Elder Ray: Bull Power, Bear Power and the slope of the average
Elder Ray, also called the Elder-ray index, draws two histograms side by side: Bull Power, how far each bar's high reached above an exponential moving average, and Bear Power, how far its low fell below it. Alexander Elder designed the pair to be read together with the direction of that same average, so the trend decides which side to trade and the histograms decide when. This page explains the calculation in words, how to choose the length of the average, the two misreadings that cause the most trouble, and what you can adjust when the strategy generator tests Bulls Power and Bears Power on your own contracts.
What it measures
Elder Ray is Alexander Elder's way of separating the two camps behind every bar. He presented it in his book Trading for a Living in the early nineties, treating an exponential moving average of closes as the market's consensus of value: roughly the price buyers and sellers have agreed on lately. The high of a bar is the furthest the buyers managed to lift price during that bar, and the low is the furthest the sellers managed to push it. Measuring both extremes against the consensus line gives two readings per bar. That is the core of the idea: instead of folding buying and selling pressure into one oscillator, the indicator keeps them apart, so you can see which side is stretching the market away from value and which side is giving ground.
The calculation takes three steps. First, compute an exponential moving average of the closing price over the chosen length. Second, subtract that average from the bar's high; the result is Bull Power, positive when the high reached above the average and negative when even the high stayed below it. Third, subtract the same average from the bar's low; the result is Bear Power, negative when the low dipped beneath the average and positive when even the low held above it. Both are usually drawn as histograms around a zero line. Because a high can never sit below its own low, Bull Power is always at least as large as Bear Power on the same bar, and the gap between the two is simply that bar's range. A wide bar spreads the pair apart, a narrow bar squeezes them together, and where the pair sits relative to zero tells you where the bar traded relative to value.
That geometry gives the readings their meaning. In a healthy uptrend, Bull Power stays positive while Bear Power spends most of its time slightly below zero; when a strong advance lifts even the lows above the average, Bear Power turns positive, a sign that sellers cannot reach value at all. In a downtrend the picture flips: Bear Power stays negative, and a negative Bull Power means buyers cannot even lift the high back to the average. Both readings are in the contract's own price points. The Elder Force Index, Elder's other well-known tool, is easy to confuse with this one, but it weighs each close-to-close change by volume rather than measuring the high and low against an average. The two share an author and a zero line; they answer different questions.
How it behaves in real markets
Elder's own rules start from the average, not from the histograms. The slope of the exponential average sets the direction: while it rises, only long setups count and Bear Power is the histogram to watch; while it falls, only shorts count and Bull Power is the one to watch. The classic long setup is Bear Power below zero but rising, meaning price has pulled back under value while the sellers' reach is shrinking. The mirror short setup is Bull Power above zero but falling during a downtrend, a rally that pokes above value with fading strength. Elder gave extra weight to divergence: in an uptrend, a new price low that comes with a shallower Bear Power trough suggests sellers are tiring. He also used the pair for adding to positions, for example adding to a long when Bear Power dips below zero and then rises back above it. Elder-ray is often described as part of his Triple Screen approach, where a longer timeframe sets the direction first, though it can also be used as a method of its own.
The main setting is the length of the exponential average, and it moves the trend filter and both histograms at once. A shorter average hugs price closely, so highs and lows rarely stray far from it: readings stay small, flip sign often, and the slope changes direction quickly, which suits fast intraday timing but produces more false turns. A longer average trails further behind, so readings grow larger, stay one-sided through extended moves, and the slope becomes a steadier filter that reacts later. Elder's original default was a fairly short average on daily bars. On intraday futures charts, choose a length that spans the swings you actually trade, and judge it by how cleanly its slope separates trending stretches from ranges on your own contract rather than by matching a figure from a book. Any level you compare the histograms with should likewise be set in points that suit the contract's current volatility.
Two misuses come up again and again. The first is taking one histogram alone, without the trend filter. Bull Power is positive on most bars of any rising market and on many bars inside a range, so a rule that buys whenever Bull Power crosses above zero simply buys every time the high pokes above the average, which happens constantly in sideways trading. Likewise, a deeply negative Bear Power looks like heavy selling, yet inside a rising trend it is exactly the pullback Elder wanted to buy. The histograms describe how far price reached from value; only the slope of the average says which of those reaches is an opportunity and which is a cue to stand aside. The second misuse is comparing raw readings across contracts. Because both readings are in price points, an index future quoted in tens of thousands of points produces Bull Power figures that dwarf those of a bond or currency future, and the same contract shows larger readings after its price level has climbed. A threshold copied from one chart to another, or from an early part of the history to today, means something different in each place. Relate the readings to the contract's own recent range, or to ATR, before comparing them.
