Tactical Asset Allocation

Content provided by Irene Bauer PhD

Balancing Strategy and Adaptability

Adaptability is the cornerstone of success in investment management. Pioneers of dynamic asset allocation, such as Ray Dalio, have often highlighted the value of flexibility in adapting to changing market conditions. Dalio's principles of balancing risk and return through thoughtful rebalancing strategies resonate strongly with the essence of tactical asset allocation. In a world where markets shift rapidly, tactical asset allocation (TAA) stands out as a dynamic strategy to seize opportunities and enhance client portfolios. The challenge is to balance a long-term strategic vision with the adaptability needed to capture short-term opportunities. This article provides a practical look at how tactical asset allocation can fit into wealth management, improving client outcomes.

Understanding Tactical Asset Allocation

Tactical asset allocation can offer meaningful outperformance over strategic asset allocation. Historical studies and practical evidence suggest that wealth managers using TAA may achieve an additional 0.5% to 2% in annualized returns, depending on market conditions and the effectiveness of their strategies. More importantly, TAA can improve risk-adjusted returns, reducing drawdowns and smoothing volatility through responsive adjustments to market conditions.

Listen to the Podcast Series*

Tactical asset allocation connects the steady foundation of strategic asset allocation – designed to meet long-term financial goals and risk preferences– with the dynamic nature of financial markets. By adjusting asset class weightings based on current market conditions and outlooks, wealth managers can add value through active decision-making, without losing focus of the core strategy.

TAA is more than just timing the market, it involves responding to valuation discrepancies, macroeconomic trends, geopolitical risks, and general riskiness in the markets. By carefully navigating between strategic objectives and market opportunities, wealth managers can provide clients with a dynamic and responsive approach to achieving their clients’ financial goals.

Tactical Adjustments

A successful TAA strategy starts with a clear understanding of the client's objectives and risk tolerance. For instance, a wealth manager might conduct a detailed questionnaire to gauge a client's level of comfort with market volatility and their financial aspirations, such as saving for retirement or funding a child's education. This information helps shape the foundation of a portfolio tailored to align with the client's unique needs. Similar to strategic asset allocation, which we have discussed in a previous episode, these factors guide wealth managers in creating a portfolio. In fact, one usually starts with a strategic portfolio as the core allocation. Then we may implement tactical adjustments by going over- and underweight certain assets and potentially extending the asset list with additional themed tilts.

The key to TAA is a disciplined process based on quantitative analysis, rigorous research, and well-defined guidelines. Wealth managers can use tools such as moving averages, sentiment analysis, and macroeconomic data to identify favourable exposures. However, it's crucial to communicate the reasons behind these moves to clients, emphasizing that tactical adjustments are short-term and meant to complement the long-term strategy.

Balancing Opportunity and Risk

The purpose of TAA in a wealth manager's toolkit is to provide an advantage while keeping the portfolio within the same risk profile or, depending on the mandate, up to a certain risk level. Constantly adjusting portfolios in response to every market fluctuation can lead to overtrading, high transaction costs and tax inefficiencies. To avoid this, choosing signals that reflect the periodicity of the rebalancing – faster signals for daily or weekly rebalancing and slower signals for a monthly to quarterly rebalancing schedule – and incorporating a trading cost estimate in any rebalancing optimization framework will help identify allocation changes that are estimated to return higher performance.

Implementing Tactical Asset Allocation Within a Mean-Variance Framework

In one of our previous episodes, we looked at the mathematical framework for strategic asset allocation. We will now extend the mean-variance optimization problem, taking account of the changes in forecasted returns. In mathematical terms, the optimizer aims to establish a new set of weights, w, to maximize expected returns, α, while at the same time controlling the risk of the portfolio and transaction costs:

The vector α defines the expected returns for the next period and the optimizer is maximizing the weighted expected returns, while being penalized for taking on too much risk or unnecessary trading. In tactical asset allocation, α defines the expected returns for the next period, which might be 1 to 3 months. This is different from strategic asset allocation where the expected return is usually the average long-term historical return.

This is an iterative optimization problem to control the risk level of the portfolio. One calculates the asset weights for a certain value of the parameter λ. If the risk is too high, we will increase λ (and similarly decrease it if the risk is too low) and the new portfolio will have a higher penalty on risk, thus leading to a lower-risk portfolio – while also fulfilling all the additional constraints. The iterative process is repeated until a threshold is reached or λ is either close to zero or very large. The latter could arise if one is trying to find a higher-risk solution with only low-risk assets in the asset mix (λ close to 0, i.e. do not restrict the volatility of the portfolio) or trying to find a low-risk solution with only higher-risk assets in the mix (λ is very large and still no solution can be found).

The drawback with mean-variance optimization is that small changes in expected returns can at times produce quite different outcomes. To mitigate this, one usually adds constraints to each of the assets, for example allowing a maximum of 15% for any one asset and much smaller upper bounds for some thematic tilts.

Often one might start with the strategic core portfolio, which is fixed weight, and then allow the assets to move between say -50% and +50% of the fixed weight. One may also allow some additional tilts to the portfolio, for which the lower bound could be 0.

