This article has been motivated by the tendency of investors to search for THE PERFECT indicator. This article aims to illustrate the advantages of combining different indicators, even though they might sometimes seem to contradict themself.

‍

For this purpose, we will use indicators developed by Alquant.

‍

In this article, we will focus on three indicators that we believe can be effectively combined to create a simple but proactive tactical equity exposure management system that should be able to detect crises of various origins, thus protecting investors’ assets from a variety of threats.

‍

Presentation of Alquant’s indicators and how to apply them to a portfolio

‍

The first indicator is the Credit indicator. This indicator tracks current credit conditions, which should be considered as a leading indicator for the stock market. Why? Because it is often the case that when credit dries up (credit conditions are unfavorable), all the bad things happen. In fact, in an unfavorable credit environment, companies that live on debt are unable to obtain further financing, default, or even go bankrupt, which in turn harms the overall stock market. Alternatively, it can be argued that the lower liquidity and lower volatility of the corporate bond market generally force credit investors to take materiality considerations into account when responding to new information. Indeed, credit investors are looking for yield rather than capital gain. Thus, if there is evidence that on an aggregate level credit investors are willing to react and thus trade on new information, this should be interpreted as a serious warning sign.

‍

The second indicator is the Macroeconomic (Macro) indicator. This indicator takes a bird’s eye view of the real world economy (macroeconomic perspective) and does not only focus on the financial markets, which is crucial to identify financial bubbles and exaggerations. However, unlike many macroeconomic indicators, Alquant uses daily and weekly published economic data instead of monthly or quarterly data to ensure very responsive indications. In addition, rather than focusing on the current level or even the current trend of economic activity, we focus on the change in the rate of change (i.e., the sign of the second derivative) to anticipate whether the trend will not continue, striving for improved predictive power. Hence, overall this makes Alquant’s Macroeconomic indicator not necessarily a good proxy for the current state of the economy but rather an indicator employing “high-frequency” macroeconomic data to anticipate future short-term equity performance.

‍

The third indicator is the Vega indicator. This indicator aims to anticipate changes in the volatility of the US stock market. In fact, predicting future volatility is orders of magnitude easier than predicting future returns because volatility tends to cluster empirically. This phenomenon does not violate the market efficiency hypothesis because accurate volatility prediction does not conflict with the correctness of underlying asset and option prices. The idea, then, is to use the fact that volatility itself is volatile and to some extent predictable to construct a Vega indicator that predicts volatility spikes. The Vega indicator is the origin of Alquant and therefore special to us; it can be considered Alquant’s secret recipe

‍

All three indicators have been designed to help adjust the equity exposure to the prevailing market conditions, thus enabling proactive and easy-to-implement risk management. Each day, the indicators provide a risk level ranging from 0% to 100%. Thus, the equity exposure should be inversely correlated to the risk level; the simplest implementation of this is 100% minus the current risk value.

‍

Statistical analysis

‍

Obviously as one can intuitively guess from the section above, although the three indicators have the same purpose, namely dynamically adjusting equity exposure, they are quite different and thus independent. This is confirmed statistically by comparing Kendall’s tau correlations. We can see from the correlation matrix below that our indicators have correlations close to 0, which strengthens our hypothesis of independence.

‍

‍

A correlation close to zero means that there is no linear relationship between the indicators meaning that when one indicator identifies risk or a warning sign the others on average won’t. In other words, they simply behave differently.

‍

Okay, but why is a low correlation good? Because this is when the diversification benefit of combining is maximized. If you take football as an example, assuming you could clone Cristiano Ronaldo (or Messi) eleven times won’t make a great team. You would prefer adding players which might overall be less skilled than Ronaldo but at least have other skills and behave differently.

‍

It is worth mentioning that this observation of low correlation holds on average. However, during major turbulences and crises, the correlation of our three indicators tends to increase strongly. This means that, for example, during the financial crisis, all three different sources agreed that something very serious was happening. If we look again for an analogy, we could say that it is a bit like if all the newspapers report a news story, there is a good chance that it is true and not just fake news.

