Subprime Mortgages Weren't Main Trigger of Housing Bubble. Or Were They?

This story is puzzling. Apparently a new academic study found that "subprime lending was not the main trigger of [the] real estate bubble." I think most housing market observers would agree with that—subprime mortgages didn't really take off until well after the beginning of the housing bubble. So if subprime mortgages weren't the primary factor, what was? According to the study, the primary factor was:

the considerable 2003 pullback of government-sponsored financial service corporations Fannie Mae and Freddie Mac from the credit market and their replacement by aggressive, private mortgage securities issuers.
...
[I]n 2003, political, regulatory and economic factors . . . forced the two [GSEs] to significantly slow their lending volume. Private funding in the form of asset-backed securities and residential mortgage-backed securities replaced conventional, conforming mortgage-backed securities as the prevalent source of mortgage capital.

The new credit environment allowed looser underwriting standards and increased tolerance for riskier, high-yield loan products. Such products included adjustable-rate mortgages with low initial “teaser” rates, Alt-A loans that did not require income verification and nonowner-occupied investor products. This borrowing climate provided previously marginal borrowers with additional access to credit. The credit market shift led to a record increase in total mortgage volume and pushed up home prices with momentum characteristic of a bubble.

Huh? "Subprime mortgages" are, by definition, mortgages that don't meet Fannie and Freddie's conforming loan guidelines. If Fannie and Freddie's pullback from the credit market, and the resulting rise of mortgages that didn't meet Fannie and Freddie's guidelines, was the primary factor, then the primary factor was subprime mortgages.

Now, it's possible that the authors of the study use a different definition of "subprime" (though I doubt it, since it would be incredibly misleading not to use the widely-accepted definition of "subprime" in a paper about the housing bubble). It's also possible that the article simply mischaracterizes the study's findings.

Am I missing something, or does this article just not make sense?