Transaction Cost Modeling

Two of the most important goals of asset managers are (1) generating outperformance, and (2) implementing portfolios in the best way possible. At Robeco quantitative strategies, we strive to achieve outperformance for our clients by gaining exposure to factor premiums.

Factor investing entails allocating to factors shown to have a premium. Academic research shows that factor premiums have better risk/return profiles than market-capitalization weighted indices. Several well-known factor premiums for equities are, e.g, the low-volatility premium (Blitz and van Vliet, 2007), the size premium (Banz, 1981), the value premium (Fama and French, 1992) and the momentum premium (Jegadeesh and Titman, 1993). Factor premiums are not only present within equities, but also within other asset classes. For instance, the low volatility effect is also present in credits (see, e.g., Houweling et al., 2012) while value and momentum also exists in bonds and commodities (see, e.g., Asness, Moskowitz and Pedersen, 2009).

However (2) implementing portfolios in the best way possible is just as important. Translating a factor investment strategy from theory to practice is not easy. Strategies that look profitable on paper turn out to be losing after trading costs. Classical momentum factor portfolios exhibit high turnover and take considerable positions in small-cap stocks, characteristics that make it difficult to adopt such a strategy in reality . The problems are visible throughout portfolio implementation: not all stocks are liquid enough to absorb a large position, some stocks have higher transaction costs than others, and some strategies generate a lot of turnover on the portfolio level.

At Robeco we take transaction costs into account in the quant investment process. Transaction costs are usually measured by the distance between the execution price and the price at the moment you decided to make the investment decision, which is also known as the ‘implementation shortfall’. We already have a proprietary transaction costs model, but we are convinced that with our expanded dataset of Robeco’s executed transactions, this model can be further improved. We have tens of thousands of trades that we executed over the past years to use in the estimation process.

At this stage there are a lot of open questions. Examples of research questions:

Curious to know what kind of projects we offer?
Available projects
  • How do we get the best estimate of trading costs, before executing? What variables to use and in what kind of model?
  • What is the best trading costs estimate after we executed the trade? The stock price will have moved due to our trade execution, but also because of general market movement and other market participants trading. If we can disentangle our own market impact from other factors we can use smaller samples to accurately calibrate the model.
  • Once we have the best cost estimate, how do we use the new insights from this project to trade in smarter ways?

This Super-Quant internship will be in Quant Research in close collaboration with the Trading Desk at Robeco, with over 35bln in quant assets and over 50bln dollar traded each year in equities alone your research will have impact. You will be responsible for analyzing a proprietary  dataset of historical transactions, improving the cost model that is used in various live strategies and thinking critically about trade execution. In the meanwhile you will get the chance to learn from researchers and traders alike, and you will experience first-hand the world of quantitative asset management, in which theory and practice collide.

Due to the technical nature of the project and the wide variety of techniques that can be used to model transaction costs, candidates from a wide variety of masters are welcome (Econometrics/Math/Physics/Data science etc.). However enthusiasm and affinity with Finance and Trading is required.


R. Almgren and N. Chriss. Optimal execution of portfolio transactions. Journal of Risk, 3:5–40, 2001

Zarinelli, Elia, et al. "Beyond the square root: Evidence for logarithmic dependence of market impact on size and participation rate." Market Microstructure and Liquidity 1.02 (2015): 1550004.

Perold, Andre F. "The implementation shortfall: Paper versus reality." The Journal of Portfolio Management 14.3 (1988): 4-9

Almgren, Robert, et al. "Direct estimation of equity market impact." Risk 18.7 (2005): 5862.

Wilma de Groot, Joop Huij, and Weili Zhou, “Another Look at Trading Costs and Short-Term Reversal Profits”, Journal of Banking and Finance, January 2012