Following Hedge Funds Through 13F Filings Is Not a Good Idea

by: M F

Okay, maybe I should be more clear. Blindly following hedge funds through 13F filings is a terrible idea. There are countless sites that focus on hedge fund filings and they are not going to tell you they offer little value, because then they have no reason to exist. Building a backtesting engine is not a trivial exercise (believe me it is incredibly non-trivial to do it correctly). Some of the i-banks offer products based on hedge fund tracking, but most base their data off FactSet/LionShares which has survivor bias, rendering it dangerous.

To give you an example, let’s take a look at the most viewed portfolios on StockPickr. James is a friend of mine and I am not trying to pick on him here but his is the best of the free sites (including two that give me motion sickness). The top five most viewed portfolios on StockPickr are: Buffett, Soros, Icahn, Pickens, and Renaissance.

But here is the problem - the investor has no idea if following these funds is worthwhile. These five portfolios have been viewed over FIVE MILLION times.

Here is the kicker:

If you take a simple portfolio of the top 10 holdings, four of the five portfolios have no excess returns over their benchmark. On average these fund clones underperformed their benchmark by 10% (i.e. I used the S&P 500 except in the case of Pickens where I used S&P Energy). If you include transaction costs, it is even worse. In other words, investors are wasting their time and their money. In many cases it is costing them money by tracking funds that underperform the S&P 500. On average these funds did about 10% worse than the S&P 500 since 2000.

Using AlphaClone I have been able to identify value added algorithms and alpha in the areas of:

1. Persistence in Returns - Do top managers continue to outperform? Are there factors that contribute to over/under performance?

2. Manager Selection - Can you identify managers ahead of time that will outperform, and if so, how?

3. Static Hedging - Is it possible to hedge the portfolios in a way to increase risk-adjusted returns, and if so, how?

4. Dynamic Hedging - Is it possible to dynamically hedge the portfolios in a way to increase risk-adjusted returns, and if so, how?

5. Security Selection - What holdings yield the best return for a manager?

6. Custom Strategies - Are there other strategies that can yield better returns? What about the most popular stocks? New buys from the manager?

All 6 of these valued added findings cannot be learned from sites that do not let you backtest properly to 2000 without survivor bias in filings and stocks (this is huge by the way).

No, I am not going to tell you what these findings are (they will be the basis for some upcoming funds), but if you are an AlphaClone user and find some gems drop me a line and I am happy to chat with you. It is something I have not published nor intend to.

And just so you don’t think I am grumpy I will throw all of these sites a bone - following the top holdings is not the best way to follow these funds.

[And don't even get me started on the social investing sites - I am biased here, but I would much rather Warren Buffett, Bruce Berkowitz, David Einhorn, and David Dreman manage my money than some teenager on his computer trading microcaps with 10,000 shares daily traded. Maybe a longer post on this topic later].

How about a little quiz? The first correct answer in the comments will get a free copy of my book The Ivy Portfolio when it prints in a week…

What fund manager’s top 10 holdings' clone beats the market by 10% a year since 2000 (and actually outperforms his fund)? The current holdings are:

WLP

UNH

WCG

URI

TAL

FUR

WTM

WSC