Simple moving average (SMA)

The plainest of the moving averages. A window length is chosen, the closes inside that window are added and divided by the count, and the result is plotted against the last of them. As each period ends the newest close enters the window and the oldest leaves it, so the average moves. Every price inside the window counts equally, which is the single property that distinguishes it from its weighted and exponential relatives.

Equal weighting has a consequence that surprises readers watching the line. The average changes for two reasons at once: because a new price arrived, and because an old price dropped out of the window. A large value leaving the far end can turn the line even while recent prices are doing nothing at all, which means the shape of the line is not always about what the market just did. The lag is a function of the window, conventionally described as about half its length.

The conventions built on it are all bounded and all disputed. Price on one side of the average is used as a trend filter, a shorter average crossing a longer one is used as a signal, and the line itself is watched as an area where interest has appeared. None of that is derivable from the arithmetic, which computes a mean and says nothing about the future. Whether equal weighting or exponential weighting is preferable is a genuine disagreement, and the choice of window is the classic place where a rule is fitted to the sample that produced it.