Dynamic Portfolio Optimization Through Periodic Rebalancing
Summary
The document explains how portfolio theory can be adapted when holdings and weights change over time. It describes dynamic optimization as repeatedly estimating allocations at scheduled rebalancing dates, using data available up to each date, rather than treating one allocation as permanent. Rolling historical windows provide inputs for each new estimate.
The answer says investors commonly rebalance monthly, quarterly, or semi-annually, and identifies volatility as a more forecastable input than expected returns for adjusting weights. It gives no empirical test or performance evidence, and its discussion is a high-level description rather than a specific optimization procedure. The suggested cadence is general guidance; the document does not address transaction costs, intraperiod trading, or how to choose a rebalancing schedule for a particular portfolio.
Key ideas
- Dynamic portfolio optimization re-estimates allocations at each rebalancing date.
- Rolling historical data can inform each new allocation estimate.
- The answer describes monthly, quarterly, and semi-annual rebalancing as common horizons.
- Volatility is presented as easier to forecast than expected returns and therefore a common input for reweighting.
Tags
Full text
# How can we quantify time varying portfolios? # How can we quantify time varying portfolios? In portfolio management, it is assumed that the assets and the weights in the portfolio are static and do not change in time. By the help of this static structure of the portfolio, we can talk about the standard deviations and returns of the portfolio in order to quantify risk and return. Am I correct? If so, I have a question. What if the portfolio is varying in time, how can we apply portfolio theory? Say, new assets are allocated, or sold in every minute. Thanks ## Answer by develarist (score 1, accepted) https://quant.stackexchange.com/a/53499 The mean-variance model in portfolio theory does suffer from being static. This is why it has been extended to dynamic portfolio optimization, which captures investors' behavior of rebalancing their portfolios periodically. They rebalance because, as time varies, individual assets returns and volatilities change and diverge from their measurements they had during the initial allocation. It is not normal to rebalance every minute though. Investors hold their portfolios for much longer horizons so that they rebalance monthly, quarterly, or semi-annually. After all, assets hardly change over the course of one minute, but can change dramatically by the end of a month's time. Since it is hard to tell what the portfolio will be like in the future, allocations that should be optimal the next quarter, for example, have to be estimated from current and historical data. Upon reaching each rebalancing period, you repeat this process using preceding historical data that has accumulated up to that specific point in time in the form of rolling windows. This is the typical design of forecasting models, where volatility is typically the basis for re-weighting the portfolio, mainly due to volatility being much more easy to forecast than expected returns. In summary, in dynamic portfolio optimization, the static model is just repeated (re-estimated) for each new rebalancing period as just described using new historical data.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.