Suppose there’s a person named Xiaofei. At the close on August 22, $PHA closed at 0.0245. He set himself a rule: only look at daily candles that have closed. Today’s low must be lower than yesterday’s low; equal doesn’t count. At the same time, today’s close must not be below yesterday’s close; a tie counts. If those conditions are met, he buys at the close and sells at the next day’s close. If the conditions still hold on the day he sells, he sells and buys back in. He doesn’t add to his position while holding. His colleague Xiaocang converted all 1000 USDT into the coin at 0.0245 on the same day, planning to hold it until the end. The replay below excludes fees.
Over these 90 days, there were 9 days that met the rule. Xiaofei bought 9 times and sold 9 times, with no buybacks, and ended with cash.
He got his first signal on the second day of the window. On August 23, the low was 0.0232—the lowest price of the entire 90 days—and the close of 0.0251 was back above the previous day’s close of 0.0245. He bought in and made 3.59% by the next day’s close, bringing his balance to 1035.86.
He stayed out for 4 candles. On August 29, he bought at 0.0251 and lost 3.98% by the next day’s close. He sold at 994.59. The day he sold was also the day with the lowest close of the 90-day period, at 0.0241.
He stayed out for another 16 candles. On September 16, he bought at 0.0273 and made +13.92% by the next day’s close, bringing his balance to 1133.03.
After staying out for 6 more candles, he bought on September 24. The low of 0.0449 dipped below the previous day’s low of 0.0451, and the close of 0.051 was back above the previous day’s close of 0.0475. The next day, September 25, the price touched the 90-day high of 0.1007 intraday and closed at 0.0867. He sold for a 70.00% gain, bringing his balance to 1926.15. At 0.0867, the close was 253.88% above the starting price and was the highest close of the 90 days. A new closing high appeared 12 times in total, with September 25 being the last. Xiaocang’s balance was 3538.78 on the same day; he had held through the preceding rise. From when Xiaofei sold on September 17 at 0.0311 until he bought back in on September 24 at 0.051, he was out of the market for those candles.
He then stayed out for 15 candles. On October 11, he bought at 0.061 and lost 8.20% by the next day’s close, bringing his balance to 1768.27. This was the biggest loss among his 9 completed trades, and the day of the strategy’s deepest drawdown. After a gap of 2 candles, he bought at 0.0508 on October 15 and made +4.92% by the next day’s close, bringing his balance to 1855.29. After staying out for 4 more candles, he bought at 0.0612 on October 21 and made just 0.33% by the next day’s close, bringing his balance to 1861.35. After a gap of 2 candles, the close on October 25 was 0.0484, exactly the same as the previous day’s close, which counts under his rule. He bought in and made +3.72% by the next day’s close, bringing his balance to 1930.57—the strategy’s highest balance.
After staying out for 8 candles, he made his final trade on November 4, buying at 0.0389. He lost 5.91% by the next day’s close, bringing his balance to 1816.43. He then stayed out for another 14 candles, all the way to the end of the period. On November 19, the close was 0.0301.
Results for these 90 days (drawdown calculated using daily closes):
Xiaocang, continuously fully invested: 1228.57 USDT, +22.86%, with a maximum drawdown of 67.01% (from the September 25 closing high of 0.0867 and balance of 3538.78 to the November 14 close of 0.0286 and balance of 1167.35). On November 19, the close was 0.0301; it would need to rise another 188.04% to return to that closing high. Over the same period, $BTC fell -19.58% from its August 22 close to its November 19 close. Both started with 1000 USDT.
Xiaofei, holding overnight only on days when “the low breaks below yesterday’s low, but the close gets back above yesterday’s close”: 1816.43 USDT, +81.64%, with a maximum drawdown of 8.20% (from his September 25 balance of 1926.15 to 1768.27 after selling on October 12). Of his 9 completed trades, 6 were profitable and 3 were losing. The +70.00% trade contributed the most; excluding it, the other 8 trades, still starting with 1000 USDT, would have ended at 1068.49 USDT, +6.85%.
