Trailing stop strategy backtest results

The short answer

Answers here: Does a trailing stop improve a trend strategy? · What does the ATR trail do to the win rate? · Long only, or both ways?

What the script does

EMA Cross with Trailing Stop, from the strategy library.

The numbers, run by run

Net and holding the symbol are after costs over the same period. Held-out is the verdict on the last third of the history judged alone: survived, decayed, did not survive, or too few trades to judge.

Year by year

9/21 cross, 3 ATR trail on SPY, long and short, the same rule long only, and SPY itself, each calendar year after costs. A partial first or last year counts the sessions it has.

Every trade

The 94 trades of 9/21 cross, 3 ATR trail on SPY, long and short, oldest first: side, entry and exit days, fill prices after slippage, and the return after both commissions. The other runs' trades open below.

The script

Pine v6, exactly as it ran for 9/21 cross, 3 ATR trail on SPY, long and short. The variants change only the inputs listed in the table.

//@version=6
strategy("EMA Cross with Trailing Stop", overlay=true, default_qty_type=strategy.percent_of_equity, default_qty_value=100, initial_capital=100000, commission_type=strategy.commission.percent, commission_value=0.05, slippage=1, margin_long=0, margin_short=0)

src      = input.source(close, "Source", group="Lengths")
fastLen  = input.int(9, "Fast EMA", minval=1, group="Lengths")
slowLen  = input.int(21, "Slow EMA", minval=2, group="Lengths")
useLong  = input.bool(true, "Trade longs", group="Direction")
useShort = input.bool(true, "Trade shorts", group="Direction")
atrLen   = input.int(14, "ATR length", minval=1, group="Risk")
trailMul = input.float(3, "Trail distance (ATR multiples)", minval=0.5, step=0.5, group="Risk")
actMul   = input.float(1, "Activation profit (ATR multiples)", minval=0, step=0.5, group="Risk")

fast = ta.ema(src, fastLen)
slow = ta.ema(src, slowLen)
atr  = ta.atr(atrLen)

longSig  = ta.crossover(fast, slow) and barstate.isconfirmed
shortSig = ta.crossunder(fast, slow) and barstate.isconfirmed

// strategy.exit works in ticks: convert the ATR distances
trailTicks = trailMul * atr / syminfo.mintick
actTicks   = actMul * atr / syminfo.mintick

pFast = plot(fast, "Fast EMA", color=color.new(#16A56A, 0), linewidth=2)
pSlow = plot(slow, "Slow EMA", color=color.new(#E5484D, 0), linewidth=2)
fill(pFast, pSlow, color=fast > slow ? color.new(#16A56A, 90) : color.new(#E5484D, 90), title="Trend fill")

if longSig
    strategy.close("Short")
    if useLong
        strategy.entry("Long", strategy.long)
        strategy.exit("Long trail", "Long", trail_points=actTicks, trail_offset=trailTicks)
if shortSig
    strategy.close("Long")
    if useShort
        strategy.entry("Short", strategy.short)
        strategy.exit("Short trail", "Short", trail_points=actTicks, trail_offset=trailTicks)

How it was tested

Alpaca daily data since 2016, 0.05% commission and 0.025% slippage on every fill, the last third of the history held out, textbook settings and nothing tuned. The engine, fills and what it does not model are written out once, in data and methodology.

Questions