Use a Volatility Based or Volatility Adjusted Stop Loss
What follows is a discussion about stop losses and how standard deviation and Average True Range (ATR) can be used in reference to volatility. To go directly to information about our volatility-compensating stop loss calculator (a calculator that does not require that the user be able to compute standard deviations or the ATR), Click here
Stop Loss Concepts
Volatility based or volatility adjusted stop losses can reduce the frequency of premature and unnecessary sales. We think they are the best choice when there are no obvious regions of support to use as a reference. They are far better than using a straight percentage. They can allow just enough "wiggle room" so the stop won't be triggered by normal "noise" as the stock works its way to higher price levels.
The basic stop-loss is an order to sell the stock "at the market" if it drops to a certain price. When you buy a stock, you are relatively sure it is about to rise. What if you are wrong? The use of a stop loss is endorsed by nearly all successful traders and investors. It is one of the best tools a person can use to limit losses while letting profits run.
Experienced traders make it a rule to always enter a stop-loss order whenever they take a position. That's because they have learned not to invest or trade without a "safety net." It is the way they draw a line in the sand and say "this much of a drop I can tolerate, but no more." Use of a stop loss enables a trader to be on a fishing trip or otherwise occupied, and if the stock falls more than it "should," it is sold. When the stock declines to the stop price, the order changes to a "sell-at-the-market" order, and the stock is immediately sold at whatever the current market price is.
The order is usually good for a limited time or it is "GTC" (Good-Till-Canceled). Stop loss "At-the-market" orders are usually preferred over "limit" orders because a limit order can result in your not selling at all. That is, for a "limit" order to be executed, the stock would have to be at the "limit" price. When a stock is falling rapidly, it is generally better to get out than to quibble over price.
Traders and investors will usually want their orders to be activated somewhat below a price where the stock is likely to get support. For example, if the stock "bounces" every time it drops to $47, that means there is support at $47. Therefore, an experienced investor will place his stop loss below $47, because it will take an unusual amount of selling pressure to break through that support and trigger the stop loss. A trendline is also a region of support (because the stock tends to "bounce" off that line). If a stock is in a rising trend, and the trendline defining its rise is at $39, a trader might set the his stop loss order at $38. Inexperienced investors sometimes use a fixed percentage for stop loss placement. For example, they might place their stop 10% below their purchase price. The problem with that is that the market does not "know" or "care" where you bought a stock. If a stock falls X% below your purchase price, it means absolutely nothing to the market. The market is much more "interested" in its own patterns of support and resistance. The market "remembers" where support is. That is why a stock will tend to "bounce" when it declines to a level where support exists. By definition, "a line of support" on a stock's chart is where people are waiting to buy shares.
However, there are times when the pattern of a rising stock does not manifest any clearly defined levels of support that can be used to formulate a stop loss strategy. That is when an alternate approach must be used. Probably the most effective of these is to measure the volatility of the stock and from that measurement determine the maximum range of excursion likely for the stock over a given time. A common practice of traders is to set their sell order just beyond the limits of the stock’s probable range of excursion. This approach makes use of the laws of probability. That is, a sample of the stock's behavior pattern is taken and a statistical analysis is performed to see how much fluctuation is "normal" for the stock (this can be achieved without mathematical expertise on your part, so don't get discouraged). The stop loss is set outside the normal "lurch & jerk" range of the stock. Hence, the normal "noise" (random fluctuation) in the stock’s behavior is not likely to trigger an unnecessary sale. Such an event is, by definition, improbable. Only an event that causes the stock to fall an unusual (improbable) amount relatively quickly would do that. (We provide a means by which you can base the stop loss for a stock on measurements of the stock's volatility without having to perform the volatility calculations. We'll get to that later.)
For example, on 1/3/08 support for Valero (VLO) was at about $68.05. During the following 20 days, the stock dropped to $47.80. If a volatility based stop loss order were used, it could reasonably have been placed at about $67.75. Such a stop would have saved the investor from having a loss of more than 29%. The stop loss would have triggered an automatic sell order "at the market" when the stock fell to $67.75. The stock did hit that price, and then it continued to drop.
It must be acknowledged, though, that even if the stop loss were placed at $67.75, the stock may not have been sold at that price. Remember that the stop loss becomes a "market order" when the stock falls to the trigger price. The order would then be executed at the best price available for the next transaction. A frightening news item might be announced that causes the stock to gap down from above to well below the price of your stop loss order. If that happens, your stock will be sold at the next available price below the gap. At the bottom of this page is a link for more information. Click on it and you will be taken to a page with chart illustrations of various ATR settings.
