Why 13F Filings Cannot Reproduce a Quant Fund’s Performance
Summary
The discussion critiques attempts to infer a quantitative hedge fund’s returns by equal-weighting its reported stock holdings. Such estimates are unreliable because 13F disclosures arrive quarterly and with a delay, and cover only US long equity positions. They omit short positions, derivatives, and non-US holdings, leaving much of a fund’s actual portfolio and its hedges unknown.
The response also notes that filings reveal positions without explaining their purpose. A reported stock may reflect a directional view, a sector exposure, or another component of a larger strategy. For that reason, 13F-based tracking is more appropriate for low-turnover, long-only investors than for a complex quantitative fund. The discussion distinguishes the performance reputation of a closed fund from that of other funds offered to outside investors, while acknowledging uncertainty about the details. It provides limitations of the filing-based analysis rather than a complete performance assessment.
Key ideas
- 13F filings report US long equity positions on a delayed quarterly schedule.
- The filings omit shorts, derivatives, and international positions, obscuring the full portfolio and its hedges.
- A disclosed holding does not reveal why a fund owns it or how it fits into a larger strategy.
- Filing-based tracking is most informative for low-turnover, long-only US equity investors.
- The reported performance of one fund should not be assumed to describe a firm’s other funds.
Tags
Full text
# Specific Hedge Fund Filed Returns # Specific Hedge Fund Filed Returns https://mebfaber.com/2008/12/27/tracking-jim-simons-renaissance-technologies/ But, like always, I will let the data speak for itself. Top 10 holdings, back to 2000, equal weighted through 12/19/2008: Fund Data Annualized Return: -9.8% Volatility: 19.5% MaxDD: -69.1% (Max DD is maximum peak to valley drawdown, measured monthly.) Maybe, just maybe they didn't do that good? Else But why are they so bad at us stock investing? Why do they make their no employee funds bad for no reason? And by the way 13f is the net long position, any hedge position which is net long would appear. Sales would appear as well so the information has to be accurate. I guess the answer is assuming only their short positions matter? Sounds strange. ## Answer by Helin (score 6, accepted) https://quant.stackexchange.com/a/49792 The short answer is that these simulated performance based on public filings is more or less useless. - 13Fs are only filed on a quarterly basis with a meaningful delay. By the time you see the filings, the positions may have changed drastically. - 13Fs only include US long positions. We don't know what shorts are paired against these longs; we don't know what derivatives are in the portfolio; and we have no visibility into non-US holdings. - Along the same line, because of RT's trading style, we have no insights into why a position exists in the first place. Did they buy a stock because they liked the company, or is it part of a sector bet, or is it just a placeholder for some unknown factor? Without these insights, blindly following the positions can be problematic. So 13Fs are really only useful if we are tracking a low turnover, mostly long-only investor making equity investments in the US. This is mostly true for funds/companies like Berkshire (concentrated long positions in the US, and he buys companies because he likes them), but this does not fit RT's profile at all. - When people talk about RT's stellar performance, they're mostly referring to the Medallion fund. Some performance statistics can be found here. Medallion is closed and only has employee money at this point (I could be wrong). RT's other funds that are offered to external clients are nowhere near as good, based on what I've read.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.