Skip to content
All library documents

ANTICOR Portfolio Rebalancing: Returns, Costs, and Assumptions

Article Quant Q&A · Author: Joshua Ulrich
Read the original

Summary

The document discusses the ANTICOR portfolio algorithm, introduced in a paper on learning to outperform the best stock. It characterizes the method as a windowed form of constant rebalanced portfolio allocation and notes that its appeal rests on strong risk-adjusted returns, if those can be achieved in practice.

The central caveat is that the idealized analysis assumes zero transaction costs and tax-free returns. Trading costs can erode performance, and the method may be sensitive to the window size used to estimate relationships. The document points to an empirical thesis and related work on constant rebalanced portfolios, but provides no results or detailed comparisons of its own. These comments therefore identify assumptions and possible drawbacks rather than establishing how ANTICOR performs across markets or implementations.

Key ideas

  • ANTICOR is described as a windowed version of constant rebalanced portfolio allocation.
  • Its risk-adjusted returns are presented as a potential advantage, conditional on achieving them in practice.
  • The analysis assumes zero transaction costs and tax-free returns.
  • Transaction costs and sensitivity to the window size are potential practical limitations.

Tags

Portfolio constructionBacktestingEquitiesRisk management
Full text
# What are the advantages / disadvantages of the ANTICOR algorithm?


# What are the advantages / disadvantages of the ANTICOR algorithm?












The algorithm is introduced in the paper, Can We Learn to Beat the Best Stock.

The obvious advantage is superior risk-adjusted returns (if you can actually achieve them). Transaction costs and sensitivity to window size seem to be potential disadvantages.

I'm hoping someone can provide more insightful advantages / disadvantages.

## Answer by Tangurena (score 4, accepted)

https://quant.stackexchange.com/a/1083

One of the major assumptions is that you have zero transaction costs. Another one is that your returns are tax-free. Otherwise it looks to me to be a windowed version of CBAL (constant rebalanced).

A more technical analysis can be found at: Castonguay, Portfolio Management: An empirical study of the Anticor algorithm (An MS thesis)

Covan and Gluss, Empirical Bayes Stock Market Portfolios (CBAL)

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.