6 of 11 Backtests of Our Listed EAs Never Took a Short. What the Long/Short Line of the Report Shows

The MT5 report splits trades into long and short and gives a win rate for each side. It is a line most people skip. We pulled it from the backtest reports of the EAs we currently list on the Market: 11 reports covering 10 products (one EA appears on two symbols).

Two-sided tests first, sorted from "short side wins more often" to "long side wins more often", then the long-only tests. Profit factor and maximum equity drawdown are for the whole test.

EA | Long trades (win rate) | Short trades (win rate) | Profit factor | Max equity drawdown

Atlas Diversified Trend Portfolio | 1,118 (32.83%) | 75 (38.67%) | 1.48 | 19.74%

Trinity Multi Asset Trend | 224 (42.41%) | 165 (42.42%) | 1.79 | 14.44%

Audcad Peace Mean Reversion | 2,630 (74.37%) | 2,774 (73.86%) | 1.53 | 36.18%

Gold King Breakout Grid | 2,236 (84.21%) | 598 (78.76%) | 4.67 | 42.75%

Bitcoin Comet Trend Follower | 62 (50.00%) | 66 (42.42%) | 2.19 | 18.17%

Donchian Trend Engine (free) | 631 (34.55%) | 0 | 1.50 | 10.34%

Megamax Donchian Trend (EURJPY) | 571 (25.22%) | 0 | 1.51 | 17.66%

Megamax Donchian Trend (USDJPY) | 629 (34.98%) | 0 | 1.55 | 16.27%

Monthly Burst USDJPY | 639 (37.56%) | 0 | 1.50 | 11.73%

Nasdaq Dip Buyer | 54 (61.11%) | 0 | 1.69 | 32.63%

Overdrive Usdjpy Aggressive | 877 (41.16%) | 0 | 1.33 | 29.37%

Six of the eleven tests contain zero short trades: five tests on USDJPY or EURJPY, and one on the Nasdaq index. A long-only system is not wrong. But if the market rose over the test period, the backtest cannot tell you what the EA does in a market that falls for several years, so open the price chart for the tested period and look before you trust the result.

- Trinity Multi Asset Trend: 224 longs at 42.41%, 165 shorts at 42.42%. As symmetric as a result gets.

- Audcad Peace Mean Reversion: 2,630 longs at 74.37%, 2,774 shorts at 73.86%. Thousands of trades per side and a gap of half a point.

- Atlas Diversified Trend Portfolio: 1,118 longs at 32.83% against 75 shorts at 38.67%. The short side wins more often, but it is 75 trades out of 1,193. This is close to a long-only test with a few shorts in it.

- Bitcoin Comet Trend Follower: 62 longs at 50.00%, 66 shorts at 42.42%. A 7.6-point gap on about 60 trades a side.

- Gold King Breakout Grid: 2,236 longs at 84.21% against 598 shorts at 78.76%. The largest sample with a clear tilt to the long side. It also carries a 42.75% maximum equity drawdown in that test, so read its win rates next to that number.

The short side has the higher win rate in 2 of the 5 and the long side in 3. Five tests is too few to say anything general about direction.

Win rate per side is not profit per side. The summary we keep does not break profit down by direction, so we are not claiming that either side earned more. A side can win more often and still earn less if its winners are smaller.

Ethereum Trend Follower is left out because we do not have the long/short split for its current version (v1.55), and we did not want to quote an older report. Two more listed EAs are absent: one is being re-measured and one has no comparable single report.

- Look at the long/short line of any report before the headline. If one side is zero, check what the market did over the test period, and ask to see a period in which it fell.

- If both sides trade, look at the counts before the percentages. A gap of several points on 60 trades is noise until proven otherwise.

- These are historical simulations of EAs we chose to keep on sale, not a live record, and they differ in symbol, period, tick model and risk preset.

The two EAs in this article: Trinity Multi Asset Trend https://www.mql5.com/en/market/product/182640 · Bitcoin Comet Trend Follower https://www.mql5.com/en/market/product/181831