If you got in low, it's been really hard to not make money.Absolutely true.
But there have been plenty of down days, and what's suspicious is that Goldman made money on all of these except one (though see my final statement in this post - I don't fully believe that the numbers that are posted mean what they are supposed to mean).
Since July 3 (random date a few months ago), if you look at S&P 500 returns, there have been 50 up days and 36 down ones. To make money on each and every one of those movements screams "exploited loophole", if that's actually what happened. I'll leave it to other interested parties to calculate the probability of that happening by chance; it depends on your assumptions, but there's no way that it's significant enough to consider.
Or maybe they've fully solved the problem of market timing. But they're really just not that smart (there's not an analyst alive that believes you can nail day to day movements at 99% accuracy).
In this field, another maxim applies: never attribute to genius that which can be adequately explained by malice. Here, we've got motive and opportunity for malice, so I'm very hesitant to believe that they're playing fair...
However, I also don't necessarily believe that these numbers mean what we think they mean, since I don't know how they were generated, or what sorts of accounting or trading tricks may have been played to "smooth out" the risk profile and ensure very few losing days, as long as the long term average was positive. The appearance of zero-risk trading has some very tangible benefits, and there's a lot of stuff you can do to shift a distribution of returns so that it looks a lot less volatile than it really is...that whole CDS bonanza had a lot to do with shifting risk to the tails to snatch at predictable returns the rest of the time, so who am I to say there's not something similar going on here?