Imagine someone named Xiao Ran, who does accounting at a tax office in a county town. He closes the books by calendar month. Some of his colleagues have been holding on without selling, but he thinks positions that span months make the books hard to reconcile, so he set a rule for $RUNE , using 10000 USDT as the starting capital: by UTC calendar month, buy at the close of the first completed daily candle of the month, then sell at the close of the last completed daily candle of the month, ignoring everything that happens in between. If the window ends before the month is over and he is still holding, value the position at the latest close.
We recalculated this using Binance spot daily candles (UTC, August 24 to November 21, 2026, 90 completed daily candles). The window only starts on August 24, so August 1 is outside these 90 days and there is no buy signal for him in August. He stays out of the market for those eight days.
While he is out, the close on August 24 is 0.480, and the close on August 31 is still 0.480, a close-to-close change of 0.00%. But on August 30, it closes at 0.457, the lowest close in these 90 days, before returning to 0.480 the next day. He doesn’t buy until September 1, when it closes at 0.470, lower than the close on the first day of the window.
The September trade goes well. He sells on September 30 at 0.800, for a monthly gain of 70.21% and cash of about 17021.28. The highest close in the entire window is also on September 30, so his monthly sell date coincides with the highest close. The next day, October 1, it closes at 0.778, 2.75% below his sell price, and he buys back in according to the rule.
On October 4, it closes at 0.796, bringing the value on paper to about 17415.09. But the rule says to wait until the end of the month to sell. He sells on the 31st at 0.631, taking an 18.89% loss on this trade and leaving him with about 13805.17 in cash. When he buys again on November 1, the close is already down to 0.553, 12.36% below the October 31 sell price. It closes at 0.640 on the 4th, then at 0.486 on the 20th, putting the value on paper at about 12132.58. The window ends on the 21st, when it closes at 0.491. The month isn’t over yet, so he is still holding, with about 12257.40 on paper—22.57% above his starting capital. He bought three times and sold twice.
· If he had put the full amount in at the August 24 close of 0.480 and held until November 21, 10000 would have become about 10229.17, a gain of just 2.29%, about 2028.23 less than under his approach. The deepest drawdown while fully invested would have occurred on November 20, at 39.25%, with about 10125.00 on paper. His approach had a maximum drawdown of 30.33%, also on November 20, with about 12132.58 on paper;
· On November 21, the price is 38.63% below the highest close of 0.800, and would need to rise 62.93% to get back there;
· Over the same period, $BTC fell from 78992.75 to 60215.77, a decline of 23.77%. The same 10000 would have become about 7622.95. Using the same monthly buy-and-sell rule, it would have become about 7585.32, a decline of 24.15%.
Xiao Ran looks back and forth between the September 30 and November 1 rows. “The highest close landed on the day I sold for the month. But after I sold on October 31, the price had dropped further by the time I bought back in the next day. November isn’t even over, and the value on paper has already fallen from 17415.09 to 12257.40.”
I didn’t tell him he was clever to sell in September, or advise him not to buy again in November. These are all historical, completed daily candles. The latest one is November 21, and no one knows what happens next. In this window, he ended up about 2028.23 ahead of simply holding the whole time, but from the peak value on paper of about 17415.09 down to 12257.40, about 5157.69 has already been given back. Choose a different window, and both 22.57% and 2.29% would change. Without that +70.21% trade from September 1 to 30, the earlier figures would also be different. November isn’t over yet, so this rule hasn’t run its course for the days after the 21st.
Have you ever been in a situation like this: you sell at the highest close when you settle up at the end of the month, then buy back at a lower price the following month—and before the books are even closed for that month, the value on paper has already given back a chunk?
We recalculated this using Binance spot daily candles (UTC, August 24 to November 21, 2026, 90 completed daily candles). The window only starts on August 24, so August 1 is outside these 90 days and there is no buy signal for him in August. He stays out of the market for those eight days.
While he is out, the close on August 24 is 0.480, and the close on August 31 is still 0.480, a close-to-close change of 0.00%. But on August 30, it closes at 0.457, the lowest close in these 90 days, before returning to 0.480 the next day. He doesn’t buy until September 1, when it closes at 0.470, lower than the close on the first day of the window.
The September trade goes well. He sells on September 30 at 0.800, for a monthly gain of 70.21% and cash of about 17021.28. The highest close in the entire window is also on September 30, so his monthly sell date coincides with the highest close. The next day, October 1, it closes at 0.778, 2.75% below his sell price, and he buys back in according to the rule.
On October 4, it closes at 0.796, bringing the value on paper to about 17415.09. But the rule says to wait until the end of the month to sell. He sells on the 31st at 0.631, taking an 18.89% loss on this trade and leaving him with about 13805.17 in cash. When he buys again on November 1, the close is already down to 0.553, 12.36% below the October 31 sell price. It closes at 0.640 on the 4th, then at 0.486 on the 20th, putting the value on paper at about 12132.58. The window ends on the 21st, when it closes at 0.491. The month isn’t over yet, so he is still holding, with about 12257.40 on paper—22.57% above his starting capital. He bought three times and sold twice.
· If he had put the full amount in at the August 24 close of 0.480 and held until November 21, 10000 would have become about 10229.17, a gain of just 2.29%, about 2028.23 less than under his approach. The deepest drawdown while fully invested would have occurred on November 20, at 39.25%, with about 10125.00 on paper. His approach had a maximum drawdown of 30.33%, also on November 20, with about 12132.58 on paper;
· On November 21, the price is 38.63% below the highest close of 0.800, and would need to rise 62.93% to get back there;
· Over the same period, $BTC fell from 78992.75 to 60215.77, a decline of 23.77%. The same 10000 would have become about 7622.95. Using the same monthly buy-and-sell rule, it would have become about 7585.32, a decline of 24.15%.
Xiao Ran looks back and forth between the September 30 and November 1 rows. “The highest close landed on the day I sold for the month. But after I sold on October 31, the price had dropped further by the time I bought back in the next day. November isn’t even over, and the value on paper has already fallen from 17415.09 to 12257.40.”
I didn’t tell him he was clever to sell in September, or advise him not to buy again in November. These are all historical, completed daily candles. The latest one is November 21, and no one knows what happens next. In this window, he ended up about 2028.23 ahead of simply holding the whole time, but from the peak value on paper of about 17415.09 down to 12257.40, about 5157.69 has already been given back. Choose a different window, and both 22.57% and 2.29% would change. Without that +70.21% trade from September 1 to 30, the earlier figures would also be different. November isn’t over yet, so this rule hasn’t run its course for the days after the 21st.
Have you ever been in a situation like this: you sell at the highest close when you settle up at the end of the month, then buy back at a lower price the following month—and before the books are even closed for that month, the value on paper has already given back a chunk?