Is it possible that it's not supply and demand which most often determines price direction but market sentiment?
Not to get all geeky or anything, but if you want to think Old School a minute, the initial question is in deep error: there is no contradiction between "market sentiment" and either of aggregate supply or aggregate demand.
In implicit form Demand is a function of [Price; Income; Market Choices {substitutes/complements}; Tastes] -- and given a hefty dose of
De Gustibus Non Est Disputandem, we don't argue about such things as Market Sentiment.
Re Supply -- S = f[ Labor, Capital, Technology] where, playing down the eminent false dichotomies in *each* of these productive assets, the most liquid one: Money, will go elsewhere when alternative uses project a better expected return. Thus, with lesser prospects in other markets, capital would be drawn to this market, available supply grows, market price falls, expected return (eventually) declines, and the capital is withdrawn/re-allocated. (And, the same in reverse: as expected returns in other markets move sentiment in the instant market to lessen, capital is withdrawn, and a declining cycle is initiated.)
Now, the slight twist here is that we're thinking about an input market, not a product market -- so "prices" become ROICs, and capital migration is seen by asset (equity shares) sell-off -- much like the market for (any other) financial products, like loans or mortgages or whatever.
But still, Market Sentiment is very much an integral part of any definition of either Supply or Demand -- and it's plenty perilous to attempt to separate them.