Proper stop loss placement hinges on the ability to determine what is “normal” stock behavior and on the ability to define the meaning of “significance” in a decline. A drop below support would be a significant decline. Such an event should trigger a stop loss. Volatility adjusted stop losses are based on the fact that a price excursion that exceeds the laws of probability (given a stock’s measured volatility) for a given time period would also be significant. It would not be “normal” for that stock. Anticipation of these and similar events is necessary to the determination of an appropriate stop loss.

People often act as if they know that the future will be good to their stock (so they don’t use a stop, or they place it too far away). However, we live in the present, not in the future and most untrained people are uncertain and conflicted about the best place to enter the sell order. People are optimistic about the prospects of a stock when they buy it. Therefore they do not want to sell. They want to believe that they made a good decision when they bought the stock. They will therefore tend to set their stop-loss so that it is unlikely to be triggered. If a person sets the sell order low enough that it will be triggered only if a significant downturn occurs, the stock may be sold after a decline of 20% or more. The key word here is “significant.” Placing the stop so that only a significant decline will trigger it is a good idea, but what does “significant” mean?

For most people, the term is relative. It may mean a 20% drop for one person and a 7% drop for another. For us, it is an event that is “statistically significant.” That is, when an event occurs that is statistically improbable, it is a significant event. For example, say we flip a coin and estimate the odds of it landing on heads or tails. Obviously, a balanced coin will land approximately 50% of the time on heads, and about 50% of the time it will land on tails. What are the odds that it will land on its edge? Though such an event is possible, it would be a very rare occurrence. It would be extremely safe to bet that on the next toss it will not land on its edge.

Similarly, a stock’s fluctuation about its moving average can be expressed as a probability distribution. For example, an expert trader might make a statistical measurement of the stock’s price behavior and determine that its standard deviation is about.858 points (you don’t have to know what this means, just follow along), then he knows that it will be “normal” for it to vary by about 2 points within 100 days. How does he know? He knows that, in a “normal distribution,” a variation equal to about 2.33 standard deviations occurs about 1% of the time and that (2.33 x.858 = 2). The fact that a deviation from the norm that is equal to 2.33 standard deviations occurs about 1% of the time is true of all “normal distributions,” regardless of the magnitude of a standard deviation. By adjusting the amount of deviation from the norm you will tolerate, you also adjust the probability of a stop loss being executed. The result of all these computations is a bell curve. The shape of this curve is the same for all normal distributions. This is a fact of mathematics, just as Pi is always equal to 3.14592…

Thus, you can set your stop-loss at such a distance from the stock that there is only one chance in a hundred that it will be triggered because of the stock’s normal fluctuations, or you can set the odds at one in two hundred or at some other probability level. First, it is necessary to determine how much “noise” or random fluctuation there is in a stock’s behavior. Then we place the stop just outside the probability envelope of the “noise” generated by its normal day to day fluctuations. Such a setting assures us that if the stop is triggered, it is because of a price surge that is not normal for the stock. You can pre-determine just how abnormal a surge will have to be to trigger a sale.

Though market behavior is not strictly “normal,” it is close enough that we can make useful estimates of probability. Setting your stops on the basis of mathematical probabilities enables you to distance human emotions from the decision process. You don’t have to agonize over whether the stop is being set too far away or too close. You can have confidence that you have placed it where it should be placed, and that if it is triggered it is because the stock’s behavior was abnormally deviant, beyond the realm of probability, and beyond your pre-determined comfort level.