent 6 months just on MA's, between 6-10 hours a day.
Personally I found that standard MA's meaning SMA has significant lag built in, and while often giving good long entries, it mostly exits very late. Even HMA, WMA, and EMA which are better in some respects suffer from this problem, because they take too much data, which is not fresh into the account.
There is a better way. Usually indices, will no go up or drop more than 3% per day for around 96% of the time. Meaning the trade between a -3% to +3% band MOST of the time. You can run your own tests in excel to check this. This information is useful, because it means that instead of using long MA's with stale data, we could instead use much shorter time frames, and discount the price MA with -3%.
For example in test I have used an indicator I developed. Chris was kind enough to code it, the guy is a legend.
This is the long only indicator, on the same ticker as you used above:
This is the short only indicator, again based on the the same chart.
I think this is good research that you are conducting. I started off the same way, and spent 6 months just on MA's, between 6-10 hours a day.
I hope you find some of this stuff useful:
1. Video of Karen SuperTrader - she is the reason I got into indices, - youtube.com/watch?v=cXy9HoWX0es - she trades indicies - I think you might find her comments interesting
2. Good tool to run quick test to test the efficacy of the golden cross MA's - movingaveragebacktest.com/?symbol=SPX500&capital=30000&long=true&short=true&years=1&ma=30,50. I hope and think that TV will soon offer use something similar to run back tests. I found that just looking at charts can be misleading.
Personally I found that standard MA's meaning SMA has significant lag built in, and while often giving good long entries, it mostly exits very late. Even HMA, WMA, and EMA which are better in some respects suffer from this problem, because they take too much data, which is not fresh into the account.
There is a better way. Usually indices, will no go up or drop more than 3% per day for around 96% of the time. Meaning the trade between a -3% to +3% band MOST of the time. You can run your own tests in excel to check this. This information is useful, because it means that instead of using long MA's with stale data, we could instead use much shorter time frames, and discount the price MA with -3%.
For example in test I have used an indicator I developed. Chris was kind enough to code it, the guy is a legend.
This is the long only indicator, on the same ticker as you used above:
This is the short only indicator, again based on the the same chart.
I hope that gives something useful to look at,
H