Worth a Hill of Means - Takahē Capital Insights

Oct 12, 2026

Takahē Capital® Insight note describes a long-biased, short-term mean reversion trading system applied to S&P 500 e-mini futures, first tracked in August 2014 after publication by Quest Partners LLC. The review explains why many quantitative strategies decay or lose edge after disclosure, then details mean reversion concepts, typical win/loss characteristics, and vulnerabilities such as gaps and over-optimization. It summarizes entry and exit rules and evaluates in-sample performance with and without trading costs, including a variant for US 30Y treasury bond futures.

LanguageEnglish

CategoryPresentations

FormatPDF

Pages9

Reading Time3 min

AccessFree to download

Worth a Hill of Means

What to expect from the strategies in our database

Mean reversion

BTFD

Trading system overview and exit rules

In-sample performance results

Takahē Capital® Insights Worth a Hill of Means By Moritz Seibert [www.takahe.capital](www.takahe.capital)[ ](www.takahe.capital)[contact@takahe.capital](contact@takahe.capital) October 2024 Worth a Hill of Means Takahē Capital maintains a large database of quantitative strategies to support and improve its research process. While some of these strategies were developed in-house, most were sourced from books, academic papers, and the internet. One such example is the trading system featured in this Insight note: a long-biased, short-term mean reversion system trading the S&P 500 e-mini futures contract. We started tracking this system in August 2014 after discovering it in a research note published by Quest Partners LLC. It is now approaching its 10-year live anniversary in our database. This Insight note reviews its performance and highlights some interesting results and conclusions. What to expect from the strategies in our database Most of the strategies included in our database 1 are either ill-designed or have experienced substantial performance decay over time. Once the rules of a trading system are made public, its excess returns tend to decline in response to broader adoption.2 While many academic papers present systems that lack realistic trading assumptions and therefore look too good on paper, our default assumption is that the strategies shared by active traders are mediocre at best and useless at worst. That’s because professional traders are motivated to protect their IP and unlikely to compromise their competitive advantage by revealing the details of their strategies. Despite these limitations, we are convinced it’s valuable to follow a broad spectrum of these external strategies and shadow-run them in parallel to our own. If nothing else, our database creates a benchmark and a frame of reference from which we can learn. Mean reversion Opposite to trend following strategies, mean reversion systems adopt a convergent approach by exploiting price oscillation around the mean of the return distribution rather than price excursions into the tails. Generically, these systems are designed to buy dips and/or sell rallies – a method that tends to produce a large number of small winning tradesand fewer, but larger, losing trades. Kahnemanand Tversky have shown that this positive win vs. loss relationship is emotionally more palatable for most investors, whereas trend following systems can induce emotional pain because losing trades tend to outnumber the winning trades, resulting in drawdowns and a postponed sense of success. However, mean reversion systems are prone to suﬀer to a greater extent from price discontinuities, also known as gaps, producing a negatively skewed trade and return distribution. And they can very easily be over-optimized. BTFD The Quest note focuses on the illustration of irregular returns in US equity and bond markets. Quoting from their article: “By illustrating the availability of skill-less and unstable mean reversion based sources of alpha, we hope investors will exercise sincere self-analysis of 1 At the top-level, our database categorizes strategies as follows: trend, countertrend, carry, mean reversion, spreads, patterns, and setups. 2 Zhou, Lin (2017): The alpha life cycle of a quantitative strategy. Falck, Rej, Thesmar (2022): When do systematic strategies decay. McLean, Pontiﬀ (2015): Does academic research destroy stock return predictability. their methods of investment selection and also of the fees that they are willing to pay their managers.” And: “We do not trade such models, nor do we recommend for anyone to trade them as we believe it is only a question of time before substantial losses are experienced in these or similar strategies.” We couldn’t agree more. The system presented in their article is a longsided mean reversion system whose trading parameters were intentionally optimized for the S&P 500 e-mini and US 30Y treasury bond fu

FAQs

How are the system’s exits defined?

The system has three exit rules and closes the long position on whichever event occurs first: a time exit on the 4th business day since entry, a stop loss exit at entry minus 2 ATRs, and a profit target exit at entry plus 0.5 ATRs.

Oct 12, 2026

Upload by WPS_1790064749