Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.36491

Simple Portfolios that Deliver Consistent Decent Returns

Oskar Edén Wallberg·Alexander Karolin
View original source ↗Read source document ↗Export metadata ↓

Abstract

Portfolio construction has been studied extensively in both academia and industry. Its applications range from pension funds and hedge funds to private savings. The academic literature centers on methodology, commonly evaluated in settings that allow short selling, leverage, and frequent trading. Industry, on the other hand, operates under real constraints, but its data and methods are proprietary and largely unpublished. Neither is directly available to the private investor. This paper is an empirical study of established portfolio construction methods in the setting of the private investor. We consider a small investable universe of liquid US ETFs, where trading is monthly and carries a cost, and neither shorting nor leverage is permitted. Portfolio construction is posed as a convex optimization problem, its inputs are estimated from public data, and every parameter is interpretable. We denote such portfolios simple. We show that simple portfolios of as few as three ETFs deliver consistent and decent returns, without tuning to the backtest period. To find them, we have developed a framework, which we share publicly, able to backtest hundreds of portfolios on a standard laptop over a twenty year period or longer. When no return forecast is made, portfolios are constructed from risk alone, using methods such as risk parity and minimum volatility, built on a statistical factor model fitted with expectation maximization. When a forecast is made, we combine the Markowitz formulation with momentum signals and ridge regression. We report what works and what each component adds, weighing its gain against its complexity and the effort it demands of the investor. The findings are supported by numerical experiments over the past two decades, and the proposed portfolios outperform the standard 60/40 allocation.

Explore related subjects

q-fin.PM

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Oskar Edén Wallberg, Alexander Karolin. 2026-10-27. Simple Portfolios that Deliver Consistent Decent Returns. https://arxiv.org/abs/2610.36491

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Bayesian Parametric Portfolio Policies

Parametric portfolio policies map predictors into portfolio weights without modelling returns, making them a natural tool for factor timing. With many predictors, if the policy has more coefficients than observations the data cannot pin it down. We develop Bayesian Parametric Portfolio Policies (BPPP), which discipline the policy with a prior, update it with realised utility, and average the weights over the posterior. We contribute to the factor-timing literature by estimating timing as a portfolio rule rather than from return forecasts. With 242 predictors and five factors, BPPP earns 2.08 percentage points a year from timing, mostly from aggregate anomaly returns. Under a leverage limit, averaging sizes each position by its uncertainty and comes close, without a separate scale choice, to the size that maximises timing with hindsight. A turnover penalty preserves significant gross timing at lower turnover.

q-fin.PM↗
arXiv2026

Market Regime-Based Multi-Asset Portfolio Optimization Using Nested Clustered Optimization for Tactical Asset Allocation

We propose a regime-aware tactical asset allocation (TAA) framework integrating Gaussian Hidden Markov Model (HMM) regime detection with Nested Clustered Optimization (NCO) for multi-asset portfolio construction. Market regimes (Bullish, Sideways, Bearish) are identified strictly out of sample to prevent look-ahead bias, parameterizing regime-conditional covariance estimation and hierarchical clustering within NCO. The framework is evaluated on a monthly Indonesian multi-asset universe comprising government bonds (SBN), LQ45 equities, gold, CNY/IDR foreign exchange, and Bitcoin over a 56-month out-of-sample period (2021-2025). Benchmarks include equal-weight (1/N), buy-and-hold, minimum variance, hierarchical risk parity (HRP), and standard regime-agnostic NCO. Statistical significance is evaluated using the Jobson-Korkie test. The empirical results demonstrate that Regime-NCO effectively curbs tail risk, achieving a maximum drawdown of 4.44% compared with 15.27% for equal-weight and 21.17% for buy-and-hold. However, regime-agnostic NCO achieves a still lower drawdown of 2.19% and higher cumulative return (29.57% vs 5.44%). Detailed analysis reveals that a rolling 24-month HMM triggers regime transitions in 55 of 56 out-of-sample months, inducing regime drift that erodes the theoretical gains of conditional covariance estimation. These findings provide candid empirical evidence on rolling HMM behavior and highlight the necessity of transition-smoothing mechanisms for practical regime-aware portfolio optimization.

q-fin.PM↗
arXiv2026

Certifying Reward Alignment in Portfolio Reinforcement Learning

Financial reinforcement learning often translates an investment objective into an additive risk-adjusted reward. An agent can optimise that reward accurately yet implement a different financial preference. We develop a decision audit for terminal mean-variance equilibrium that separates reward misspecification from learning error. In a finite exogenous-state market, we bound the gain from changing one action using actor, reward and continuation-critic residuals. The certificate ignores state-constant critic offsets, accommodates convex action constraints, and has an independent-sample extension that does not require the equilibrium policy. A continuation-aware actor-critic method estimates the future-policy correction from samples. Across nine controlled markets and twenty seeds, it reduces policy distance by and one-step deviation gain by relative to a matched myopic-reward network. A separate PPO comparison tests the same reward intervention within a standard optimiser. Historical tests cross seven covariance estimators and 10, 30 and 49 US industry portfolios over 432 months: joint reward calibration preserves each fitted mean-variance target, but does not improve realised investment performance. Together, these results provide a way to quantify whether a learned decision implements its stated financial objective, separately from how well that objective performs in a backtest.

q-fin.PM↗