Jul 25, 2016 · 2 minutes

This is interesting: Since May Facebook’s shares -- which were having a comparatively torrid 2016-- have only increased some 2%. Meanwhile, shares in Twitter, which is still down 20% for the year, had a boost in May.

Morgan Stanley issued a research report today explaining why the reverse in fortunes happened-- and why investors were wrong to place those bets. (TL;DR: Sorry, Twitter.)

The concern around Facebook was that the growth of Snapchat may be eating into Facebook’s user engagement. Meantime, the excitement around Twitter was that the NFL deal, Olympics and political ads could give it a huge boost. Indeed, it’s not going to get much better for Twitter engagement than this election cycle. Donald Trump’s entire communication strategy centers on his unpredictable Twitter account-- making news even when he doesn’t use it, as he didn’t in the immediate aftermath of his wife’s plagiarized speech. That’s taken the whole social media battle for the White House to Twitter.

Unfortunately, this election is also mired in hate speech and Twitter continues to struggle mightily with bullying and trolls, losing more and more high profile celebrities to its unwillingness to meaningfully crack down on abuse.

Morgan Stanley focused on user engagement and found that it isn’t deteriorating on Facebook as a whole, which shouldn’t be a surprise. Facebook Live has stolen some of Twitter’s “real time” thunder, but more to the point, Facebook just has so many ways to win. It has the core site, Messenger which has 1 billion users now, and it’s still got Instagram and WhatsApp. Even if Snapchat is stealing Instagram’s thunder with a certain demographic, that’s only one part of the business.

Meantime, Morgan says engagement hasn’t bottomed at Twitter and is still deteriorating. From the report:

Has TWTR Engagement Bottomed and Are there Drivers for Upward Revisions in 2H:16? We don't believe so. Our user and app install analyses show continued user and engagement deterioration in 2Q:16, causing us to lower our ’16 MAU and ad revenue estimates by 1% and 2%, respectively. We agree TWTR will benefit from the NFL, Rio Summer Olympics, and political ad spend in 2H:16…but our new modeling where we break out each of these pieces in detail (See Exhibit 30) shows that even if these three events lead to $30mn in incremental 2H:16 revenue (which may be optimistic), TWTR is still likely to miss consensus estimates. 

This is the difference between a company that was paranoid early in its public company life, and one that was over-confident. Despite its far larger size, Facebook never assumed it would remain a dominant social platform-- or that even social media would stay the hot Web business investors wanted to back. Twitter, meantime, refused to put its weight behind once-hot companies like Vine and Periscope that it acquired, and refused to address a long standing problem with abuse and bullying, in an almost arrogant view that its core product would always be the place where sports, politics and awards shows would be discussed.

Both companies report earnings this week. Hope Twitter enjoyed this mini reversal in fortunes, because if Morgan Stanley is right, things may be about to take yet another ugly turn for Twitter even with a political/sports boost.