The hardest part of decision-making is being able to separate information from noise. In a textbook investment problem, you are given useful information which you employ to form a probability distribution of possible returns. You take all of the information and find the expected value for an investment and make the decision.
The real problem is when you have too little or too much information. The issue of too much information is one that has always vexed investors but is usually not handled in an investment class. (Case studies may be the exception, but classic finance classes do not address noisy information.) Nevertheless, there are simple solutions that have always been used to deal with this problem. We just may not speak about it in terms of noise.
A large part of the value associated with trend-following is on noise reduction and price smoothing. Think about the simple case of the moving average. For example, you take twenty days of information and average to compress into one number. This is a noise reduction technique. All of the ups and downs from highs and lows as well as the daily movement is filtered out through averaging. We cut the noise to find the true signal.
Most system research is trying to find the right trade-off between noise reduction and information efficiency. If you take too long of a moving average you are throwing out too much information. Employ a short-term moving average and you still have significant noise.
Break-out or channel systems are another form of noise reduction. If you are inside the range, you may ignore the prices and throw out this noise. This approach provides a new twist on the concept of noise traders which have often been given a bad name.
Noise reduction changes the decision process and reduces the decision choices. This reduction process may calm decision-making for the better. Silence is good.
The real problem is when you have too little or too much information. The issue of too much information is one that has always vexed investors but is usually not handled in an investment class. (Case studies may be the exception, but classic finance classes do not address noisy information.) Nevertheless, there are simple solutions that have always been used to deal with this problem. We just may not speak about it in terms of noise.
A large part of the value associated with trend-following is on noise reduction and price smoothing. Think about the simple case of the moving average. For example, you take twenty days of information and average to compress into one number. This is a noise reduction technique. All of the ups and downs from highs and lows as well as the daily movement is filtered out through averaging. We cut the noise to find the true signal.
Most system research is trying to find the right trade-off between noise reduction and information efficiency. If you take too long of a moving average you are throwing out too much information. Employ a short-term moving average and you still have significant noise.
Break-out or channel systems are another form of noise reduction. If you are inside the range, you may ignore the prices and throw out this noise. This approach provides a new twist on the concept of noise traders which have often been given a bad name.
Noise reduction changes the decision process and reduces the decision choices. This reduction process may calm decision-making for the better. Silence is good.
No comments:
Post a Comment