[fairity]

> Generally speaking, once a fund reaches the level outperformance of ARK, they underperform the market.

Source for this? It would make sense to me if your claim was that future performance for this cohort is average. But, you seem to making a stronger claim -- that is, this cohort of outperforming funds performs worse than average going forward.

[cosmojg]

I think this makes more sense if you think of it in terms of mean reversion [1], particularly as distinct from regression to the mean, that mean being the market return.

Here are a couple articles that try to analyze the phenomenon [2][3].

[1] https://www.bogleheads.org/wiki/Mean_reversion

[2] https://peeranalytics.com/2018/05/21/hedge-fund-mean-reversi...

[3] https://www.valuewalk.com/2015/05/hedge-fund-mean-reversion-...

[fairity]

Thanks for the sources.

A quote from your second source: The average subsequent performance of the historically best- and worst-performing long U.S. equity hedge fund portfolios is practically identical and similar to the market return.

A quote from your third source: Due to hedge fund mean reversion, yesterday’s nominal winners tend to become tomorrow’s nominal losers.

So, it seems like there are differing conclusions based on how you slice the data. I'm familiar with mean reversion, but I don't understand why this would predict underperformance for outperforming funds. If you flip a coin 10 times in a row and get heads 10 times in a row, you don't expect to get heads less than 50% going forward (but do expect the average rate of heads to trend back towards 50% over time).

Furthermore, if OP's strong claim ever held true, surely there would be funds that outperform by simply indexing the broad market minus the holdings of top performing hedge funds. As more and more money gets invested in these new inverse fund, the pattern the strong claim is based on would slowly cease to exist. Unless you're saying the underperformance is due to fraud or exorbitant fees.

[inthewoods]

You can find lots of research on it but first, between 85-90% of actively managed investment funds underperform their benchmark:

https://www.ifa.com/articles/despite_brief_reprieve_2018_spi...

That means it is very, very hard to outperform.

Then, if you do outperform, that tends to be an exceptional year followed by not so great results.

https://www.morningstar.com/articles/1017292/what-to-expect-...

Key comment:

"Of the 123 stock funds that gained 100%-plus between 1990 and 2016, just 24 made money in the three years following their big gain, with the average fund losing around 17% per year."

Now, is it possible that ARK is an exception - of course. But the math would tell you that is highly unlikely.

[fairity]

Interesting read, but the data seems extremely skewed.

Quote from the same article: "Eighty-eight of the 123 funds were go-go tech/Internet darlings that soared in 1999 but crashed after, losing 24.1% per year from Jan. 1, 2000, to Dec. 31, 2002."

So, it seems like the data comprising the "after-years" is highly skewed towards years when the overall market crashed, which explains the negative returns.

If your claim is true, shouldn't there be a fund that simply indexes the broad market minus top performing hedge fund holdings? If not, why not?

[inthewoods]

Most active manager performance is mean reverting. They have a good year, and then there is little evidence of persistence in their future returns. Here's a larger study from S&P:

https://www.spglobal.com/spdji/en/documents/research/researc...

"We observe little to no evidence of performance persistence among active managers, except in the large-cap value and real estate categories. For example, out of 1,034 large-cap funds that existed in the universe as of Sept. 30, 2013, only 19.73%, or 204 funds, outperformed the S&P 500. In the following year, 15.69% of those 204 funds outperformed the benchmark. By the end of the third year, none of those original 204 funds were able to outperform the S&P 500 on a consecutive basis."

What is implied from this data is that if a manager has a good year, they are unlikely to match it going forward. So only 20% beat an index, and then only 16% of those that did beat it the next year.