Technical Analysis · Intermediate · 5 min
Why indicators lag
You can explain why indicators built from past prices always lag, compare the lag of an SMA and an EMA, and judge what that means for signals.
Always behind
Every indicator on your chart has the same blind spot: it cannot see the candle that has not happened yet. It is calculated from the prices that already exist.
That sounds obvious, but it explains a lot of frustration. Why the crossover always comes after the move has started. Why the signal looked perfect on the old chart and late on the live one. This lesson explains where the lag comes from and why it cannot be removed for free.
Built from the past
Most indicators are calculations on past prices. That makes them lagging indicators: they confirm changes after they have begun.
Take a 20-period simple moving average. When a new candle closes, the average includes that close plus the 19 before it. If the price suddenly turns, the new close is outweighed by 19 older ones, so the average turns only gradually. The longer the period, the more old prices it contains, and the larger the lag. As StockCharts puts it, the longer the moving average, the more the lag.
There are ways to reduce lag:
- Shorter periods contain fewer old prices.
- Exponential averages weight recent prices more, so they react faster than simple averages of the same length.
- Indicators of indicators, like the MACD's signal line, add another layer of smoothing, and with it more lag.
But every reduction has a cost. A faster indicator reacts sooner to real turns, and equally soon to random wiggles. In sideways markets, fast settings produce one false signal after another. Slow settings filter noise, but react late. Reviews of technical trading rules, such as Park and Irwin's, also warn that settings tuned to past data tend to look better in hindsight than in real time.
Lag is not a flaw to be fixed. It is the price of smoothing. The useful question is which trade-off fits your timeframe and style.
Four and six candles late
In the fictional chart, the price rises from 50 to a high close of 60.50 on candle 28, then falls to 51. The gold line is the 20-period SMA, the lighter one the 20-period EMA.
- The EMA keeps rising until candle 32, four candles after the high. By then, the price has fallen to 57.77.
- The SMA keeps rising until candle 34, six candles after the high. The price is at 56.30, about 7% below the high close.
- Both averages then turn down and follow the decline, always above the price.
The EMA was two candles faster. Neither saw the top. A trader using the turn of either average as an exit would have given back a good part of the last leg of the rise, which is the normal cost of waiting for confirmation.
Common mistakes
- Expecting indicators to predict turns. They describe what already happened.
- Shortening settings until signals look early. That adds false signals.
- Stacking indicators to confirm each other. They are built from the same past prices.
- Judging lag on old charts. In hindsight, the price turn is obvious, and the lag looks avoidable.
The takeaway
Indicators are calculated from past prices, so they confirm moves after they start. Faster settings lag less but give more false signals.
Check yourself
Quick quiz
4 questions. Instant feedback, nothing is saved.
-
1Tap the candle with the highest close.
Tap the chart.
The high close came on candle 28 at 60.50. Both averages kept rising for several more candles after it.
-
2Why does every moving average lag?
A 20-period average still contains 19 older closes when the newest one arrives. It can only catch up gradually.
-
3Which turned down first after the high: the 20-period EMA or the 20-period SMA?
The EMA peaked on candle 32, the SMA on candle 34. Weighting recent prices more reduces the lag, but does not remove it.
-
4What is the trade-off when choosing shorter, faster indicator settings?
Fast settings react sooner to real turns and to random wiggles alike. Slow settings filter noise but react late.
of 4 correct.