Trend Following Strategy Based on Volume Flow Indicator

Author: ChaoZhang, Date: 2023-10-10 14:51:13
Tags:

Overview

This strategy implements trend following trading based on Volume Flow Indicator (VFI). It judges the market trend direction by calculating price fluctuations and volume changes, and realizes low buying and high selling.

Strategy Principle

  1. Calculate VFI indicator: Compute VFI value based on logarithmic price change and volume, and smooth out oscillations through smoothing techniques.

  2. Determine trend direction: VFI crossing above 0 is a bullish signal, while crossing below 0 is a bearish signal.

  3. Trading signals: Go long when fast EMA crosses above slow EMA and VFI crosses above buy line; close position when VFI crosses below sell line.

  4. Stop loss: Set fixed stop loss percentage.

This strategy mainly relies on VFI to determine the trend direction, combined with moving averages to generate trading signals. VFI reflects market sentiment through price volatility and volume changes, making it a trend following indicator. Compared with single price indicators, VFI provides more comprehensive judgments and identifies trend reversal points better, filtering out consolidations.

Advantages

  1. VFI determines trends better than single price indicators, effectively filtering out consolidations and false breakouts.

  2. Moving averages provide supplementary judgements, avoiding incorrect signals from VFI in ranging markets.

  3. Fixed stop loss controls risk and facilitates risk management.

  4. Trend following mode generates excess returns without predicting reversals.

  5. Flexible parameter tuning adapts to different periods and products.

Risks

  1. VFI may generate incorrect signals during significant fluctuations.

  2. Fixed stop loss could be too wide or too narrow.

  3. Ill-configured entry and exit settings lead to over-trading or missing trades.

  4. Trend following fails to capture reversals and needs timely stop loss.

  5. Improper parameters cause premature or delayed entry.

Enhancements

  1. Optimize VFI calculation by adjusting parameters.

  2. Fine-tune moving average periods for better signal timing.

  3. Employ dynamic stop loss instead of fixed one.

  4. Add other indicators to filter signals.

  5. Optimize parameters separately based on timeframe.

  6. Test robustness across different products.

Conclusion

This strategy determines the trend direction with VFI and uses moving averages to filter out wrong signals. It realizes low buying/high selling through trend following without predicting reversals. The advantage lies in its superior trend detection over single price indicators and ability to filter out consolidations. The main risk is generating incorrect signals during fluctuations. Optimizing parameters, adding supplementary indicators and stop loss techniques can improve its stability. Overall, with parameter tuning and stop loss optimizations, this VFI based strategy can become a reliable trend following system.


/*backtest
start: 2023-10-02 00:00:00
end: 2023-10-06 21:00:00
period: 3m
basePeriod: 1m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

// This source code is subject to the terms of the Mozilla Public License 2.0 at https://mozilla.org/MPL/2.0/
// © mohanee
//This strategy is based on VFI indicator published by  UTS.  
//more details of VFI indicator can be found at  [url=http://mkatsanos.com/VFI.html]http://mkatsanos.com/VFI.html[/url] 
// I have added buy line and sell line to the indicator  and tested  SPY stock/index on one hour chart

//@version=4
strategy(title="VFI  strategy [based on VFI indicator published by  UTS]", overlay=false,pyramiding=2, default_qty_type=strategy.fixed,    initial_capital=10000, currency=currency.USD)


// Const
kMaColor = color.aqua
kNeutralColor = color.gray
kBearColor = color.red
kBullColor = color.green

kAlma = "ALMA"
kEma = "EMA"
kSma = "SMA"
kWma = "WMA"


// Input

vfi_length = input(8, title="Length", minval=1)  //default 130
vfi_coef = input(0.2, title="Coef", minval=0.1)
vfi_volCutoff = input(2.5, title="Volume Cutoff", minval=0.1)
vfi_smoothLen = input(3, title="Smoothing Period", minval=1)
vfi_smoothType = input(kEma, title="Smoothing Type", options=[kAlma, kEma, kSma, kWma])

