Receiver Operating Characteristics Slides

Receiver Operating Characteristics Slides - Investment...

Info iconThis preview shows pages 1–8. Sign up to view the full content.

View Full Document Right Arrow Icon
vestment Performance of Investment Performance of Directional Trading Strategies Òscar Jordà and Alan M. Taylor ept of Economics U C Davis Dept. of Economics U.C. Davis ontact Information Contact Information :
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
vestment Performance of Investment Performance of Directional Trading Strategies Òscar Jordà and Alan M. Taylor ept of Economics U C Davis Dept. of Economics U.C. Davis ontact Information Contact Information :
Background image of page 2
bjective Objective y To evaluate investment strategies, specifically directional trading strategies, such as long-short strategies. arry trade is an example: ex ost profits y Carry trade is an example: ex-post profits et y Let y Then the trader decides on the direction of the carry depending on and where is called a classifier. Òscar Jordà and Alan M. Taylor Investment Performance of Directional Trading Strategies 3
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
emarks Remarks y A binary evaluation based on direction only carries a lower burden y However, even a slightly better than 50-50 assifier can quickly generate significant profits classifier can quickly generate significant profits. y E. g. suppose a strategy generates a 55% success f picking the direction of carry. In the period of picking the direction of carry. In the period 1989-2008 the strategy would have delivered 4.06% annual avg. return with a 1.02% standard deviation, a Sharpe ratio of 1.14 and a monthly return skew of 2.20 Òscar Jordà and Alan M. Taylor Investment Performance of Directional Trading Strategies 4
Background image of page 4
pproach Approach y Non-parametric y independent of investor loss function (unlike typical binary evaluation scores, e.g. Brier, uadratic log) quadratic, log) y But where optimal operating point can be tailored investor preferences to investor preferences y Classification can be achieved without having to rely on binary choice models y Methods have well-known asymptotic approximations Òscar Jordà and Alan M. Taylor Investment Performance of Directional Trading Strategies 5
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
The Basic Ingredient: The Receiver Operating Characteristics (ROC) Curve y Origin s of binary classification evaluation perhaps l P i (1884) i l M f as early as Peirce (1884) Numerical Measure of the Success of Predictions and his science of the method y ROC origin: Peterson and Birdsall (1953) in radar signal detection OC applications: y ROC applications: y Biostatistics/Medicine: Lusted (1960) y Psychometrics: Swets and Pickett (1982) y Atmospheric Sciences: Mason (1982) y Machine Learning: Spackman (1989) Òscar Jordà and Alan M. Taylor Investment Performance of Directional Trading Strategies 6
Background image of page 6
The Classification Problem - Definitions y Using the same notation as the introduction, define: y True positive rate: alse positive rate: y False positive rate: y Also define: y
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 8
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 30

Receiver Operating Characteristics Slides - Investment...

This preview shows document pages 1 - 8. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online