I figured I'd go a little more vertical, a la Michael Stoke's post at MarketSci, while working on the horizontal structure and content of this site. I wanted to follow up with MarketSci's series of posts on the Golden Cross and run some backtests on a broader range of markets.

System Rules:

Rules for both systems are the same except for the length of the SMA's being used

1. Enter position at the open if the previous day's faster SMA(10,50) crosses the slower SMA(50,200)
• Long Entry = 10(50) SMA > 50(200) SMA
• Short Entry = 10(50) SMA < 50(200) SMA
2. Exit position at the open if the previous day's faster SMA(10,50) crosses the slower SMA(50,200) in the opposite direction
• Long Exit = 10(50) SMA < 50(200) SMA
• Short Exit = 10(50) SMA > 50(200) SMA

• Risk 1% of Equity on each Trade
• Position Size = ((Account Equity * 1%)/2 ATR)
• Transaction Slippage = 2 ticks
• Markets Traded: ES, NQ, TF, GC, SI, HG, CL, HO, NG, RB, ZC, CC, CT, KC, ZW, LE, HE, ZS, SB, AUD, GBP, JPY, EUR, CHF

10/50 SMA

Full Report

50/200 SMA

Full Report

The 50/200 SMA Crossover Strategy resulted in much more robust results, with the 10/50 having an extended drawdown of 10+ years.

The trade distributions exhibit the classic fat tails of trend follow strategies. Without any stops in place there were a large number of trades that resulted in a loss greater than the 1% used to calculate positions size. Possible follow-up with stops added.

System Rules:

Rules for all systems are the same except for the type (SMA,EMA) and length(10/50,50/200) of the MA's being used

1. Enter position at the open if the previous day's faster MA(10,50) crosses the slower MA(50,200)
• Long Entry = 10(50) MA > 50(200) MA
• Short Entry = 10(50) MA < 50(200) MA
2. Exit position at the open if the previous day's faster MA(10,50) crosses the slower MA(50,200) in the opposite direction
• Long Exit = 10(50) MA < 50(200) MA
• Short Exit = 10(50) MA > 50(200) MA

• Risk 1% of Equity on each Trade
• Position Size = ((Account Equity * 1%)/2 ATR)
• Transaction Slippage = 2 ticks
• Markets Traded: ES, NQ, TF, GC, SI, HG, CL, HO, NG, RB, ZC, CC, CT, KC, ZW, LE, HE, ZS, SB, AUD, GBP, JPY, EUR, CHF

10/50 SMA

Full Report

10/50 EMA

Full Report

50/200 SMA

Full Report

50/200 EMA

Full Report

Without digging deeper into the trade data and looking at the periods of divergence between the SMA and EMA strategies, I find it interesting how quickly the outperformance of the 10/50 SMA vs the 10/50 EMA was reversed, then how much the two tracked each other.

The 50/200 cross strategies are long term strategies which makes less trade data available (even going back 25yrs). I'm cautious not to put too much weight on the outperformance of the 50/200 SMA over the 50/200 EMA strategy, in light of what was seen in the 10/50 cross strategies.

System Rules:

Simple Donchian 40/20 Strategy

Rules for both systems are the same except for the type (SMA,EMA) of the MA's being used

1. Enter position on break of 40 day Donchian Channel
Long Entry at 40 day High
Filter: only enter Long position if 50 MA > 200 MA
Short Entry at 40 day Low
Filter: only enter Short position if 50 MA < 200 MA
2. Exit position on break of 20 day Donchian Channel in opposite direction
• Long Exit at 20 day Low
• Short Exit at 20 day High

• Risk 1% of Equity on each Trade
• Position Size = ((Account Equity * 1%)/2 ATR)
• Transaction Slippage = 2 ticks
• Markets Traded: ES, NQ, TF, GC, SI, HG, CL, HO, NG, RB, ZC, CC, CT, KC, ZW, LE, HE, ZS, SB, AUD, GBP, JPY, EUR, CHF

Donchian 40/20 Without Filter

Full Report

Donchian 40/20 SMA Filter

Full Report

Donchian 40/20 EMA Filter

Full Report