‍

‍

Next, let’s look at the autocorrelation of each indicator. The autocorrelation of an indicator is its correlation with the same signal lagged by several days. This metric allows us to quantify the inertia of a signal, i.e. its reactivity and therefore the investment horizon to which it is best suited. In other words, a highly autocorrelated indicator means that if today is described as risky, tomorrow will probably be described as risky again, which leads to indicators with long risk-off or risk-on periods and few switches.

‍

‍

From the autocorrelation graphs above, we can draw the qualitative conclusion that the indicators have very different inertias, ranging from the most reactive and changeable (Macro) to the most autocorrelated indicator (Credit). Therefore, each of the three indicators has a distinct sensitiveness.

‍

Backtests

‍

Let’s start our analysis with a backtest where we apply our indicators on a stock index: the S&P 500. As explained in the previous section, the exposure chosen each day is 100% minus the value of the indicator the previous day. In this section, we introduce the Combined indicator, which is the arithmetic average of the Credit, Macro, and Vega indicator. The objective will be to evaluate the interest of diversification between the indicators.

‍

‍

The performance graph above clearly shows that, as suggested by the autocorrelation analysis, the Credit indicator has long risk-off periods. At the same time, except for the financial crisis and COVID-19, it is much more difficult to identify when the Macro indicator is risk-off, making it look very similar to the S&P 500.

‍

Note: The pain ratio is a risk-return trade-off metric developed by Zephyr Associates that compares the value-added relative to the risk-free rate to the depth, duration, and frequency of losses. The return component of the pain ratio is the annualized return on the investment relative to the risk-free investment. Generally, a short-term cash investment is used as a risk-free investment. The denominator of the pain ratio is the pain index, an integral measuring the depth, duration, and frequency of losses.

‍

By analyzing the backtest statistics, we can observe that each of the indicators is already performing well on a stand-alone basis. Thus, each of the indicator overlay strategies has significantly reduced volatility. Thanks to this feature, the Sharpe ratio is significantly increased compared to a buy and hold S&P 500 portfolio. In addition, each indicator also reduced the maximum drawdown by a considerable amount, making such overlaid investments less prone to drawdown.

‍

From a pure return perspective, two indicators offer much higher returns than a simple passive investment in the S&P 500. One might be concerned about the returns offered by the Credit indicator, but it is worth remembering here that with an average exposure of only 30%, it achieves almost the same returns as the S&P 500.

‍

However, if someone only looked at the average return (or CAGR) and the performance graph, they might be tempted to ask questions like:

‍

Which indicator is the best?

‍

Why not just keeping only the best performing indicator?

‍

Why trust the Credit indicator when it underperforms the Macro?

‍

However, keeping in mind the results obtained in the second section, namely a low correlation between the different indicators as well as different inertia, and the intuition gained in the first section, namely that these indicators are focused on different market forces, one could assume that by combining these three indicators, the backtested results will be improved. Alternatively, one could be confident that Aristotle was right and that “the sum is more than its part”.

‍

But the most striking information we can extract from these statistics is that the combined indicator provides the highest risk-adjusted return, with the highest Sharpe ratio and lowest maximum drawdown in absolute terms. This confirms that diversification among robust indicators offers better long-term performance.

‍

Conclusion

‍

This article showed that the three indicators analyzed were of irreplaceable and non-redundant interest compared to the others. In other words, each indicator presented has its own dynamics and relevance for the timing of actions. In addition, each of them allows us to manage risk in the short or medium term. As we cannot predict the future, each indicator can provide useful information for portfolio management and is legitimate. We, therefore, recommend using them together, without favoring any indicator in particular. In this way, the timing of stocks will be diversified and adapted to different types of situations and risks.

‍

Disclaimer

‍

This content is advertising material. This content as well as all information displayed on Prisma or 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.

‍