I pulled the Binance SHIBUSDT daily candles. By UTC close, the most recent 90 completed daily candles run from August 12 to November 9; the November 10 candle is still in progress and is not included. The daily gains and losses mentioned below use both the day's open-to-close move and the previous close-to-current close move. The previous 20-day low is the lowest price among the 20 daily candles that had already closed before that day; for the first few days, this uses candles from before the 90-day window.
Over the 90 days, $SHIB went from an open of 0.00000449 on August 12 to a close of 0.00000443 on November 9, down 1.34%. Over the same period, $BTC rose 18.50%, and $ETH rose 13.42%. Of the 90 candles, 47 were bullish, 40 were bearish, and 3 had the same opening and closing price. On August 12, the price fell 2.00% from open to close, closing at 0.00000440; from that candle's close to the November 9 close, it rose 0.68%.
The highest price was 0.00000627 on September 22, and the lowest was 0.00000426 on November 8, a gap of 47 days. The price rose 47.18% from the low to the high; measured from the August 12 open to the high, it rose 39.64%. The November 9 close was 29.35% below the high, so it would need to rise 41.53% to get back there; it was 3.99% above the low.
Over these 90 days, I also counted how often the low fell below the previous 20-day low and whether the close got back above that level. This happened 11 times, and the close recovered above the level 9 times. In the remaining 2 cases, although the day's low fell below it, the close stayed below the level.
In both cases where the close failed to recover, the breach was greater than 1%. There were 5 breaches of less than 1%, and the close recovered in all of them. Of the 6 breaches of at least 1%, the close recovered in 4. The 2 cases where it did not were October 29 and November 8: on October 29, the previous low was 0.00000476, and the day's low reached 0.00000460, 3.36% below the previous low; the close was 0.00000463, still 2.73% below it. On November 8, the previous low was 0.00000444, and the low reached 0.00000426, 4.05% below it; the close was 0.00000437, still 1.58% below. The November 8 low was also the lowest price of these 90 days.
On October 29, the open was 0.00000497 and the close was 0.00000463, a 6.84% drop from open to close and also a 6.84% drop from the previous close. The next day, October 30, closed at 0.00000481, up 3.89% from the previous close. Its low was 0.00000446, which dipped below the previous 20-day low of 0.00000460, but the close recovered above that level, finishing 4.57% higher. Looking further ahead, the close 5 days later was unchanged from the October 29 close. By November 9, the close was 0.00000443, down 4.32% from the October 29 close and 6.93% below the 0.00000476 level that had failed to hold that day.
Only 1 day has elapsed since November 8. On November 9, the open was 0.00000437 and the close was 0.00000443, up 1.37% from open to close and also up 1.37% from the previous close. It was 3.99% above the November 8 low, but still had not reached the 0.00000444 level that failed to hold that day, remaining 0.23% below it.
Split the 90 days into three periods. From August 12 to September 10, the price rose 12.25% from open to close; from September 11 to October 10, it rose 7.54%; and from October 11 to November 9, it fell 18.42%. The three periods had 15, 16, and 16 up days, respectively; 15, 12, and 13 down days; and 0, 2, and 1 days with the same opening and closing price. Over the most recent 14 days (October 27 to November 9), the price fell 9.41% from open to close; over the most recent 7 days (November 3 to November 9), it fell 3.28%. Average daily trading volume over these 90 days was about 4.4883 million USDT, and over the most recent 7 days it was about 1.8821 million.
The price action 1, 3, and 5 days after the close recovered in those 9 cases was mixed. After 1 day, the average gain was 0.16%, with 5 advances and 4 declines among the 9 cases; after 3 days, the average gain was 0.30%, with only 3 advances. Of the 8 cases with 5 days of data, the average gain was 0.77%, but this was mainly driven by the 21.24% gain following the September 16 case. Excluding that case, the remaining 7 cases were down an average of 2.15% after 5 days, with 6 declines and just 1 advance. Both cases where the close failed to recover were up the following day, by an average of 2.63%; the October 29 case was unchanged after 5 days. The sample is small, especially with only 2 cases where the close failed to recover, so it doesn't show that “a recovery means the price will rise” or that “if it fails to recover, it will keep falling the next day.”
These figures only show that over these 90 days, $SHIB when the price touched the previous 20-day low, the close recovered more often on days when the breach was less than 1%; in all 5 cases of a shallow dip below the level, the close ended back above it. They cannot tell you that 0.00000426 is the bottom, or that the close will necessarily fail to recover the next time the price breaches the level by 1%. This is a measure of the price action over these 90 days, not a prediction of future movements.
When you look at a pattern like $SHIB , where the price dips to the previous low multiple times but the close often recovers above the level, would you pay more attention to whether all 5 shallow dips recovered, or first check whether the price quickly returned to that level after the two times it failed to hold?