Use common sense when you purchase. It is not wise to place a stop loss 10% below your purchase price if you purchase the stock when it is 12% above its support. Trending stocks tend to re-visit the rising support levels that define a rising trendline. That means your 10% sell order is likely to be triggered even if nothing of significance has happened. If a stock returns to its trendline, that does not mean it has made a significant change in direction. It is normal for stocks to do that. If, on the other hand, you buy when the stock is only 1% to 2% above support, your 10% stop loss will not be triggered unless the stock undergoes enough selling pressure to cause it to fall through all the support that exists immediately above your stop order. It is not easy for a stock to decline that much through significant buying demand, and it will have done so if your stop loss is triggered. Such an event would mean something very significant and very negative has happened. Under those conditions, you should want your sell order to be executed, and assuming you bought just above support, it would not be necessary for you to take a 10% loss in the process.
Psychologists have measured the dispersion of the frequency distribution of I.Q. scores in a population. They call this measurement the standard deviation. On the Stanford Binet intelligence scale, the standard deviation is 16 points. Thus, the "normal" range is plus or minus 16 points from the average. Since the center of "normal" is defined as 100, for this I.Q. test, a "normal" I.Q. will fall in the range between 84 and 116. Only 2% (2 people out of 100) have an I.Q. that is 2 standard deviations above average, or 132. The greater the number of standard deviations a person scores above 100, the more rare his score is. So, a person with an I.Q. that is 3 standard deviations above 100 would occur about once in a room of 1000 randomly distributed people. Similarly, if a person could measure the dispersion of the frequency distribution of stock price spikes from the norm, he could adjust his stop loss accordingly. For example, if he wanted to place it so that there would be only 1 chance in 50 that it would be triggered, he would place it 2 standard deviations from the norm. That is, he would apply a multiplier (the number "2" in this illustration) to the standard deviation measurement. If SD is one standard deviation, then 1 x SD would be subtracted freom the high low or close to have a stop loss with approximately a 16% probability of being triggered by random fluctuation. At the low minus 2 x SD, the odds of the stop loss being triggered by random fluctuation would be approximately 2%. For more on the probabilities of a stop loss being triggered, see Probabilities .
One of the most popular measurements of dispersion in reference to stock prices and volatility is the Average True Range (ATR). Many experienced traders use the ATR instead of the standard deviation for determinimg stop loss placement. The ATR is not the same as the standard deviation. It does give different results, but the investor can use a similar approach for determining where to place a stop loss as we used in the above paragraph. Once the ATR is calculated, a multiplier is applied to it (as we did above with the standard deviation) to adjust the probability of stop loss execution so that it suits the investor's tolerance for risk. Thus, our hypothetical trader might place his stop loss 2 ATRs below the highest low, close, or high reached by the stock since its purchase. Two ATRs below the highest high could be used to lock in most of any upward surge. On the other hand, a person might keep the stop loss 2 ATRs below the highest low to make it less likely that the stop loss will be triggered simply because of a random fluctuation. More aggressive traders will use even tighter stop losses. Some will place the stop order 1.5 ATRs below the highest close, or 1 ATR below the highest high, or whatever. These more aggressive traders have a shorter investment time horizon and are attempting to capture the gains from short-term surges in price (where the percentage gain is greater relative to the amount of time invested). The specifics of order placement are tied to the investment time horizon and the volatility of the stock. The investor must determine the kind of stock moves he wants to capture and how much of a pullback he is willing o tolerate before selling. When he decides how much decline he is willing to allow, if he is using a multiple of the ATR in his determination, he is also factoring in the volatility of the stock.
Think about it. Stop-loss orders are for the investor similar in function to the safety lines used by mountain climbers. Of course you are confident that the stock is going to rise when you make your purchase. That's why you bought the stock. Everyone makes mistakes! A stop-loss is simply your safety line if you are wrong. If you do not believe you can be wrong, do yourself a favor and stay away from the stock market. Nearly all of the top traders and investors incorporate the stop-loss as an integral part of their discipline. A trader who does not take such protective measures, is like a circus high-wire performer who abandons the use of a safety net. He is asking for trouble. If you do not use a sell order "safety net," and if your stock plunges while you are away from your monitor, or if you are too slow to act because of confusion or indecision (perhaps your indicators are unusually ambiguous), you could lose nearly all of the value of a position (margin traders could even lose more than what they invested). As we said before, top traders who cannot find appropriate levels of support in a chart as a basis for positioning a stop loss, will use volatility-adjusted stops, and they adjust them upward as the stock rises. A volatility-based stop loss automatically positions the sell order so that it compensates for the volatility of the stock and for the risk tolerance of the investor. Straight percentage stops cannot do that.This page is merely an introduction to the subject of stop losses. For additional detail and illustrations based on our own tool for computing stop losses Click here
All pages on this website are protected by copyright
1590 Adams Avenue #4400
Costa Mesa, CA 92628 USA.