Common ways to use it (concepts)
Bear Power rising below zero in an uptrend
With the average sloping up, a Bear Power reading below zero that turns higher marks a dip under value where selling is losing reach. This is Elder's core long setup; its mirror, Bull Power above zero and turning lower while the average slopes down, gives the short.
A shallower trough at a new price low
When price makes a lower low but the Bear Power trough is shallower than at the previous low, sellers pushed less far below value the second time. Read with a rising or flattening average, it suggests the decline is losing force; on its own it does not mark a turn.
When even the lows stay above value
A positive Bear Power means the whole bar traded above the average. A run of such bars describes a market that sellers cannot pull back to value, a condition some traders use to hold a position longer or to avoid fading the move.
Pairing it with other indicator families
The partner Elder intended is the average itself. A moving average of the same kind and length supplies the slope that turns the histograms from raw readings into setups. In a rule-based strategy the two pieces are usually written as separate conditions: one checks that the average is rising or falling, and Bull Power or Bear Power supplies the timing within that direction. Using the same length for both keeps them describing the same consensus line.
MACD is the natural contrast. Its histogram folds all momentum into one series built from closes, while Elder Ray keeps buyers and sellers apart and uses the extremes of each bar. When the MACD histogram is still rising but Bull Power peaks lower, the move is still advancing on closes while the highs reach less far above value, a detail the single series cannot show.
For the scale problem, ATR is the simplest companion. It is also measured in points, so expressing Bull Power or Bear Power as a share of the current average range makes readings comparable across years and between contracts. A trend-strength gauge such as ADX can also decide whether the slope of the average deserves trust, since a slightly rising average in a flat market says little about direction.
In AlphaPrime
In AlphaPrime, Elder Ray appears as two built-in groups in the strategy generator, Bulls Power and Bears Power, so a strategy can use either histogram alone or both together. Each group has an Indicator period N, the indicator's lookback, and its level and cross conditions also take a Threshold, the level the reading is compared with. A threshold at zero gives the plain above-value and below-value readings of Elder's rules; a threshold away from zero asks for a reading that reaches further before the condition holds. The Current-bar shift lets a condition read an earlier bar instead of the latest one.
Each group offers the same set of conditions. For Bulls Power they include Bulls power crosses up through the level, Bulls power is above the level, Bulls power is rising and Bulls power turns up, each with its downside mirror; Bears Power has the matching set, such as Bears power is below the level and Bears power turns up. Elder's pullback setup maps directly onto these words: Bears power is below the level with the threshold at zero, together with Bears power turns up. The two groups measure how far the high and low reach but leave the slope of the average to you, so adding Moving average is rising or Moving average is falling from the Moving Average group supplies the trend filter Elder insisted on.
The generator crosses these conditions with rules from other families and backtests the candidates on the GPU in parallel. You can set out-of-sample validation thresholds, so a histogram rule that fit only one stretch of history or one price level can be set aside before you go further. Strategies that hold up can be exported as PowerLanguage strategy code for trading platforms, so every Elder Ray condition can be read line by line before you simulate it.
FAQ
Can I use Bull Power or Bear Power on its own?
You can, but it was not designed that way. Elder read both histograms with the slope of the same exponential average: the slope picks the side, the histogram picks the moment. Without that filter, a zero-line cross mostly says the bar touched the average, which happens constantly in a range. Pairing the histogram with a rising or falling moving average condition is the usual remedy.
What length should I use for Elder Ray?
There is no universal setting. A shorter average gives quicker, smaller readings and a more nervous slope; a longer one gives larger, steadier readings that turn later. Elder's default came from daily charts, so on intraday futures the useful question is which length separates trend from range on your contract, and a backtest can answer that.
Is Elder Ray divergence reliable?
A gap between price and Bear Power or Bull Power shows that one side is reaching less far beyond value than before; it does not time a reversal by itself. Elder took it only in the direction of the average's slope. Treat it as context and test it on your own contract, including how often it appears with no turn following.