Consistent and Efficient Dynamic Portfolio Replication with Many Factors
Bibliographic record
Abstract
Factor investing involves choosing securities to construct portfolios with particular risk–return profiles. With the proliferation of benchmark-tracking exchange-traded funds (ETFs) virtually any risk–return profile can be reconstructed; the challenge is to find the right ETFs because the number of relevant ETFs is very large. This article proposes an innovative modification to the resampling procedure used in many popular machine learning methods for reducing the dimensionality of this problem. The proposed method allows selection of the specific ETFs used to replicate returns, taking the total costs of using the optimal portfolio to dynamically track returns into consideration. Existing variable selection algorithms are not designed to incorporate rebalancing costs, which are accumulated over time. The methodology is illustrated by replicating hedge fund returns with ETFs. The results show that, by selecting the right replication instruments in a way that is consistent with an investor’s economic utility instead of using purely statistical losses, the investor can save around 60 bps per year. TOPICS:Exchange-traded funds and applications, statistical methods, simulations, big data/machine learning Key Findings • A modified LASSO approach is developed for replication when variables are selected from many potential factors and transaction costs are accounted for in a dynamically consistent way. • By accounting for investor’s economic utility instead of purely statistical losses, the improved portfolio optimization procedure saves investors around 60 bps per year out of sample. • The new cross validation procedure is applicable for a wide range of problems in a time series context, when overfitting and transaction costs are major concerns of the model user.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.010 | 0.032 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".