The automated MTR-TM was down for a few days for a couple of reasons.
First, maintenance was performed on the site a couple of weeks ago and some of the historical signal data was updated and it was incorrect. There were 97 signals showing on the web site but there should have been 60 based on the results in AMIBroker. This impacted historical trading signals prior to 2007 and more recent signals may have been pushed by 1 day.
Second, one of MTR members, Kevin, discussed in the Forum that he wanted to run some tests using the signals and compare to S&P 500, NASDAQ, and other indexes.
Kevin performed some great work with Excel to check the MTR-TM signals vs. the S&P 500, NASDAQ, and other indexes and emailed the results. It was Kevin's email that prompted us to look at the data and saw the bug with the 97 signals vs. the 60. The trade signals from the MTR-TM did not correlate enough to the movements in the S&P 500 or NASDAQ to make holding ETFs based on the indexes a viable long term strategy using MTR-TM signals. The 4% Model, as with the MTR-TM, main purpose is to determine if overall the market would move higher or lower. Then based on these signals a trader could trade in that direction but it is clear just sitting on ETFs that fit the S&P 500 or NASDAQ 100 would not be the best call.
There was some work already being performed on the MTR-TM formula and the model on the site now represents the updated logic. Looking for a simple strategy to trade these signals it is clear a trader can use an ETF that closely correlates to the Value Line Arithmetic Index such as the S&P 400 MidCap ETF.
The Market Timing Model documentation was updated to discuss this in more detail. Please Section II and III.
We thank Kevin for his work in testing the MTR signals vs. other indexes. This helped to uncover a data issue on the web site and prompted additional research as to better trading methods using the MTR-TM signals.