The hedge fund industry is trying to get rid of its biggest problem

The hedge fund industry has a crowding problem.

There are, it seems, too many hedge funds making too many of the same bets.

As a result, hedge fund performance has lagged relative to broad stock indexes.

But according to Goldman Sachs' latest hedge fund trend report, funds spent the second quarter of the year trying to undo at least some of this crowding that, it seems, has led to stagnant industry-wide returns.

"Although hedge funds maintained their sector and factor tilts coming into 3Q, the extended period of popular long position underperformance reduced the amount of hedge fund crowding relative to 1Q 2016," Goldman wrote in its report.

"Exhibit 8 shows a measure of hedge fund 'crowding' calculated as the effective number of stocks in the aggregate hedge fund portfolio. Using this measure, crowding reached a record high in 1Q 2016. It then declined sharply in 2Q as funds spread assets into new investment ideas in response to the painful and prolonged underperformance of hedge fund VIPs. Similarly, our VIP basket saw a turnover of 18 stocks this quarter, only modestly above the long-term average of 16 stocks per quarter but the highest rate since 2012."

In English, Goldman's analysis effectively says hedge funds are still betting on what they, as an industry, have been betting on - "maintained their sector and factor tilts" - but also made a major effort to put money into new ideas during the quarter.

But if the hedge fund industry's performance problem is about too many funds being in too many of the same stocks, the density of hedge fund portfolios still presents some major barriers.

The following chart from Goldman shows that almost 70% of hedge fund assets are comprised of the average fund's top 10 holdings. Said another way, the industry is mostly betting on just a handful of ideas.