The bulls are kidding themselves if they think the sentiment data support their optimism.

To be sure, the data do provide some superficial support for their case: Stock market exuberance is slightly lower today than it was at the market's late-April high, even though the major market averages are higher today than then. Since bullishness normally tends to rise and fall with the market itself, these developments are encouraging from a contrarian point of view.

But only to a very limited extent. What the bulls are overlooking is that sentiment at stock market tops behaves far differently than at bottoms.

Sentiment's behavior in recent weeks is consistent with the market being in the middle of an extended topping process.

The key thing that many of the bulls are overlooking is that bottoms, but not tops, tend to be short and violent. In a typical major bottom, the market will plunge in one final cathartic seizure and then, on high volume, turn around on a dime.

Tops, in contrast, do not carve out an upside-down "V" pattern that is the mirror opposite of a bottom. Instead, the market will roll over, with different sectors hitting their respective highs over a relatively long period of time.

So the bulls can't conclude much from recent sentiment developments.

Take the last bull market top on the bull and bear market calendar maintained by Ned Davis Research, on May 19, 2015. That top ushered in a 9-month bear market that ended on Feb. 16, 2016. Though the broad market averages didn't fall enough during that period to meet the semi-official definition of a bear market as a drop of at least 20%, many secondary averages did. The Russell 2000 index, for example, fell by more than 26%.

Consider the behavior during that bear market of Investors Intelligence's sentiment survey of investment newsletters. At the May 2015 top, the percentage of monitored newsletters that were outright bullish was five percentage points lower than where it had stood the previous February, three months earlier.

In other words, that bull market top was not accompanied by a spasm of irrational exuberance followed by an immediate plunge. Had that happened, of course, it would have been the mirror opposite of how sentiment behaves in a typical major bottom. But that's not what happened; instead, the topping process took three months.

At the February 2016 bottom, in contrast, the bullish percentage in the Investors Intelligence survey hit its low the same week as did the market itself. The bottoming process, from a sentiment perspective, took only a couple of days.

This contrast reflects the experience of just one sentiment indicator in one particular bear market, of course. But an analysis of not just that indicator but also that of the American Association of Individual Investors and the Hulbert Stock Newsletter Sentiment Index shows that this experience in 2015-2016 was the rule, not the exception. Tops, in contrast to bottoms, are gradual drawn-out affairs.

None of this means we are at a major market top, I hasten to add. The bull market may very well last for much longer. My point is that the sentiment data do not provide support for that possibility.

In short: We need to exercise caution so as not to get caught up in the bullish euphoria that inevitably accompanies new all-time market highs.