Suppose there’s someone called Xiaoshu who works part-time watching over the machines at a community activity center in the county town at night. On August 17, $STEEM opened at 0.03503. Old Shen, who often came by to print documents, said he’d bought and was just holding, and that even if it fell on Thursday, there was no need to rush in and buy the dip. Xiaoshu set himself another rule: whenever the UTC Thursday daily candle closed below Wednesday’s close, he would buy 1000 USDT worth at Thursday’s closing price, then count forward 4 more closed daily candles—which would be the following Monday’s close—and sell at that day’s closing price. If the previous trade hadn’t been sold yet, he would skip any new signal.
I ran this strategy on Binance spot STEEM/USDT closed daily candles from August 17 to November 14 (UTC, 90 candles in total). The November 15 candle hadn’t closed yet, so it wasn’t included. There were 13 UTC Thursdays during this period, each with a Wednesday close to compare against. Thursday’s close was higher than Wednesday’s on 4 days and lower on 9 days; there were no unchanged closes.
None of the 9 signals occurred while a previous trade was still open. On one occasion, the scheduled sell day’s candle hadn’t closed yet, so it wasn’t included. In practice, 8 trades were completed: 3 winners and 5 losers. No two trades were handed off on the same day, and there was no case of selling one day and buying again the next. After cycling 1000 USDT through 8 trades, the final amount was 1193.32, a compounded gain of +19.33%, averaging +2.64% per trade. The first 4 trades had 2 winners and 2 losers, with a compounded gain of +27.54%, reaching a balance of 1275.38; the last 4 had 1 winner and 3 losers, with a compounded loss of -6.43%. The best trade bought at the September 10 close of 0.04406 and sold at 0.05566 on September 14, for +26.33%. The worst trade bought at the August 27 close of 0.04507 and sold at 0.04308 on August 31, for -4.42%. Excluding the best trade, the remaining 7 trades had 2 winners and 5 losers, with a compounded return of -5.54%.
If Old Shen had bought at the August 17 open and held until the November 14 close at 0.04169, his 1000 USDT would have become 1190.12, a gain of +19.01%. Over the same period, $BTC was up +1.99% from the August 17 open to the November 14 close. Xiaoshu finished his 8 trades with 1193.32, almost the same as Old Shen’s 1190.12.
The turning point was a trade in mid-September. Xiaoshu bought at the September 10 close of 0.04406. On September 13, a Sunday, the intraday high reached 0.093—the highest price in these 90 days. Based on his purchase price, that high represented a gain of 111.08%, bringing his balance to 2017.56. But his rule required him to wait until Monday, September 14, to sell at the close. The closing price that day was 0.05566, locking in only +26.33% and bringing his balance to 1207.50. If Old Shen had held from the open to that high, his balance could have reached 2654.87 (+165.49%). Excluding this trade, Xiaoshu’s other 7 trades were actually losing trades.
On September 21, he sold his third trade, bringing his balance to 1292.56. He then stayed out until October 1, when he bought again. The highest closing price in these 90 days was 0.06934 on September 26. Old Shen’s balance at that close could have reached 1979.45 (+97.94%); Xiaoshu was out of the market that day, still at 1292.56. From the intraday high of 0.093 to the November 14 close, the price fell 55.17%; to recover, it would need to rise another 123.08. The two points were 62 days apart.
The Thursday, November 12 close was 0.04589, down 0.24% from the November 11 close of 0.046, so the rule would have triggered another buy. But the scheduled sell day was November 16, and that candle hadn’t closed yet, so this trade wasn’t included. November 14 was a Saturday; the close was down 9.47% from the November 13 close of 0.04605, closing at 0.04169.
Old Shen held all the way through and is up +19.01%. Xiaoshu had more losing than winning trades among his 8 trades, yet ended with almost the same amount, thanks to that one trade in mid-September—and he still didn’t capture Sunday’s high. The above is based on historical data from closed candles; it doesn’t indicate what will happen next and isn’t investment advice. If you were choosing between buying after a UTC Thursday close-down and selling the following Monday, or simply holding until now, which would you choose?
I ran this strategy on Binance spot STEEM/USDT closed daily candles from August 17 to November 14 (UTC, 90 candles in total). The November 15 candle hadn’t closed yet, so it wasn’t included. There were 13 UTC Thursdays during this period, each with a Wednesday close to compare against. Thursday’s close was higher than Wednesday’s on 4 days and lower on 9 days; there were no unchanged closes.
None of the 9 signals occurred while a previous trade was still open. On one occasion, the scheduled sell day’s candle hadn’t closed yet, so it wasn’t included. In practice, 8 trades were completed: 3 winners and 5 losers. No two trades were handed off on the same day, and there was no case of selling one day and buying again the next. After cycling 1000 USDT through 8 trades, the final amount was 1193.32, a compounded gain of +19.33%, averaging +2.64% per trade. The first 4 trades had 2 winners and 2 losers, with a compounded gain of +27.54%, reaching a balance of 1275.38; the last 4 had 1 winner and 3 losers, with a compounded loss of -6.43%. The best trade bought at the September 10 close of 0.04406 and sold at 0.05566 on September 14, for +26.33%. The worst trade bought at the August 27 close of 0.04507 and sold at 0.04308 on August 31, for -4.42%. Excluding the best trade, the remaining 7 trades had 2 winners and 5 losers, with a compounded return of -5.54%.
If Old Shen had bought at the August 17 open and held until the November 14 close at 0.04169, his 1000 USDT would have become 1190.12, a gain of +19.01%. Over the same period, $BTC was up +1.99% from the August 17 open to the November 14 close. Xiaoshu finished his 8 trades with 1193.32, almost the same as Old Shen’s 1190.12.
The turning point was a trade in mid-September. Xiaoshu bought at the September 10 close of 0.04406. On September 13, a Sunday, the intraday high reached 0.093—the highest price in these 90 days. Based on his purchase price, that high represented a gain of 111.08%, bringing his balance to 2017.56. But his rule required him to wait until Monday, September 14, to sell at the close. The closing price that day was 0.05566, locking in only +26.33% and bringing his balance to 1207.50. If Old Shen had held from the open to that high, his balance could have reached 2654.87 (+165.49%). Excluding this trade, Xiaoshu’s other 7 trades were actually losing trades.
On September 21, he sold his third trade, bringing his balance to 1292.56. He then stayed out until October 1, when he bought again. The highest closing price in these 90 days was 0.06934 on September 26. Old Shen’s balance at that close could have reached 1979.45 (+97.94%); Xiaoshu was out of the market that day, still at 1292.56. From the intraday high of 0.093 to the November 14 close, the price fell 55.17%; to recover, it would need to rise another 123.08. The two points were 62 days apart.
The Thursday, November 12 close was 0.04589, down 0.24% from the November 11 close of 0.046, so the rule would have triggered another buy. But the scheduled sell day was November 16, and that candle hadn’t closed yet, so this trade wasn’t included. November 14 was a Saturday; the close was down 9.47% from the November 13 close of 0.04605, closing at 0.04169.
Old Shen held all the way through and is up +19.01%. Xiaoshu had more losing than winning trades among his 8 trades, yet ended with almost the same amount, thanks to that one trade in mid-September—and he still didn’t capture Sunday’s high. The above is based on historical data from closed candles; it doesn’t indicate what will happen next and isn’t investment advice. If you were choosing between buying after a UTC Thursday close-down and selling the following Monday, or simply holding until now, which would you choose?