ANY VIEWS EXPRESSED HERE ARE THOSE OF THE AUTHOR AS OF THE DATE OF PUBLICATION, ARE BASED ON AVAILABLE INFORMATION, AND ARE SUBJECT TO CHANGE WITHOUT NOTICE. INDIVIDUAL PORTFOLIO MANAGEMENT TEAMS MAY HOLD DIFFERENT VIEWS AND MAY MAKE DIFFERENT INVESTMENT DECISIONS FOR DIFFERENT CLIENTS.
BEFORE INVESTING, INVESTORS SHOULD CONSIDER THE RISKS THAT MAY IMPACT THEIR CAPITAL. THE VALUE OF YOUR INVESTMENT MAY BECOME WORTH MORE OR LESS THAN AT THE TIME OF THE ORIGINAL INVESTMENT. PLEASE REFER TO THE RISK SECTION IN THE PDF AVAILABLE FOR DOWNLOAD BELOW.
- Our “contextual” alpha model differentiates stocks based on issuer characteristics in an attempt to identify which inefficiencies will be most important for specific securities.
- We use a variety of bottom-up stock-selection models and opportunistic or event-driven models, reflecting our belief that a strong quant process requires multiple independent sources of alpha and diverse time frames.
- Our proprietary risk model is tailored to the factors used in the alpha model and this alignment can be important in the trade-off of sources of risk and return. We think it helps us better distinguish intentional risks from unintentional risks, and be more efficient and targeted in our risk budgeting.
- By combining cost estimates from our transaction-cost model with stock-specific alpha forecasts, we seek to improve stock selection by avoiding seemingly attractive ideas where the execution costs would significantly reduce or eliminate the excess-return potential.
We believe that successful quantitative investing depends on a fully integrated process, which includes alpha generation, risk modeling, and portfolio construction. Weakness in any of these areas can compromise otherwise well-conceived investment ideas. In this paper, we provide an overview of three proprietary tools we’ve developed to help ensure that our quant process stands up to the test — our alpha, risk, and transaction cost models.
Q: What’s different about our alpha model?
Our research has shown that certain factors have historically been closely associated with stock outperformance. Further, the alpha of these factors has been empirically strong and pervasive across time and geography, and can be traced to a combination of behavioral inefficiencies, market structure, and risk premiums. We believe quantitative investing techniques are ideal for detecting and systematically exploiting these factors.
However, we have found that many of the inefficiencies that quants exploit do not work equally well for all stocks. With this in mind, we designed a “contextual” model that differentiates stocks based on issuer characteristics in an attempt to identify which inefficiencies will be most important for specific securities. This gives us a different perspective from many quant managers. In addition, we’ve built our models to tap a diverse range of alpha sources. These differences in our models can help us pursue…
To read more, please click the download button below.