Failsafe:

Protection in extreme down markets and very negative conditions:

One can never fully protect a portfolio and still let it grow faster than you might expect from t-bills.  But one can try to anticipate potential faults.  We have implemented a number of protections. For example, after extremely high volatility days, as measured by the VIX index we reduce our market exposure, and put a portion of the money in the money market.  We have also implemented another such protection that is available specifically to our program. This program works in the short term projecting one single day and is very specific in what it does and how it works. When we project a buy signal for the next day and it fails. (the market goes down) there is usually a good reason because, in going down, the market had to overcome the positive influences that we had measured. When the market goes down more than about 1% it is very significant, and it should not be neglected.  Now if the next day's signal is an "Out" or "Sell" there is no problem since the signal is either neutral or is in the same direction the market moved. When the very next signal is a "buy" and the previous day's buy signal failed by over -1% it is a strong warning to get out and wait for another day. Since as we have seen under proof that the market will prefer to move in the same direction as the previous day. It is a caution worth heeding. It doesn't happen very often. Usually less than once a year. But it keeps us more in synch with the market movement and more protected against a snowballing market crash.  We have worked very hard to insure that every element that is part of this program actually makes sense.  Just because something seems to work in the time frame that you are testing, does not mean that it will work in other time frames. That is why it is very important that every element that is part of a stock market analysis program can be justified on its own merits.