This is a replay of a rule using Binance spot PHAUSDT and BTCUSDT historical daily data (UTC daily candles, closed data from 2026-08-22 to 11-19); it is not a prediction of what will happen in the future. These 9 overnight trades added up to +81.64% over the 90 days, mainly thanks to that one night on September 25. On the same day, Xiaocang’s balance reached 3538.78, before later falling back to 1167.35. Would you use “today’s low breaks below yesterday’s low, but the close gets back above yesterday’s close” to decide whether to hold overnight?
Over these 90 days, there were 9 days that met the rule. Xiaofei bought 9 times and sold 9 times, with no buybacks, and ended with cash.
He got his first signal on the second day of the window. On August 23, the low was 0.0232—the lowest price of the entire 90 days—and the close of 0.0251 was back above the previous day’s close of 0.0245. He bought in and made 3.59% by the next day’s close, bringing his balance to 1035.86.
He stayed out for 4 candles. On August 29, he bought at 0.0251 and lost 3.98% by the next day’s close. He sold at 994.59. The day he sold was also the day with the lowest close of the 90-day period, at 0.0241.
He stayed out for another 16 candles. On September 16, he bought at 0.0273 and made +13.92% by the next day’s close, bringing his balance to 1133.03.
After staying out for 6 more candles, he bought on September 24. The low of 0.0449 dipped below the previous day’s low of 0.0451, and the close of 0.051 was back above the previous day’s close of 0.0475. The next day, September 25, the price touched the 90-day high of 0.1007 intraday and closed at 0.0867. He sold for a 70.00% gain, bringing his balance to 1926.15. At 0.0867, the close was 253.88% above the starting price and was the highest close of the 90 days. A new closing high appeared 12 times in total, with September 25 being the last. Xiaocang’s balance was 3538.78 on the same day; he had held through the preceding rise. From when Xiaofei sold on September 17 at 0.0311 until he bought back in on September 24 at 0.051, he was out of the market for those candles.
He then stayed out for 15 candles. On October 11, he bought at 0.061 and lost 8.20% by the next day’s close, bringing his balance to 1768.27. This was the biggest loss among his 9 completed trades, and the day of the strategy’s deepest drawdown. After a gap of 2 candles, he bought at 0.0508 on October 15 and made +4.92% by the next day’s close, bringing his balance to 1855.29. After staying out for 4 more candles, he bought at 0.0612 on October 21 and made just 0.33% by the next day’s close, bringing his balance to 1861.35. After a gap of 2 candles, the close on October 25 was 0.0484, exactly the same as the previous day’s close, which counts under his rule. He bought in and made +3.72% by the next day’s close, bringing his balance to 1930.57—the strategy’s highest balance.
After staying out for 8 candles, he made his final trade on November 4, buying at 0.0389. He lost 5.91% by the next day’s close, bringing his balance to 1816.43. He then stayed out for another 14 candles, all the way to the end of the period. On November 19, the close was 0.0301.
Results for these 90 days (drawdown calculated using daily closes):
Xiaocang, continuously fully invested: 1228.57 USDT, +22.86%, with a maximum drawdown of 67.01% (from the September 25 closing high of 0.0867 and balance of 3538.78 to the November 14 close of 0.0286 and balance of 1167.35). On November 19, the close was 0.0301; it would need to rise another 188.04% to return to that closing high. Over the same period, $BTC fell -19.58% from its August 22 close to its November 19 close. Both started with 1000 USDT.
Xiaofei, holding overnight only on days when “the low breaks below yesterday’s low, but the close gets back above yesterday’s close”: 1816.43 USDT, +81.64%, with a maximum drawdown of 8.20% (from his September 25 balance of 1926.15 to 1768.27 after selling on October 12). Of his 9 completed trades, 6 were profitable and 3 were losing. The +70.00% trade contributed the most; excluding it, the other 8 trades, still starting with 1000 USDT, would have ended at 1068.49 USDT, +6.85%.
This is a replay of a rule using Binance spot PHAUSDT and BTCUSDT historical daily data (UTC daily candles, closed data from 2026-08-22 to 11-19); it is not a prediction of what will happen in the future. These 9 overnight trades added up to +81.64% over the 90 days, mainly thanks to that one night on September 25. On the same day, Xiaocang’s balance reached 3538.78, before later falling back to 1167.35. Would you use “today’s low breaks below yesterday’s low, but the close gets back above yesterday’s close” to decide whether to hold overnight?