SPY Intraday X/Y/N Micro-Reversal Strategy Backtest Report

(Sample Period: 2024-02-27 to 2026-02-27)

I. Research Objective

This backtest validates an intraday short-term strategy: enter the market when the price deviates from the previous close by a certain magnitude `X`, wait for a slight reverse regression of `Y` to take profit; if not, force close the position at 15:55, no overnight holding.  

Your focus is on whether a stable profitable range exists for the "naked short then buy" (short position) strategy.

 II. Data & Backtest Settings

- Underlying: `SPY.US`

- Data: 1-minute K-line, two-year sample

- Valid complete sessions after cleaning: `487` (390 minute bars per session)

- Parameter grid: `X=0.5%~3.0%(step 0.1)`, `Y=0.05%~0.50%(step 0.05)`, total 260 combinations

- Nominal capital per trade: `10,000 USD`

- Commission: `1.1 USD/trade (flat)`

- Entry: Next minute's opening price after the signal

- Exit: Reach target `Y` or forced close at 15:55

Net profit per trade is calculated as:  

`shares = floor(10000 / entry_price)`  

- Short: `net = (entry - exit) * shares - 1.1`

- Long: `net = (exit - entry) * shares - 1.1`

 III. Key Results

 1) Short "Winning Range" Established

According to your screening criteria: `Win Rate>=80%`, `Number of Trades>=30`, `Net Profit Per Trade>0`, there are `28` valid parameter sets for short positions.  

Parameters are mainly concentrated in:

- `X`: 1.1%~2.3% (most dense at 1.7%~1.9%)

- `Y`: 0.05%~0.20% (most dense at 0.10%)

Representative combinations are as follows:

| X(%) | Y(%) | # Trades | Win Rate | Net Profit Per Trade(USD) | Total Net Profit(USD) |

|---:|---:|---:|---:|---:|---:|

| 1.8 | 0.20 | 72 | 80.56% | 4.26 | 306.64 |

| 1.9 | 0.20 | 62 | 80.65% | 3.86 | 239.27 |

| 1.7 | 0.10 | 144 | 90.28% | 2.46 | 354.78 |

| 1.8 | 0.10 | 129 | 89.92% | 2.59 | 333.60 |

 2) Characteristics of Combined Long/Short Mode

When combining long and short, the combination with the highest net profit per trade often corresponds to a lower win rate (approx. 50%~69%) and larger drawdown.  

This indicates that "high win rate" and "high profit per trade" are difficult to achieve simultaneously under the same parameters, requiring a trade-off.

 IV. Your New Question: Drawdown When Target Profit is Increased to 0.3%/0.5%/1.1%

After rerunning with fixed `Y` values, the following results are obtained (taking the combination with the highest net profit per trade for each Y):

| Y(%) | Mode | Optimal X(%) | # Trades | Win Rate | Net Profit Per Trade(USD) | Max Drawdown(USD) |

|---:| --- |---:|---:|---:|---:|---:|

| 0.3 | Short | 2.1 | 35 | 74.29% | 5.75 | 402.38 |

| 0.5 | Short | 1.6 | 51 | 56.86% | 9.93 | 402.38 |

| 1.1 | Short | 1.6 | 39 | 23.08% | 11.48 | 641.38 |

The conclusion is clear:  

The larger `Y` is, the higher the profit per trade, but the win rate drops significantly, and drawdown increases (especially at 1.1%).