If most actively managed funds don't beat the index why not choose the opposites?

Question

This is probably a dumb question and I'm likely missing something obvious but here's my reasoning-- where am I going off the rails?

1) Most actively managed funds do not beat the stock market index over the long-term.

2) This implies that fund managers are performing worse than a random guess. Possibly because industries/companies that the managers think will do well appear to be attractive to many people at the same time so they result in short term gains but quickly become overvalued.

3) Therefore, if we took a total stock market index and sold the funds that active managers were purchasing, and buy the ones that they sold this collection of stocks should outperform the market on average over the long term.

Basically if (A + B)/2= C Where A is the long-term growth of the stocks chosen by managers and B is the long term growth of all other stocks and C is the long-term average of the stock market and A is consistently lower than C then that implies that B is higher.

Or is it just that managers are no better than chance and the reason that they under-perform is that the MER and transaction costs eat into their profits?

Answer 1

@BenVoigt's answer is pasted below so people searching for questions with no answers can avoid this one.

Your final paragraph is correct. The managers are not enough better than dumb luck to pay the management and trading costs. But they aren't necessarily worse than dumb luck either, certainly not by enough to make your "opposites" proposal work.

Answer 2

There are a number of factors that will not make your theory work. First off actively managed funds do tend to profit in most years. So if you do exactly the opposite, you would lose money.

Secondly there is a cost to trading. When you buys-sell-buy your costs are triple that of those that just buy. Basically that is what an index fund does, just buy. Doing the opposite of an actively managed fund will carry the same trading costs.

Third there is management cost associated with the fund. Those of index funds tend to be a lot less because a research staff is not needed, nor the IT tools. Just buy what is in the index. Doing the opposite would likely reduce the management cost, because in a way decisions are made for you. How would you collect the data though?

I did a research project using AI on picking stocks. The problem I could not solve is an answer to "when do I sell"? I see this an an incalculable problem. Examining the chart of FB is illustrative. On 3/16/18 FB was $185, by 4/6/18 it was 157, on 5/11/18 it was 187. Depending on the day/hour/second you sold and re-bought could vastly change the profitability of the fund. Passive funds solve this problem by not selling until the stock is kicked from the exchange. Active fund guess and often guess sub-optimally.