Of course Hillary and Bernie are at the top, since they are the only two major contenders on the Democrat side, so split the pool between them.
What's interesting is that how Scott Walker exited the race, and Jeb! scaled back his spending, because their donations dropped precipitously. Even though they got huge donations last quarter, they spent the money as fast as they could. Presidential campaigns are like venture capital that way: you spend money aggressively in order to make more money. If you are right, this strategy wins, if you are wrong, you go bankrupt quickly. (And going bankrupt quickly is preferable to being a zombie barely hanging in there).
Or, instead of the aggressive strategy, you could just play it cool for a bit, waiting for the press to get tired of Trump. As we say last election cycle, the press would suddenly get excited about a candidate for a bit, they'd shoot up in the polls, start to overspend, then the press would get bored, and their campaigns would go bankrupt. Those who lasted till the end were those who played a more conservative campaign. Maybe banking a bit of money now in order to tide you through the press's fickleness would be a good strategy.
Then there is the Rand Paul strategy. He's got a bunch of rabid followers (like me), so he can keep in the race until everyone has got tired of all the "normal" candidates. His father Ron used that strategy, and it works well.
Another way to look at candidate popularity (instead of polls and donations) is betting sites, as tracked by this site. While polls point to Trump and donations point to Carson, that site claims Rubio has the best chance of getting the Republican nomination by far. If you believe this is wrong, and if your beliefs are right, then you can go onto numerous online betting websites (including with Bitcoin) and make the appropriate bet.
In theory, online betting sides are the worst way to predict the winner. Yes, I said worst. That's because if there was enough liquidity in the market for them to be statistically valid, then there would also be enough liquidity for hedging. Consider Rubio. Now that he is leading among betters, he can then invest in betting on his competition. Should his campaign falter (donations drop, lose popularity), then the odds of his competitors will go up. He can thus sell those contracts and make a ton of money, enough to be overcome the drop in donations. Or, everyone can hedge their bets for candidates they know will be hostile to their interests. For example, all 1%ers should be buying contracts betting on Bernie Sanders winning. Sure, they'll lose money as Sanders taxes them to death, but they'll earn a lot of money at 10 to 1 odds on the betting sites.
There have been lots of article on the "wisdom of the crowds", but sadly, none have pointed out this hedging angle that makes their predictions inherently inaccurate.