Paper: arXiv 2308.06882
Authors: Dhruv Desai, Ashmita Dhiman, Tushar Sharma, Deepika Sharma, Dhagash Mehta, Stefano Pasquali
Abstract
Mutual fund categorization has become a standard tool for the investment management industry and is extensively used by allocators for portfolio construction and manager selection, as well as by fund managers for peer analysis and competitive positioning. As a result, a (unintended) miscategorization or lack of precision can significantly impact allocation decisions and investment fund managers. Here, we aim to quantify the effect of miscategorization of funds utilizing a machine learning based approach. We formulate the problem of miscategorization of funds as a distance-based outlier detection problem, where the outliers are the data-points that are far from the rest of the data-points in the given feature space. We implement and employ a Random Forest (RF) based method of distance metric learning, and compute the so-called class-wise outlier measures for each data-point to identify outliers in the data. We test our implementation on various publicly available data sets, and then apply it to mutual fund data. We show that there is a strong relationship between the outlier measures of the funds and their future returns and discuss the implications of our findings.
Complexity vs Empirical Score
- Math Complexity: 5.5/10
- Empirical Rigor: 7.5/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper employs advanced supervised distance metric learning with Random Forests (ML), but its rigor is validated by testing on public datasets and mutual fund data, showing an inverse correlation with future returns.
Research Flowchart
flowchart TD
A["Research Goal<br>Quantify miscategorization effects"] --> B["Data<br>Mutual Fund Datasets"]
B --> C["Methodology<br>RF-based Distance Metric Learning"]
C --> D["Compute<br>Class-wise Outlier Measures"]
D --> E{"Analysis<br>Test on Public Datasets & Apply to Mutual Funds"}
E --> F["Key Findings<br>Outlier measures predict future returns"]
F --> G["Implications<br>Improved peer analysis & allocation decisions"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a supervised distance‑metric learning method using Random Forest proximities to compute a continuous outlierness measure for funds relative to their assigned categories.
Source sentence
We pose fund miscategorization as a distance metric based outlier detection problem, and in turn device a continuous measure of outlierness of the funds with respect to their assigned category (as well as all other categories);
Assumes 3
- Random Forest proximity matrix can be interpreted as a meaningful similarity measure between funds. model
Source sentence
For the proximities among all pairs of data-points in the dataset generated from RF, the distance matrix is defined as 1−Proximity, where the Proximity is an 𝑛×𝑛 matrix whose entries are one of the pair‑wise proximity measure as defined above
- Morningstar category labels accurately reflect the true portfolio composition of funds. data
Source sentence
based on funds’ portfolio composition, we chose features from the feed which would describe the same.
- The outlier measure is inversely related to future fund returns. statistical
Source sentence
We then employ the outlier measure on the fund data and show that there is inverse correlation between the fund’s outlier measure and its returns.
Data
- UCI Toy Datasets (Iris, MNIST, Wine, Breast Cancer, Car Evaluation) benchmark public
Source sentence
To benchmark our implementation, we test it on various toy datasets for classification task from the University of California Irvine (UCI) repository[3]: Iris flower data, MNIST digits, Wine, Breast cancer and car-evaluation.
- Morningstar Fund Data fundamentals commercial
Source sentence
We sourced our data for funds from Morningstar Data warehouse data feed.
Costs
Not applicable
Finds 2
- Funds identified as outliers tend to have lower future returns, demonstrating an inverse relationship between outlierness and performance. performance
Source sentence
We then employ the outlier measure on the fund data and show that there is inverse correlation between the fund’s outlier measure and its returns.
- Outlier measures computed with respect to non‑assigned categories can reveal novel funds that are outliers across all categories. statistical
Source sentence
Using this approach we can identify novelties in the dataset, i.e., a data point which is an outlier to all other classes.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 7 related papers). Coverage is partial, so this is neither replication nor refutation.
3 papers in the archive
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.