Argh! Yahoo has gone downhill again. Now the charts are garbage.
</rant>

This can be backtested if we define a "steep market decline" (when moving into junk) as a decline of 50%, and "recovered market" (moving back to the index) as the S&P500 reaching its former high. The junk could be defined stocks that have fallen at least 75% from their former high...

Anyway....
SPY (S&P500) has only had one decline of 50% since inception of 1993. That was the 2008 bear market.
Only 2 declines more than 40%, that was the 2002-2003 market.
Two more at 25%, 2011 and 2020.

There are not enough "big" falls to test this theory with.

I suspect that there wouldn't be a good way to reliably pick the junk stocks that will rebound greater than the market. There's just too many junk stocks, and most of them are just that -- JUNK.

One potential candidate might be the opposite of the large S&P500 stocks but not quite junk, like the S&P 600 small cap ETF. But when I plotted SPY & IJR on testfol.io, they fall & rise in unison. But the embarrassing thing is, since inception (May 2000), IJR beats SPY handily. More volatile, though.

One thing that I have reluctantly become convinced of is that for a long-term investor, trying to do timing is futile and counterproductive. Well-defined timing rule-based strategies of course, not handwaving ad-hoc untestable strategies.

All that timing does (a well-defined good strategy, that is) is to reduce the volatility and avoid the large drawdowns. But they also miss out on some of the subsequent gains. Long-term, the return is slightly reduced over buy-and-hold.

The biggest advantage of timing is when the portfolio is short term, like for a 75+ year old retiree. Instances where the bounce after the crash may not come before the end.