The market discounts what it thinks will happen, not what will happen. When the market is totally wrong about what will happen (e.g. 99-2000, early 2001, 2006-07), you get a large price move totally at odds with what accurate forecasters & analysts think will happen. The latter get caught because they underestimate the degree to which the market can be totally wrong for quite some time.
If you can forecast changes in market expectations, but not what will happen on a fundamental basis, then it makes sense to play the market move, even if it will ultimately reverse and go lower. Your general approach should be to have on a position consistent with your long-term view most of the time, but to book profits and make counter-trend trades when market expectations are due for a shift (e.g. at price/sentiment blowoff extremes).
If you cannot forecast changes in market expectations, but you can forecast fundamental developments, then you need a strategy for situations like the last 2 months, where market action and likely future fundamental action diverge significantly. It was just the same in early 2001 - tech staged a spectacular rally after the nasdaq had halved in a few months. That rally failed and the market ultimately had another huge fall to new lows. Ditto for the Nikkei during the 90s, there were many significant rallies despite ultimately going much, much lower.
Anyone trading purely on fundamental forecasts, who does not have any timing ability, needs to be aware that at any point there could be a 50% counter-trend rally in the index, *even though they are right about it going lower*. 1929-32 had a 50% sucker rally, Nikkei 1990-2003 had more than one 50% sucker rallies. Asia 1997-98 had a big sucker rally late in 97/early 98. Brazil 2000-2002 had a 30% sucker rally.
That means you either need to develop some timing skills, trade small size, be willing to take gigantic drawdowns, or purchase out-the-money options to limit your losses in the event that a large counter-trend rally takes place.
One important implication of all this is that when market expectations are going to shift significantly, then (if you can forecast this) it makes sense to flip your positions from short to long, *even if you know you are right in the long-term*.