//These are adde by me for the strategy purpose  BEGIN
vfi_buyLine = input(-4, title="Buy Line", minval=-10)
vfi_sellLine = input(5, title="Sell Line", minval=-10)
stopLoss = input(title="Stop Loss%", defval=5, minval=1)
//These are adde by me for the strategy purpose  END

vfi_longEMA = input(200, title="Long EMA", minval=1)
vfi_shortEMA1 = input(50, title="short EMA1", minval=1)
vfi_shortEMA2 = input(9, title="short EM2A", minval=1)

vfi_showTrend = input(false, title="Visualize Trend")
vfi_showFill = input(true, title="Apply Filling")
vfi_showMa = input(true, title="Show Moving Average")
vfi_maType = input(kSma, title="Moving Average Type", options=[kAlma, kEma, kSma, kWma])
vfi_maLength = input(30, title="Moving Average Length", minval=1)
vfi_almaOffset = input(0.85, title="• ALMA - Offset (global setting)", minval=0.0, maxval=1.0, step=0.05) // more smoothness (closer to 1) vs. more responsiveness (closer to 0)
vfi_almaSigma = input(6.0, title="• ALMA - Sigma (global setting)", minval=0.0, step=0.05) // the larger sigma the smoother ALMA


// Functionality

isRising(sig) =>
    sig > sig[1]
    
isFlat(sig) =>
    sig == sig[1]

vfi_trendColor(sig) =>
    isFlat(sig) ? kNeutralColor : isRising(sig) ? kBullColor : kBearColor
    
vfi_color(sig) =>
    isFlat(sig) ? kNeutralColor : sig > 0 ? kBullColor : kBearColor
    
osc_color(sig) =>
    sig == 0 ? kNeutralColor : sig > 0 ? kBullColor : kBearColor

smooth(t, sig, len) =>
    ma = float(sig)         // None
    if t == kSma            // Simple
        ma := sma(sig, len)
    if t == kEma            // Exponential
        ma := ema(sig, len)
    if t == kWma            // Weighted
        ma := wma(sig, len)
    if t == kAlma           // Arnaud Legoux
        ma := alma(sig, len, vfi_almaOffset, vfi_almaSigma)
    ma

calc_vfi(fviPeriod, smoothType, smoothLen, coef, vCoef) =>
    avg = nz(hlc3)
    inter = log(avg) - log(avg[1])
    vInter = stdev(inter, 30)
    cutOff = coef * vInter * close
    vAve = smooth(kSma, volume[1], fviPeriod)
    vMax = vAve * vCoef
    vC = min(volume, vMax)
    mf = avg - avg[1]
    vCp = iff(mf > cutOff, vC, iff(mf < -cutOff, -vC, 0))
    sVfi = sum(vCp, fviPeriod) / vAve
    vfi = smooth(smoothType, sVfi, smoothLen)
    
value_vfi = calc_vfi(vfi_length, vfi_smoothType, vfi_smoothLen, vfi_coef, vfi_volCutoff)
value_ma = smooth(vfi_maType, value_vfi, vfi_maLength)

longEMAval= ema(close, vfi_longEMA)
shortEMAval1= ema(close, vfi_shortEMA1)
shortEMAval2= ema(close, vfi_shortEMA2)

color_vfi = vfi_showTrend ? vfi_trendColor(value_vfi) : vfi_color(value_vfi)
color_osc = vfi_showFill ? osc_color(value_vfi) : na
color_ma = vfi_showMa ? kMaColor : na


// Drawings

plot_vfi = plot(value_vfi, title="VFI", color=color_vfi, linewidth=1)
plot_fill = plot(0, color=color_vfi, editable=false)
fill(plot_vfi, plot_fill, title="Oscillator Fill", color=color_osc, transp=75) 
hline(vfi_buyLine, color=color.green, title="Buy Line", linewidth=2, linestyle=hline.style_dashed)
hline(vfi_sellLine, color=color.purple, title="Sell Line", linewidth=2, linestyle=hline.style_dashed)
plot(value_ma, title="MA", color=color_ma, linewidth=2)

strategy.entry(id="VFI LE", long=true,  when=crossover(value_vfi,vfi_buyLine)  and ( shortEMAval1 >= longEMAval  ))

//strategy.close(id="VFI LE", comment="Exit",   when=crossunder(value_vfi,vfi_sellLine))
strategy.close(id="VFI LE", comment="TP Exit",   when=crossunder(value_vfi,vfi_sellLine) and close>strategy.position_avg_price)
//strategy.close(id="VFI LE", comment="Exit",   when=  (shortEMAval1 > shortEMAval2 )  and crossunder(close, shortEMAval2))

//stoploss
stopLossVal =   strategy.position_avg_price -  (strategy.position_avg_price*stopLoss*0.01) 
strategy.close(id="VFI LE", comment="SL Exit",   when=crossunder(value_vfi,vfi_sellLine) and close < stopLossVal)




More