Jim Simons is a decorated mathematician (known for his work on the Poincare conjecture) but is probably better known as the founder of the wildly successful hedge fund Renaissance Technologies (commonly known as 'RenTech'). His Medallion Fund has racked up a phenomenal return since 1988 (more than 30% annualized) and was up 80% in 2008. The Medallion Fund has capacity constraints, however, and is now closed.
As a result in 2005, Simons launched the Renaissance Institutional Equities Fund (RIEF) with the expectation that it would have a much larger capacity (on the order of $100Bn dollars). The mandate is to outperform the S&P 500 index by 4-6% each year and would hold longer duration trades.
In 2008, RIEF lost 16% and was down 6% in 2009. In 2010, RIEF returned 16.5%.
We take a look at Simon's 13F filings over time to get an idea of what the core names in the REIF portfolio have been:
as can be seen, many names are held for several quarters at a time and the turnover in the top 10 names isn't terribly high. Which then begs the question, how would a 13F "replication" strategy work on this data?. If we were to buy the top 5 or 10 positions (as a percentage of total reported longs) each quarter after the 13Fs were released and held those positions for the following quarter...what would that return stream look like?:
…buying only the top position in RenTech's filings each quarter is a little volatile, but buying the top 5 or 10 has actually done very well over the last few years. Buying the top 5 positions in RenTech's 13F (the week after it's released) would've earned 95% in 2009 (vs. 26.4% for the S&P), 20% in 2010 and down only 1.8% in 2011 (vs. about -6% for various Long/Short Equity Hedge Fund Indices):