Paper: arXiv 2411.07949
Authors: Chutian Ma, Paul Smith
Abstract
We consider a simplified model for optimizing a single-asset portfolio in the presence of transaction costs given a signal with a certain autocorrelation and cross-correlation structure. In our setup, the portfolio manager is given two one-parameter controls to influence the construction of the portfolio. The first is a linear filtering parameter that may increase or decrease the level of autocorrelation in the signal. The second is a numerical threshold that determines a symmetric “no-trade” zone. Portfolio positions are constrained to a single unit long or a single unit short. These constraints allow us to focus on the interplay between the signal filtering mechanism and the hysteresis introduced by the “no-trade” zone. We then formulate an optimization problem where we aim to minimize the frequency of trades subject to a fixed return level of the portfolio. We show that maintaining a no-trade zone while removing autocorrelation entirely from the signal yields a locally optimal solution. For any given “no-trade” zone threshold, this locally optimal solution also achieves the maximum attainable return level, and we derive a quantitative lower bound for the amount of improvement in terms of the given threshold and the amount of autocorrelation removed.
Complexity vs Empirical Score
- Math Complexity: 7.5/10
- Empirical Rigor: 2.5/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper employs advanced stochastic calculus, AR(1) process modeling, and optimization theorems, indicating high mathematical complexity. However, the empirical evidence is purely theoretical, lacking backtests, real data, or implementation details, placing it in the Lab Rats quadrant.
Research Flowchart
flowchart TD
A["Research Goal: Minimize trades for fixed portfolio return<br>in single-asset strategy with transaction costs"] --> B["Model Inputs: Signal w/ Autocorrelation & Cross-Correlation"]
B --> C["Methodology: Optimize 2-parameter controls"]
C --> D{"Strategy Controls"}
D --> E["Linear Filter: Adjusts Autocorrelation"]
D --> F["Symmetric No-Trade Zone: Hysteresis"]
E --> G["Optimization Objective: Minimize Trades"]
F --> G
G --> H["Findings: Remove Autocorrelation + Maintain No-Trade Zone = Locally Optimal Solution"]
H --> I["Outcome: Maximizes Return & Quantifies Improvement Bound"]
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 simplified single‑asset portfolio model with two one‑parameter controls—a linear filter to adjust signal autocorrelation and a symmetric no‑trade threshold—optimizing to minimize trade frequency for a fixed return level.
Source sentence
We consider a simplified model for optimizing a single-asset portfolio in the presence of transaction costs given a signal with a certain autocorrelation and cross-correlation structure. In our setup, the portfolio manager is given two one-parameter controls to influence the construction of the portfolio.
Assumes 6
- The signal {X_t} is an i.i.d. Gaussian sequence with mean 0 and variance 1. model
Source sentence
More formally, let {Xt} ∼N(0, 1), t = 0, 1, 2, . . ., denote a sequence of independent identically distributed (iid) Gaussian random variables.
- The noise {ε_t} is an independent i.i.d. Gaussian sequence, independent of {X_t}. model
Source sentence
Let {εt} ∼N(0, 1), t = 0, 1, 2, . . . denote another iid sequence of Gaussian random variables, independent from the first.
- Returns are generated by Y_t = ρ X_t + √(1‑ρ²) ε_t with fixed 0<ρ<1. model
Source sentence
Let 0 < ρ < 1 be a fixed correlation coefficient, and define the sequence {Yt} of random variables via Yt = ρXt + p 1 −ρ2ϵt.
- The filtered signal ˜X_t is obtained by exponential moving average with parameter α, yielding an AR(1) process with autocorrelation α. model
Source sentence
Given 0 ≤α < 1, define Xt EMA −−−→˜Xt = p 1 −α2(Xt + αXt−1 + α2Xt−2 + α3Xt−3 + ...) (3) Observe that ˜Xt−1 and Xt are independent and that, by construction, ˜Xt is an AR(1) process with autocorrelation α.
Show 2 more
- Portfolio positions are limited to a single unit long or short. execution
Source sentence
Portfolio positions are constrained to a single unit long or a single unit short.
- Transaction costs are captured solely by trading frequency; no explicit cost model is assumed. execution
Source sentence
In this paper, we do not assume a specific model for transaction costs. Instead, we measure the negative effects of transaction costs by trading frequency.
Data
No dataset named in the text.
Costs
Unclear - sources disagree
Source sentence (the authors' words)
In this paper, we do not assume a specific model for transaction costs. Instead, we measure the negative effects of transaction costs by trading frequency.
Finds 2
- Maintaining a no‑trade zone while removing all autocorrelation from the signal (α=0) yields a locally optimal solution that attains the maximum attainable return for any given threshold. theoretical
Source sentence
We show that maintaining a no-trade zone while removing autocorrelation entirely from the signal yields a locally optimal solution. For any given "no-trade" zone threshold, this locally optimal solution also achieves the maximum attainable return level, and we derive a quantitative lower bound for the amount of improvement in terms of the given threshold and the amount of autocorrelation removed.
- For any fixed no‑trade threshold, the locally optimal solution also provides a quantitative lower bound on the improvement in return as a function of the threshold and the removed autocorrelation. theoretical
Source sentence
For any given "no-trade" zone threshold, this locally optimal solution also achieves the maximum attainable return level, and we derive a quantitative lower bound for the amount of improvement in terms of the given threshold and the amount of autocorrelation removed.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 9 related papers). Coverage is partial, so this is neither replication nor refutation.
1 paper 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.
Cited References (22)
- Stochastic optimal control: the discrete-time case (1996)
- The O rnstein– U hlenbeck process as a model of a low pass filtered white noise (2008)
- Trades, quotes and prices: financial markets under the microscope (2018)
- Time series analysis: forecasting and control (2015)
- Multi-period trading via convex optimization (2017)
- Multiperiod consumption and investment behavior with convex transactions costs (1979)
- Portfolio selection with transaction costs (1990)
- Multiperiod portfolio optimization with multiple risky assets and general transaction costs (2016)
and 14 more
- Dynamic trading with predictable returns and transaction costs (2013)
- Forecasting, structural time series models and the K alman filter (1990)
- Portfolio selection with transactions costs (1976)
- Optimal properties of exponentially weighted forecasts (1960)
- Time series: modeling, computation, and inference (2010)
- Moving averages for financial data smoothing (2013)
- Lifetime portfolio selection under uncertainty: The continuous-time case (1969)
- Optimum consumption and portfolio rules in a continuous-time model (1975)
- Operators on inhomogeneous time series (2001)
- An introduction to high-frequency finance (2001)
- Forecasting economic time series (2014)
- Non-linear and non-stationary time series analysis (1988)
- Time series analysis (2020)
- Multiperiod portfolio optimization with multiple risky assets and general transaction costs (2016)
0 of 22 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.