Sunday, July 15, 2012

In a Hedge Fund13F Replication Strategy, Does the Delay Matter?

An oldie but goodie from World Beta (Mar. 20, 2012):
13Fs: Does the 45-Day Delay Matter?
I’ve written dozens of posts here on the blog since ’06 on the topic of 13F research (older post here for example).  An analyst and I used to cobble together the 13Fs and backtests by hand, a long an arduous process.  However, over the years that process has allowed a number of insights to flow through that are hard to realize any other way than just getting your hands dirty.  Findings such as what funds to track, what holdings to track, what strategies to track, etc have been discussed here and in our book The Ivy Portfolio – often insights that have not been presented elsewhere.  So, below I wanted to tackle a topic that I hear almost every single time 13Fs come up – namely, does the 45 day delay matter in tracking these top hedge fund manager stock picks?

Below we do a quick test, and note it is not comprehensive.  There are inherent biases no matter how you chop up the data (how many funds to include, long/short only or entire universe, include dead funds, regress the returns based on turnover and AUM? etc etc)  but we look at 20 funds we have been following for years on the blog.  We compare reblancing on the 13F filing date to rebalancing a portfolio at the quarter end (ie look ahead bias investors do not have).  It shows how a portfolio constructed without the 45-day delay compares to a portfolio with publicly available information.  Tests go back to 2000, total return data with no transaction costs.

Q:  So, does the 45-day delay matter?
A:  A little.

Some of our readers may recognize that answer as the punch line to a paratrooper joke.