On March 23 the S&P 500 closed down nearly 34% from all-time highs seen just the month before. It was the fastest 30% bear market from all-time highs in stock market history.

Then without warning the market snapped back in a big way, rising nearly 18% in just three days and 31% through the close May 11.

The swift decline in stocks was painful but understandable. The virus closed off most of the economy on the spot in March. Nothing like this had ever happened before.

The snapback rally is more of a head scratcher for most investors. How could stocks continue rallying in the face of the worst economic data since the Great Depression? Tens of millions are unemployed. Thousands of businesses are gone. Many more will follow.

One explanation would be the 34% drop from late-February through late-March was pricing in those economic numbers. The simplest explanation would be the trillions of dollars of support from both the federal government and Federal Reserve. This was the fastest, sharpest bear market in history but also the fastest, biggest stimulus package ever assembled.

There is, however, a third possibility. Many of history’s great crashes have exhibited head-fake rallies that offered investors a false sense of hope that proved to be fleeting.

During the Great Depression stock market crash there was a 47% rally from late-1929 until the early Spring of 1930. It did not last of course. Before that rally stocks had fallen 45%. After rising almost 50%, they would go on to fall by more than 80%.

That crash also included monthly gains of 8%, 9%, 12% and 14% before all was said and done along with rallies of 23%, 27% and 35%. I cannot imagine the amount of false hope each of these rallies must have given investors.

The nasty 1973-1974 bear market that cut the market in half saw a 20% bounce before it was over. Even the 2007-2009 market crash gave investors a false glimmer of hope with a gain of more than 25% in the midst of the crash that was eventually relinquished.

The bear market from 2000-2002 saw three separate rallies of around 20% before finally settling in at a bottom more than 50% lower than the peak. On a spreadsheet, it often looks like market crashes are a straight line down but that’s typically not the case:

The Great Depression is always the worst-case scenario but the dot-com crash may be second on that list. It included not only the bursting of the tech bubble but also the Enron scandal and 9/11.

Even after bottoming in October of 2002 and quickly rising more than 20%, there was yet another 15% decline before finally taking off until 2007.

There’s a good reason why it’s so difficult to tell the difference between a dead cat bounce within the context of a bear market and an actual market bottom. When stocks do eventually bottom, they tend to see strong gains coming out of the gate.

Here are the 3 and 6-month returns from the bottom of past S&P 500 bear markets:

I am sure every one of these recoveries was called a dead cat bounce at the time. It’s hard to trust the market to give you gains when all it’s done recently is take them away.

The stock market can move hard and fast in both a dead cat bounce and a bear market bottom. There are people who will tell you with certainty that they know for sure whether this bounce or the next is the true bottom of this market crash. There are always backtests or historical parallels to back any market stance.

It will only look clear with the benefit of hindsight. This crash has been so swift and severe that it will likely play by its own set of rules. That means no warning ahead of time before it bottoms and no certainty about the difference between a dead cat bounce and an actual end of the bear market.

The only trade I’m certain about at the moment is going long humility and short hubris.

Ben Carlson, CFA, is the director of institutional asset management at Ritholtz Wealth Management. He may own securities or assets discussed in this piece.