When I said dividend, I should have specified 'special dividend' one that is paid out once in a large cash payment. "The upside on buying shares back is that i

When I said dividend, I should have specified 'special dividend' one that is paid out once in a large cash payment. "The upside on buying shares back is that it doesn't force people to reinvest in the company - it just shrinks the ownership base as the company shrinks." That upside is really not important for professional investors when dealing with mid cap or larger companies. Assuming best case that management teams are genuinely looking out for shareholders rather than inflating the stock price for their options, management teams are not very good at predicting when their own stock is under/over valued. Most investors want them to return cash to shareholders in most cases as liquidity for companies like IBM is so big that it doesn't matter if they must buy back shares. Since it is professional investor's jobs to value the share price, they are in a much better position to decide whether the share should be bought or not. Also, dividend stocks have supposedly performed better over history, but dividends is not a recognized alpha factor by the main finance research relied upon by market participants(1). Dividends go in and out of favor depending on a number of factors, right now, its probably the market's expectation of raising rates set against the background of increasing demand for income. Professional investors like dividends because they can simply put that money to work somewhere else. (1) There are papers that say dividends are a factor, but I'm not aware of one that is widely implemented. Anything beyond fama-french's 3 factor model doesn't have widespread acceptance yet. There are a number of other models in use, especially by quant funds, but they are not public and tested in public.