Introducing the Concept of Robust Optimization

One has to keep in mind that the forecast returns are only a rough estimate of any future returns. To mitigate the drawbacks described above when obtaining future weights, one adds another term to the objective function of the optimization problem:

In this case VCVF is the variance-covariance matrix of the forecast alphas, representing the uncertainty of these forecasts (in reality this is often limited to a diagonal matrix as it gets very hard to determine correlations and covariances among the different asset forecasts). κ in the above term is the penalty factor for the uncertainty of the forecasts – similarly to the factors λ and τ, by increasing/decreasing κ, one adds more/less weight to that uncertainty.

In layman’s terms, adding the additional term to the optimization problem reduces drawbacks from not being able to obtain forecast returns that are more than rough estimates for bullish, bearish, or neutral. Can you forecast the returns of the S&P 500 for the next month or quarter to an accuracy of at least one percentage point with any certainty? Most likely not. And most likely not, except in the case of low risk assets such as money market exposure.

Example of Tactical Asset Allocation

To illustrate a real example using ETFs, we examine a simplified case study of a balanced portfolio. This example, though limited to five assets, effectively demonstrates how a structured approach to TAA can achieve measurable outperformance.? The portfolio is denominated in GBP and includes the following assets: UK short-term gilts, UK corporate bonds, global equities, Nasdaq stocks, and gold. For this analysis, we constrain the portfolio's volatility to a maximum of 10%.

Initially, the portfolio is modelled as a fixed-weight strategic allocation with the following weights: UK short-term gilts (IGLS) – 35%, UK corporate bonds (UKCO) – 10%, global equities (IWRD) – 40%, Nasdaq (EQQQ) – 10%, and gold (SGLP) – 5%. Subsequently, we employ an optimization process that permits each asset's weight to vary by ±50% from the baseline allocation. The optimization is performed monthly, using data available at the end of each period. Asset alphas are derived from a combination of macroeconomic indicators and trend-following signals. At each iteration, the risk adjustment factor, λ, is calibrated to ensure that the portfolio’s predicted volatility does not exceed the 10% target.

Below, we present the historical asset weight changes, where each vertical line represents a specific point in time. The graph highlights fluctuations in the equity allocations (IWRD and EQQQ), fixed-income securities (IGLS and UKCO), and gold (SGLP).

Backtest Results

The backtest results are derived by calculating signals at the end of each month and implementing trades based on these weights on the second business day of the following month. The performance statistics for the GBP-denominated portfolio are summarized below. To thoroughly evaluate any strategy, many variations of the strategy should be tested, for example expanding the asset universe, setting different risk thresholds, and applying the methodology across multiple currencies.

The historical performance for the backtest is as shown below. For this we calculated the signals at the end of each month and assume that we trade on those weights on the second business day of the next month.

This results in the statistics as given below for the GBP portfolio. In order to test such a strategy, one should test many variations of it – adding some more assets, checking higher versus lower risk levels and running the same strategy in different currencies. In our study, the examples were extended to similar USD and EUR portfolios. Equity and gold allocations were maintained, while the fixed-income components were adjusted to include short-term US/EUR government bonds and all-maturity US/EUR corporate bonds, replacing their UK counterparts. And similar to the GBP portfolio, the fixed-weight portfolios assumed monthly rebalancing to baseline weights, whereas the TAA portfolios allowed ±50% deviation in asset weights. In all cases, the portfolio volatility was capped at 10% annually. Notably, the fixed-weight USD portfolio exceeded this limit slightly, reaching 10.3%, while the TAA approach successfully reduced it to the target of 10%.

Tactical Asset Allocation in Action

This simplified example underscores key considerations in evaluating asset allocation algorithms. In practice, portfolios often comprise a broader range of assets, potentially a dozen or more. It is crucial to acknowledge the inherent limitations of any model; TAA serves as a robust starting point for portfolio rebalancing, but an informed judgment and a potential discretionary override from an investment committee can enhance outcomes. Sometimes the override might also just mean to not rebalance, thus reducing trading costs.

All in all, a decent improvement to performance given one is simply re-weighting the allocations every now and then!

Source: All data from 31 Dec 2012 to 18 Dec 2024.

About Algo-Chain

Algo-Chain helps Wealth Managers & Financial Advisors who are looking to create their own Model Portfolios and increase their share of the value chain via their White Label Model Portfolio Service.

https://www.algo-chain.com/WealthManagerSolutions

Disclaimer

*The podcast provided by Allan Lane & Irene Bauer has been converted from their own original content, into a podcast using Generative AI tools and the voices used in the podcast are not their own.??All information provided has been fact checked.

The content referred to in this podcast is targeted at professional Wealth Managers & Financial Advisors and may not be suitable for all investors. Twenty20 Solutions Ltd does not provide, and nothing in this podcast should be construed as, investment or other advice. It is not intended that anything stated in this podcast should be construed as an offer, or invitation to treat, or inducement for you to engage in any investment activity. The information in this podcast relating to model portfolios & individual funds suggested by Algo-Chain is purely for research and educational purposes only.