Retail Power

An oscillator from −100 to +100 for Binance futures: how steadily, over the last day, the product of volume and the price deviation from its 1000-minute average has been rising.

What it is

Retail Power combines two things into one number: how far the price has moved from its average, and how much volume traded during that deviation. The model then looks at which way this product has been changing over the last day. A positive value means price and volume pushed it up, a negative value means they pushed it down.

The model is computed on the server from 1-minute candles of Binance USDT perpetual futures, for all trading coins. The value is recalculated every second and saved to history once a minute. The current, still open candle is included in the calculation.

note

Despite the name, the model does not separate retail traders' trades from the rest. It uses only the close price and the volume of 1-minute candles.

Retail Power is available in three forms:

• the Retail Power widget: a table of all coins with the current value;
• the [SFM] Retail power indicator: a histogram of the value below the chart;
• the [SFM] Retail power highlighter indicator: background highlighting on the price chart itself. The charts inside the widget show it too.

The widget is part of the Max plan because it shows the whole market at once. The chart indicator is available from Pro because it shows a single coin. The widget opens from the "Screeners → Microstructure → Retail Power" menu.

Formula

1. The deviation of the price from its 1000-minute average (about 16.7 hours), in percent:

$$PRATIO = \frac{Close}{SMA_{1000}(Close)} × 100 - 100$$

2. The deviation strength: the deviation multiplied by the candle volume in coins:

$$Pw = Volume × PRATIO - PRATIO$$

3. An RSI with a period of 1440 minutes (one day), computed on the $Pw$ series rather than on price, and rescaled to −100…+100:

$$RetailPower = (RSI_{1440}(Pw) - 50) × 2$$

The RSI is computed on exactly the last 1440 changes of $Pw$, with no smoothing over earlier history. So the value can be read more simply: it is the share of the net change of $Pw$ over the day in the sum of all its swings over the same period.

• +100: over the day $Pw$ only rose;
• 0: rises and falls of $Pw$ cancelled each other out;
• −100: $Pw$ only fell.

Widget

Column	What it shows
Symbol	Coin
Value	Retail Power value, from −100 to +100

By default the table is sorted by Value in descending order. When a coin changes in the top five rows, the row flashes for a second.

There is one filter, Value. A positive number keeps coins with a value not lower than the one entered, a negative number keeps coins with a value not higher than it. For example, Value = −10 shows coins at −10 and below. The global filters of the control panel and the blacklist also apply to the table.

Indicators

[SFM] Retail power

A histogram below the chart with a dashed zero line.

Element	When
Green bar	value above zero
Red bar	value zero or below
Green background	value above +5
Red background	value below −5

[SFM] Retail power highlighter

Highlights the background of the main chart. There is one color, red, from transparent to saturated. The brightness grows with the absolute value: highlighting starts above 5 and is at its maximum above 15. The highlighting ignores the sign: both +12 and −12 give the same color. Check the sign in the widget or in the histogram.

Settings

• Symbol (only in [SFM] Retail power): the coin to take the model from. By default, the chart coin.
• Filter noise?: hides values with an absolute value below 1.

How to read it

• A value above +5: over the last day the product of volume and the price deviation from its average has been rising steadily. $Pw$ rises when the price is above its 1000-minute average and volume increases, or when the price climbs toward the average from below.
• A value below −5: the same in the opposite direction: the price is below its average and moves further away on rising volume, or the price falls toward the average from above.
• Values near zero: there is no steady direction over the day.

Combining price and volume in $Pw$ makes the sign not always obvious: when the price is below its average ($PRATIO$ is negative), rising volume pulls $Pw$ down. So read the value together with the price chart, not on its own.

Limitations

• Binance USDT perpetual futures only.
• Volume is taken in coins, not in dollars.
• The model is slow: the 1000- and 1440-minute windows react to market changes with a lag of hours.
• The calculation needs 2,441 one-minute candles of history, so a new coin has no value for its first 1.7 days.
• There is no public backtest of the model yet. We do not claim that the value predicts price moves.