- 1998: Volume 7, No. 3 A Moving Average Primer

Moving averages come in various flavors. They range from simple, to weighted, to smoothed and exponential. Traders use them alone or in combinations as crossovers, even triple crossovers. They use them in oscillators as a moving average convergence/divergence, and in bands. The basic underlying assumption here is that markets move in a trending fashion more often than not.

Not everyone believes this is a true statement. There is a contingent of academics who feel market movements are random in nature (the "Random walk theorists"). I think you can fairly easily prove to yourself that "random walk" is bunk. Just look at any chart of any commodity of at least six months in length. You will see the trends unfold before your eyes. (Just keep your eyes peeled since that's when they are the best!) When demand for a particular commodity, or a financial asset, is stronger than supply, prices (and therefore the market) will move in an uptrend. When supply is overwhelming demand at any particular point in time, the market will trend downward.

Markets will trend up or down, but they can at times move sideways as well. An erratic up/down type affair, a trendless market, will wreak havoc temporarily with any trend-following system. These are the periods you need to use your discipline and patience to persevere.

The good news is, I've found markets are engaged in up- and/or down-trends for longer periods of time, and more often, than they are in sideways trends. This is why moving average techniques put the odds in your favor.

Incidentally, it is easy to determine trends after the fact by looking at a chart (see figure 1 and figure 2 , for instance). It is not all that easy in the thick of battle. The news, fundamentals, will not help you at all, since the news is always most bullish at the top and most bearish at the bottom.

Figure 1

Figure 2

The Simple Moving Average

The simplest moving average is fairly easy to construct and it's called (surprise) the simple moving average (SMA). It can come in any number of days (selected by the trader). Here's the formula:

SMA = (P1 + P2 + P3 + ... + PN) / N

P is the price of the commodity being averaged and N is the number of days in the moving average.

The value of an SMA is determined by the values that are being averaged and the time period. A 10-day SMA shows the average price for the past 10 days. A 20-day SMA, the average price for the last 20 days, etc. Moving averages can be calculated based on opens, closes, highs, or lows, and even the average of the day's ranges. I recommend only using the close or the settlement price for each day. I feel this is generally the most important price of the day, since this is the price used to calculate margin calls. If the market closes on the high, or in the high range, most of the short players (unless they shorted right at the high(s)) will have funds transferred out of their accounts. This makes the shorts a bit weaker and the longs a bit stronger, at least for the next day.

Let's construct a five-day SMA of crude oil based on the closing prices. Assume the closes the past five days were 2105, 2110, 2115, 2120 and 2125. The five-day SMA is 2115. If on the sixth day, the market closes at 2155, it will cause the five-day SMA to rise to 2125, the average of the last five days divided by five.You always drop the oldest closing price, and add today's closing price. In other words, when using the five-day, you would always drop the sixth (oldest) day. With a 10-day you drop the eleventh (oldest) day, etc.

5 SMA day 1 = (2105 + 2110 + 2115 + 2120 + 2125) / 5 = 2115

5 SMA day 2 = (2110 + 2115 + 2120 + 2125 + 2155) / 5 = 2125

The direction of the trend is determined by the direction the SMAs are moving and by comparing today's settlement price with the moving average. In the simple example above, the trend is up since the close on day 2 (2155) is higher than the SMA (2125). On day one the close was 2125 and the SMA was 2115 so the trend is still up and we are long. On day two the close was 2155 and the SMA 2125, so the trend is still up and we stay long. When the average turns down and under the closing price, it generates a sell signal.

You can connect each day's MA value on a chart and this produces a line. You can chart this line right onto a price chart to generate trading signals. As long as the line on any particular day is under the closing price, the trader would stay long; the trend is up. Once the line crosses over the closing price, the trader would go short; the trend has turned down. If long, and the line crosses over the closing price, the trader would reverse the position by selling double the amount of contracts owned. The problem is, if you do this every time the line crosses price, especially when using shorter-term averages, you can get whipsawed (bounced back and forth with small losses and commissions eating you up). Many times a market will trade in a wild range, moving wildly up and down in the same session, but as I've mentioned before, I believe the closing price is the most significant.

As a simple rule when using MAs, I suggest waiting for the close to penetrate the MA to generate a signal. Even better, you should wait for a "two-day close" (two consecutive days of penetration of the MA on the close), to signal a change in trend and therefore a change in position. Of course, by watching for the close you do take on additional risk, since in a volatile market, or a wide-ranging day, the price at the close could be far under or above the MA. If you waited for the close on Black Monday (the day the Dow closed 500 points lower than Black Friday) you were in deep water. This is why it additionally makes good money management sense to always use an ultimate down and out point (in other words, a stop) for any position. You should do this when using any technical system, or fundamental for that matter. When using an MA, place your stop at some point away from the MA to create an approximate maximum percentage loss for those few abnormal moves that happen each year. Abnormal moves are quite rare, but they do occur, and these are the ones which have the potential to cause a catastrophic loss. Rule 1 is to always avoid the catastrophic loss (that's the one so big it renders us unable to continue trading).

