Swing Index

Technical Analysis, Studies, Indicators:

Welles Swing Index (SI)

Description

The Welles Swing Index was developed by Welles Wilder to evaluate price change direction and price strength in relation to price swings. Swing Index uses last two bars data to define possible trend change over longer period of time. In particular, Swing Index helps to spot moments when investors change their trading behavior - moments when the changes of price (previously aligned to the swings of price) starts to move in opposite to swings direction, when previously strong and consistent group of investors (Bulls or Bears) becomes weaker and short-lived only. Since Swing index uses only 2 bars' data to define possible longer trend development, it is mostly attracts short-term intraday traders who are mostly interested in short-term price swings.

Technical Analysis, Signals and Trading Systems

To understand better swing index, we have to take deeper look into its calculation. The SI indicator calculations are quite complex:

SI = 50 * [pClose - Close + 0.5 * (pClose - pOpen)+ 0.25 * (Close - Open)]/ R * (K / M)

For better understanding lets split SI calculations into several parts where the first part (let's call it Weighted Change)
Weighted Change = 50 * [Close - pClose + 0.5 * (pClose - pOpen)+ 0.25 * (Close - Open)]

We called the first part as Weighted Price Change because it could be rewritten as
100 * [pClose - Close + 0.5 * (pClose - pOpen)+ 0.25 * (Close - Open)]/2
where (Close - pClose) is change between current and previous bar close prices, (pClose - pOpen) is previous bar open-close change and(Close - Open) is current bar open-close change. This is average price change with heavier weight put on a change between current and previous bar close prices. The weighed change may also call it as Typical Price Change because these calculations are quite similar to the Typical Price calculations (TP = (Close + High + Low)/3). We may call it as True Change or Average Change because it also has some similarities with True Price and Average Price calculation. However, it does not matter how we call it, we have to understand that this is price change.

By following further in the Swing Index Formula, we have

Weighted Price Change / R

where R could be one of the following depending on a price swing (strongest move)

1. R = ( | High - Previous Close| ) - 0.5 * ( | Low - Previous Close| ) +.25 * ( | Previous Close - Previous Open| )
2. R = ( | Low - Previous Close| ) - 0.5 * ( | High - Previous Close| ) +.25 * ( | Previous Close - Previous Open| )
3. R = ( | High - Low| ) +.25 * ( | Previous Close - Previous Open| )

Basically R is 2-bar price swing. By dividing the weighed price change by the price swing we receive weighted change in absolute values which evaluates the strength and location of a price change (price swing) to a price swing:

• When we have strong swing up and strong price change up, Swing index will be approaching high positive numbers.
• When we have strong swing up and small positive or negative price change  (price swigged up but then dropped down), the swing Index will be around zero.
• When we have strong swing down and strong price change down, Swing index will be approaching low negative numbers.
• When we have strong swing down and small negative or positive  price change  (price swigged down but then recovered), the swing Index will be around zero.

Weighted Absolute Change = 50 * [Close - Previous Close + 0.5 * (Previous Close- pOpen)+ 0.25 * (Close - Open)] / R

The last part in the Swing Index calculation (K/M) is actual strength of the swing: K = the greatest absolute value of  | High - Previous Close | and |Low - previous Close |. M is a maximum price change within a bar and it used to keep K/M in the 0-1 range. Otherwise, Swing index will go go above 100% or drops below -100%. Basically in this last part we compare the current swing (K) to the maximum possible swing (M) in the analyzed timeframe.

Overall , we may say, that Swing Index is used to compare a price movement (price change) to price swing with purpose of defining the strength of this price movement - to see whether price swing and price change are aligned (in the same direction) and to see whether price change is as strong as price swing. Positive SI reading world be a sign of a price up-move and negative SI readings would be a sign of price down-move. As was mentioned above, Swing Index is based on the 2-bar data and it used in intraday short-term trading - just several bars into the future. According to Technical Analysis theory,

• a buy signal could be considered when the Swing Index crosses over zero - a trader might expect short-term price movement upward - price does not move up yet, but we start to have up-swings.
• a sell signal could be considered when the Swing Index crosses below zero - a trader might expect short-term price movement downward - price does not move down yet, but we start to have down-swings.

Chart 1: Intraday SPY stock chart and Swing index: 1 bar = 1 minute.



As you may see the Swing index is quite choppy and it could be challenge to use this technical indicator in real trading. On our chart we allow to smooth this indicator, yet, it adds some lag, still it could help to reduce the number of generated signals

Chart 1: Intraday SPY stock chart and smoothed Swing index: 1 bar = 1 minute.



Also, because of the Swing Index's choppiness, Accumulated Swing Index is more popular among traders.

Swing Index Formula and Calculations

We used following abbreviations in the formulas below:

Close - Current bar Close price
pClose - Previous bar Close price
High - Current bar High price
Low - Current bar Low price
Open - Current bar Open price
pOpen - Previous bar Open price

1. First determine the values calculated for a variables, "K"

 K = the greatest absolute value of  | High - pClose | and |Low-pClose |

2. Calculate "R" variable. For that you have to get following absolute values

A = | High - pClose |
B = | Low - pClose |
C = | High - Low |

Now find the biggest of A, B and C

If A is the greatest then
R = ( | High - pClose| ) - 0.5 * ( | Low - pClose | ) +.25 * ( | pClose - pOpen | )
If B is the greatest then
R = ( | Low - pClose| ) - 0.5 * ( | High - pClose | ) +.25 * ( | pClose - pOpen | )
If C is the greatest then
R = ( | High - Low| ) +.25 * ( | pClose - pOpen | )

3. Calculate Swing Index

SI = 50 * [Close - pClose + 0.5 * (pClose - pOpen)+ 0.25 * (Close - Open)]/ R * (K / M)

where M is a Limit Move (maximum price change within a bar) - on our charts we automatically define maximum price change as the highest difference between high and low bar' prices over specified period:

M = max(High - Low)

By V. K. for MarketVolume.com