Introduction

‍

In the first article, we observed that rebalancing according to the variations of a risk indicator can under certain circumstances be more interesting than rebalancing on a fixed schedule. However, we noted that following a risk indicator can increase the rebalancing frequency, which may pose a problem for some asset managers. We also found that it is possible to smooth a signal to reduce this frequency, but such a process can significantly alter the indicator's effectiveness if it is not properly calibrated.

‍

In this second article, we will see that an excellent way to reduce noise and therefore the number of rebalancing actions for a signal is to combine it with other indicators. This diversification allows for retaining most of the value provided by the underlying indicators while making the resulting signal more stable. To demonstrate this, we will base our analysis on risk indicators developed by Alquant.

‍

Combining five Alquant risk indicators

‍

For this study, we will use a diversified indicator based on an average of five risk indicators available on Prisma (a platform of actionable risk indicators provided by Alquant). The goal is to have a robust signal that encompasses a wide range of financial and economic metrics. A diversified approach is key to achieving stable long-term performance, as it allows for the consideration of a wide range of factors that quantify aspects of economic reality. This maximizes the chances of predicting risks that could impact the markets in the short or medium term. This signal will be called the Combined Indicator in this article. All Prisma indicators vary from 0% (low risk) to 100% (high risk).

‍

Figure 1: Illustration of the 5 risk indicators combined into a single signal.

‍

This indicator provides a daily value indicating the ideal equity exposure that a portfolio should target. The higher the Combined Indicator, the greater the risk in equity markets, and consequently, the lower the optimal equity exposure.

‍

Case study 1: Dynamic equity-cash allocation

‍

Many asset managers manage their equity allocation tactically but must adhere to strict constraints. Typically, they rebalance their equity allocation at regular intervals, such as quarterly or monthly. This approach makes it challenging to add long-term value to a portfolio and difficult to hedge against sudden market downturns with a fixed rebalancing schedule.

‍

For this case study, we will consider a simple universe comprising two assets: an ETF tracking the S&P 500 index (SPY) and USD cash. To simplify, we assume the asset manager can adjust their equity and cash exposure from 0% to 100%. Our objective is to rebalance between equity and cash based on the Combined Indicator without creating a complicated and unusable rule.

‍

We will test a straightforward binary strategy: if the Combined Indicator is above a threshold value (indicating high risk), we move to a 100% USD cash position. Otherwise, we adopt a 100% US equity position. We will analyze two strategies, each using a different rebalancing threshold, and compare the results: one with a threshold at 50% and the other at 66%.

‍

Of course, this strategy could be further tailored to specific constraints, such as setting the minimum and maximum equity exposure to the lower and upper Tactical Asset Allocation (TAA) boundaries instead of 0% and 100%, while keeping the same methodology. Additionally, we could define a neutral state in which the strategy would follow the Strategic Asset Allocation (SAA) equity exposure.

‍

Figure 2: Equity-cash strategy building recap

‍

As shown in the histogram in Figure 2, the Combined Indicator usually displays low values, which means the estimated risk of investing in equities is rather low. Therefore, we trigger a rebalancing only when the indicator rises above a high fixed threshold. This high threshold filters out lower risk values, and rebalancing occurs only when a significantly high risk (above 50% or even 66%) is detected. This approach avoids rebalancing every time the indicator changes, which would result in approximately 75 rebalancings per year (for an information ratio of 0.95).

‍

The table below outlines the "simple in or out" strategies we will evaluate. This approach involves deciding whether to be fully invested in the market or to exit based on specified criteria. These strategies simplify decision-making by reducing it to a binary choice, thereby streamlining the investment process for testing and analysis.

‍

Figure 3: Exposure in equity depending on the Combined Indicator’s values

‍

Our benchmark for this analysis is a portfolio consisting of 50% Equity and 50% USD Cash, rebalanced twice a year. Now, let's examine the results for these simple strategies and evaluate the frequency of rebalancing required. A transaction cost of 0.05% is applied to each rebalancing action.

‍

Figure 4: Performance of the strategies tested, from 2008-01-15 to 2024-05-20

‍

Figure 5: Statistics of the strategies tested, from 2008-01-15 to 2024-05-20, including 0.05% transaction costs.

‍

First and foremost, the two strategies significantly outperform the benchmark in both return and risk-adjusted metrics, such as the Sharpe and Calmar ratios. Notably, these strategies dramatically reduce the number of rebalancings - from 75 (if rebalanced every time the indicator changes) to 10.7 (with a threshold of 50%) and 4.1 (with a threshold of 66%) while still generating significant Alpha, at 9% and 5%, respectively.

‍

Although these lower rebalancing frequencies do slightly reduce the information ratio from 0.95 to 0.93 and 0.58, the two threshold strategies greatly improve the information ratio per rebalancing. Specifically, the information ratio per rebalancing increases from 0.013 to 0.087 for a threshold of 50% and to 0.141 for a threshold of 66%.

‍

This reduction in rebalancing frequency, based on thresholds, is highly effective because the Combined Indicator benefits from the diversification effect. This effect stabilizes the Combined Indicator, avoiding noise and false signals, while still effectively identifying high-risk periods. Consequently, this approach not only optimizes long-term returns and risk-adjusted performance but also minimizes the frequency of rebalancing needed to manage risk.

‍

Recall that in our first article, we demonstrated the impact of reducing the rebalancing frequency of a single indicator (RSI indicator) using a standard smoothing approach. Although this method successfully reduced the rebalancing frequency, it also proportionally decreased the information ratio, leaving the information ratio per rebalancing roughly the same. This further underscores the added value of the approach presented in this second article.

‍

Let's take a quick look at the plots below, which show exposures over time.

‍

‍

Figure 6: Exposures over time of both strategies tested

‍

Figure 6 demonstrates how the Combined Indicator helps avoid unnecessary rebalancing when risks are lower, thus maximizing benefits from bull markets. Simultaneously, it enhances management during periods of high risk. Rebalancing periods are concentrated into specific, relatively short periods that require special attention.

‍

In this second article, we demonstrated how to achieve enhanced and more robust performance compared to a 50%-50% portfolio. This was accomplished by diversifying the strategy among different indicators and utilizing a threshold to trigger rebalancing, thereby maintaining manageable rebalance frequencies (less than monthly on average).

‍

In the next and last article, we will focus on a portfolio manager constrained to maintain 100% exposure to equities. We will explore how to optimize such a portfolio while keeping rebalancing actions to a minimum.

‍

Disclaimer

‍

‍This content is advertising material. This content as well as all information displayed on any of Alquant’s websites does not constitute investment advice or recommendation, and shall not be construed as a solicitation or an offer for sale or purchase of any products, to effect any transactions or to conclude any legal act of any kind whatsoever. Past performance is not a guide to future performance.