Volatility Forecast Errors and Portfolio Exposure
Year: 2026
Type: preprint
Abstract: A volatility forecast can score worse and still improve a portfolio backtest by shifting capital across assets. I examine this mechanism over 180 months for five asset-class ETFs. In the strongest historical contrast, understating equity risk relative to bond risk reduces the minimum-variance portfolio's average bond weight by 43.92 percentage points. The average shift in holdings accounts for 4.74 of the 5.01 percentage points of annualized gross return difference. In a chronological test, the equity–bond adjustment selected in the first half raises net annualized certainty equivalent by 4.96 percentage points in the second half. A group-level allocation overlay using only past weights gains 5.40 points with roughly half the turnover. Freezing the overlay before the test period retains 87% of the observed gain. Across three allocation rules, the historical improvement can be captured by shifting capital across asset groups.
Author: Yang, Fuli
Language: English
Subfield: Finance
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7225374850