By the way, option strategies can also help here. Use puts to protect long positions, and calls to protect short positions. However, most markets are normal, and in normal markets here is the general rule of thumb: on the close, if the market price has fallen below the moving average line, a sell signal is generated; on the close, if the market price has risen above the moving average line, a buy signal is generated.

The signal can be taken at either the close of the signal day, or the open of the next day. I prefer the close on the day of the signal, since it has been proven that when a market closes at an extreme of a day's price range (which many times will create the signal) it will follow through in the direction of the close the next day. So you need to know on a daily basis exactly where the average is coming into the close, and then use a "stop close only" order. Alternatively, you can place a market order just prior to the close. Make sure your broker is ready for this and has good floor communications.

There will be times when the violation of the MA could be a close call. On days like these, I would opt to wait for one more day, rather than risk taking a false signal.

Disadvantages of the SMA

Most of the traders who use moving averages use SMAs because they are simple. They are easy to calculate since all you need is a calculator, or pencil and paper-but I don't recommend them. This is because the oldest price has the same influence as the newest. Generally, the newest reflects current market conditions better than the oldest, but with the SMA they are weighted equally.

For example, suppose wheat is trading between 490 and 510 and the simple nine-day SMA is about 500, but there is one day of data when the price was 475. When this low number becomes 10 days old and is dropped, the SMA jumps. This could generate a buy signal indicating a new major uptrend. However, the market tone may not have changed at all. In other words, when one old piece of oddball data gets dropped, it has a tendency to jump the SMA. Many times this jump can be an overstatement.

How Many Days Should You Use In Your MA?

The length of the MA will greatly impact trading activity and therefore profitability. Some traders use five-day MAs, some 10-day, some 20, some 50. With very long-term traders, particularly stock market investors, the 200-day MA is popular. The length is an arbitrary decision, but the sensitivity of any MA is directly determined by its length. It is the length which determines how much time an MA has to respond to a change in price. It is a matter of "lag time." Very simply, shorter moving averages are more sensitive than longer moving averages. A five-day is more sensitive than a 10-day, and they are both more sensitive than a 20-day. The more sensitive an MA, the smaller the loss will be on a reversal signal. However, there will be a higher likelihood of a whipsaw, where a false signal occurs when a minor movement, which ultimately does not change the major trend, is enough to push the MA in the opposite direction of the settlement price, therefore resulting in a false change in position. It is false simply because the trader will subsequently need to reverse once again when the true trend reasserts itself.

Bottom line, it's important to use an MA which is long enough so it is not overly sensitive. On the other hand, if the MA is too long, the trader will tend to take too big a loss (or give up too big a portion of unrealized paper profits) before he or she is even aware of a trend change. A longer MA will keep you in a trade longer, thereby maximizing paper profits, but it can eat into realized profits because it moves too slowly. So, just like the story of the three bears, the MA cannot be too hot or too cold; it needs to be just right.

Just right, however, is not always that easy to determine, and can certainly change with market conditions. Voluminous studies have been done to determine which length is right for which market, but I believe these are useless simply because market conditions change for all markets. The silver market of the Hunt era is not the same silver market of today. Soybeans in a drought market act far differently from a normal weather market.

The Bottom Line

Moving averages are not all created equal and they are not the Holy Grail. However, if used systematically and consistently they will keep a trader on the right side of the big moves. They are a totally technical approach (rely on price only) plus it doesn't matter what market you use them on. They work in bull and bear markets, but will whipsaw the trader in a sideways or trendless market. This isn't as big a disadvantage as it may appear on the surface, since my experience says markets will spend more time trending than not trending. However, at times I have been involved in choppy, whipsaw markets that last for many weeks. If not properly capitalized, or if not disciplined, a trader can be wiped out or at the very least become demoralized and abandon the system. Usually he or she will quit just before the big move starts. Remember, you need to catch the big moves when using a moving average system or you won't win.

George Kleinman is the president of Commodity Resource Corp in the U.S. and has gained extensive experience from trading for more than 20 years on behalf of individual and commercial clients. He is the author of Mastering Commodity Futures & Options published by Financial Times Pitman Publishing. The publisher's order number is 800-462-6420. The book is also available through Trader's Library.