Surprisingly there are TWO answers and they are polar opposites and therefore a bit surprising.
But before we get to that, let me tell you what happened to me. I arrived just in time for several major crises – the Asian currency crisis was first. When Thailand devalued and others followed, the hedge fund investors who had borrowed money to amplify their returns were forced to sell investments to raise cash to pay their lenders. Since it was very hard to sell the securities that were related to the devaluations, they sold other things easier to sell.
Since these investors specialized in emerging markets, this caused a contagion that pushed other markets down that had nothing to do with the problem currencies. It occurred to me at this point that my business school theories had failed to capture what was really going on in the market. The securities that had temporarily dropped in price simply because traders needed to raise cash, bounced back not long later – producing fantastic returns for those investors willing to buy when so many needed to sell.
With this lesson in my head, I was ready for the next crisis. When Russia defaulted on their domestic debt, I was ready. I bought closely related securities that were under pressure from the resulting wave of selling but which I felt confident would rebound to produce good returns. In 1999 I made a 75% return buying bonds.

So answer number one is:

The academic community had also figured out that markets were not really efficient. A string of research papers had shown systematic mispricing relative to the efficient markets ideal. But that’s not really the whole story when it comes to mutual funds. More than half of them will typically not even match the market index returns. Are they just not smart? Despite apparent mispricing in securities markets, the academic studies up to that time had failed to find any evidence that fund managers could exploit the opportunities available to beat the market index other than by chance.
That really bothered me because I knew it could be done (I was doing it myself). Being an academic at heart, I was dismayed that no one had explained why we had opportunities and we had smart fund managers, but (except for a few exceptions) rarely could funds add value versus the index funds.
Finally a couple of professors named Cremers and Petajisto solved the puzzle. They came up with a statistical methodology that sorted the fund managers into different categories and this allowed them to find a subset that could consistently beat the index funds on average. Their key insight was a measure they called Active Share. With this, they were able to explain what was going on in the fund industry.
When mutual fund companies become large and well established the company owners and managers can earn high profits for themselves by marketing their fund well and avoiding significant risks of under-performing by noticeable amounts. Thus they risk much and gain little by trying to beat the index when they have reached a certain level of assets and tenure.

Thus answer number two is: