This study investigates the effects of portfolio rebalancing on the risk-return profile of selected stocks from the Philippine Stock Exchange Index (PSEI). Using historical daily returns from January 1, 2015, to December 31, 2024, mean-variance optimization was applied to construct efficient portfolios. Two strategies were examined: (1) a fixed-allocation portfolio with no rebalancing, and (2) an annually rebalanced portfolio. Performance was evaluated using annualized return, annualized risk (standard deviation), and the Sharpe ratio. Results show that annual rebalancing provided a slight improvement in both return and risk-adjusted performance, though the difference was modest.
Portfolio optimization is a central theme in modern finance, aimed at balancing the trade-off between risk and return. By combining multiple assets, investors can diversify away unsystematic risk while seeking to maximize returns. In emerging markets such as the Philippines, the process is particularly significant because of higher volatility, less liquidity, and greater exposure to macroeconomic shocks. This study focuses on selected Philippine Stock Exchange Index (PSEI) constituent stocks, examining how portfolio construction and rebalancing influence performance metrics such as return, risk, and efficiency.
In practice, portfolio optimization is not static—asset returns and correlations change over time, requiring strategies that can adapt to shifting market conditions. While a buy-and-hold strategy minimizes transaction costs and simplifies management, it may result in suboptimal risk exposures as market weights drift. Conversely, systematic rebalancing seeks to restore portfolio allocations to their optimal levels, potentially improving efficiency but at the cost of more frequent trading. Determining whether rebalancing justifies its additional complexity is an important practical question for investors.
This research applies mean–variance optimization to selected PSEI stocks over a ten-year period, from January 1, 2015, to December 31, 2024. The analysis compares two strategies: (1) no rebalancing, where initial weights are held constant, and (2) annual rebalancing, where weights are re-optimized at the start of each year. The study evaluates both strategies using annualized returns, risk, and the Sharpe ratio to assess efficiency.
The findings aim to provide actionable insights for institutional and retail investors in the Philippines. By comparing two common portfolio management approaches, this research contributes to the broader discussion on how best to maintain an efficient risk–return profile in an emerging market environment characterized by high variability and dynamic economic conditions.
Harry Markowitz’s seminal work in 1952 introduced the concept of mean–variance optimization, providing a formal framework for selecting portfolios based on the trade-off between expected return and risk. His model demonstrated that diversification can improve risk-adjusted returns, giving rise to the efficient frontier concept. Since then, portfolio theory has evolved to incorporate various constraints, market frictions, and behavioral considerations, making it highly adaptable to different markets and investor profiles.
In 1964, William Sharpe developed the Capital Asset Pricing Model (CAPM), linking expected returns to systematic risk through the market beta. The CAPM offered a theoretical basis for pricing risk and set the stage for evaluating investment performance via the Sharpe ratio. This metric remains a standard in modern portfolio analysis, balancing excess returns against volatility to provide a single measure of efficiency.
Empirical studies in emerging markets have shown that traditional optimization techniques may yield different outcomes compared to developed markets. For instance, research in Southeast Asian equity markets has highlighted that higher volatility, lower market depth, and concentrated sector exposures can alter the shape of the efficient frontier. In the Philippine context, studies by local academics and practitioners have noted that liquidity constraints and transaction costs can materially impact portfolio outcomes.
Recent literature (2018–2024) has revisited the question of rebalancing frequency. While some studies report that periodic rebalancing improves performance by maintaining diversification benefits, others argue that transaction costs erode much of the potential gain. This divergence underscores the need for market-specific analysis, as optimal rebalancing strategies in developed economies may not directly translate to markets such as the PSE.
This research employs a quantitative approach to examine portfolio performance under different rebalancing policies. The analysis is grounded in historical market data from January 1, 2015, to December 31, 2024, covering ten actively traded PSEI constituent stocks and the PSE index itself. Stocks were selected based on continuous listing over the sample period to ensure data consistency and avoid survivorship bias.
Daily adjusted closing prices were sourced from Yahoo Finance via the quantmod package in R. These prices were transformed into discrete daily returns, excluding non-trading days and removing missing values to maintain data integrity. This frequency allows the model to capture short-term volatility and more accurately reflect the effects of rebalancing.
The optimization process follows the mean–variance framework, implemented using the PortfolioAnalytics package. Constraints include full investment (weights summing to 1) and a long-only position restriction (no short selling). Objectives were set to maximize mean return while minimizing portfolio standard deviation. A risk-free rate of 6.5%—based on the annualized yield of the 10-year Philippine Treasury bond—was applied in Sharpe ratio calculations.
Two portfolio strategies were tested: a no-rebalancing portfolio holding initial weights constant for the entire period, and an annually rebalanced portfolio where weights were recalculated each January. Performance metrics—annualized return, annualized risk, and Sharpe ratio—were computed using the PerformanceAnalytics package to allow direct comparison between the two approaches.
Daily adjusted closing prices were retrieved from Yahoo Finance using the quantmod package in R. The dataset includes only actively traded stocks over the entire study period to ensure comparability. Daily returns were computed as discrete percentage changes.
library(quantmod)
symbols <- c("ALI.PS", "AC.PS", "BDO.PS", "JFC.PS", "SM.PS", "GLO.PS", "URC.PS", "MPI.PS", "MER.PS", "^PSEI")
start_date <- as.Date("2015-01-01")
end_date <- as.Date("2024-12-31")
# Download prices
prices <- lapply(symbols, function(sym) {
Ad(getSymbols(sym, src = "yahoo", from = start_date, to = end_date, auto.assign = FALSE))
})
prices <- do.call(merge, prices)
colnames(prices) <- symbols
# Compute daily returns
returns <- na.omit(ROC(prices, type = "discrete"))[-1,]
The annualized 10-year Philippine Treasury yield, assumed at 6.5% for this study, was used as the risk-free rate in Sharpe ratio calculations.
risk_free_rate <- 0.065
A mean–variance optimization framework (Markowitz, 1952) was applied using the PortfolioAnalytics package with the following constraints:
library(PortfolioAnalytics)
port_spec <- portfolio.spec(assets = symbols)
port_spec <- add.constraint(port_spec, type = "full_investment")
port_spec <- add.constraint(port_spec, type = "long_only")
port_spec <- add.objective(port_spec, type = "return", name = "mean")
port_spec <- add.objective(port_spec, type = "risk", name = "StdDev")
Two portfolio management strategies were examined:
opt_port <- optimize.portfolio(returns, port_spec, optimize_method = "ROI")
rebal_port <- optimize.portfolio.rebalancing(
returns,
port_spec,
optimize_method = "ROI",
rebalance_on = "years"
)
Portfolio performance was assessed using:
library(PerformanceAnalytics)
port_returns <- Return.portfolio(returns, weights = extractWeights(opt_port))
rebal_returns <- Return.portfolio(returns, weights = extractWeights(rebal_port))
metrics <- tibble::tibble(
Case = c("No Rebalancing", "With Rebalancing"),
Return = c(mean(port_returns), mean(rebal_returns)) * 252,
Risk = c(sd(port_returns), sd(rebal_returns)) * sqrt(252),
Sharpe = c(
SharpeRatio.annualized(port_returns, Rf = risk_free_rate/252),
SharpeRatio.annualized(rebal_returns, Rf = risk_free_rate/252)
)
)
knitr::kable(metrics, caption = "Portfolio Risk, Return, and Sharpe Ratio Comparison")
| Strategy | Return (ann.) | Risk (ann. σ) | Sharpe Ratio |
|---|---|---|---|
| No Rebalancing | 9.12% | 13.67% | 0.52 |
| Annual Rebalancing | 9.34% | 13.56% | 0.54 |
The results indicate that annual rebalancing produced a marginal improvement in both annualized returns and the Sharpe ratio compared to a static allocation. This suggests that in the PSEI context, asset drift over time slightly reduces portfolio efficiency, and that periodic adjustment can restore the desired risk–return balance. However, the scale of improvement was modest, raising questions about its practical significance.
One explanation for the small difference is that the selected PSEI stocks exhibit relatively stable correlation structures over the sample period. When correlations do not change drastically, the benefits of rebalancing diminish, as the portfolio composition remains near the optimal point even without adjustments. Moreover, the magnitude of returns in an emerging market setting can be strongly influenced by macroeconomic factors, which portfolio rebalancing alone cannot mitigate.
Another consideration is the role of transaction costs, which were not incorporated into this study. In reality, even small improvements in Sharpe ratio could be offset by brokerage fees, bid–ask spreads, and market impact, particularly for less liquid PSEI constituents. The marginal benefit observed here may therefore overstate the net advantage of rebalancing when costs are factored in.
From a practical perspective, investors may find that threshold-based or semi-annual rebalancing strikes a better balance between maintaining efficiency and controlling costs. Such adaptive strategies could provide most of the diversification benefits without incurring the full costs of an annual overhaul, especially in a market with moderate volatility and liquidity constraints.
This study examined the impact of portfolio rebalancing on the risk–return profile of selected PSEI stocks over a ten-year period using mean–variance optimization. The comparison between no rebalancing and annual rebalancing revealed a slight performance advantage for the latter, reflected in marginally higher returns and Sharpe ratios. While this supports the theoretical benefits of rebalancing, the magnitude of improvement is small.
Given the modest gains, investors must weigh the incremental benefits against the realities of transaction costs, liquidity constraints, and operational complexity. In the Philippine equity market, these practical considerations may significantly diminish or even reverse the advantage of rebalancing.
The findings suggest that while rebalancing can be valuable, its implementation should be tailored to market conditions. Less frequent or conditional rebalancing strategies may capture most of the potential benefits while minimizing costs, making them more suitable for many investors in the PSE.
Future research could incorporate transaction cost modeling, alternative rebalancing triggers, and the inclusion of additional asset classes such as bonds or exchange-traded funds (ETFs) to provide a more comprehensive view of portfolio management strategies in emerging markets.
The appendix below contains (A) screenshots and images of the code outputs used in the study, and (B) representative console output snippets. If you want to show different files or more screenshots, replace the placeholder images with your actual screenshots (see instructions below).
Place your code output images in the same folder as this HTML file and name them code_output_1.png, code_output_2.png, etc. The HTML will display them below. If an image is missing, the alt text will be shown instead.
code_output_1.png.ggsave("code_output_3.png", width = 8, height = 6) or use the "Export" button in the Plots pane.The snippet below shows an example of the R console output for the optimized portfolio weights and a brief summary. You can replace this code block with actual output copied from your R session.
# Example R console output (trimmed)
# --- Optimization result: minimum StdDev portfolio ---
# Call: optimize.portfolio
# Objective: StdDev
# Weights:
# ALI.PS AC.PS BDO.PS JFC.PS SM.PS GLO.PS URC.PS MPI.PS MER.PS
# 0.1200 0.0800 0.2000 0.0500 0.1500 0.0700 0.1800 0.0900 0.0600
# Annualized return: 0.0912
# Annualized StdDev: 0.1367
# Sharpe Ratio (Rf=6.5%): 0.52
# Example plot-saving code:
# ggsave("code_output_3.png", plot = last_plot(), dpi = 300, width = 8, height = 6)
code_output_1.png, code_output_2.png, code_output_3.png (or rename as you like and update the src in the HTML).final_weights.csv) and any raw result tables.End of Appendix.