EQUITY PORTFOLIO RISK MONITORING

*Please note all views are personal*

Introduction

It is well noted that timing markets for investment gains is quite hard. But still, one need to evaluate risk to help build fail-safe portfolio mechanism. Risk is often misunderstood and is often seen as the loss made on the portfolio. We should be clear that gain and loss in a portfolio is an outcome. It’s like realising Titanic ship doesn’t have adequate number of lifeboats compared to the number of people on the adverse night. Evaluating Risk in this case is about identifying inadequate lifeboats before the Titanic sailed.?

So, what are the parameters of Investment Risk which one need to evaluate and comprehend?

Investment Risk Analyst usually monitors volatility, tracking error, VaR, stress scenarios, diversification benefit, allocation and selection risk, sentiment risk, portfolio style changes, portfolio turnover, cross-sectional volatility, liquidity risk based on market and investor concentration, valuation risk, macro trends and most recent ESG parameters etc and various ways of modelling these risk parameters like modelling them over different time horizons.

A portfolio risk analyst has a hard task of differentiating risk signal from noise and communicating the same to portfolio managers and board. For example, when volatility of portfolio and benchmark goes up, it is normal for the tracking error to move up. Simultaneously, sensing high volatility, a value fund manager may move from large cap value tilted portfolio to midcap value tilted portfolio or quality value tilted portfolio. This may not be bad but if combined with upside capture ratio and downside capture ratio may give an insight if high tracking error is helping the portfolio. This combined with other indicators like valuation risk (quality portfolio may have high valuation risk), sentiment analysis and liquidity risk could throw better insights about the risk of the portfolio.

The best way of understanding the changes is to do trend analysis. Risk Manager need to follow and understand trend of various risk parameters and understand the reason for changes. It is important to understand if there are any relationship between various parameters and to distinguish increase/decrease in trend of parameters is not simply a noise but actually a signal of risk profile changes in the portfolio. The portfolio risk analyst is also expected to reassess the risk parameters to make sure it is relevant in current market scenario and its portfolio characteristics.

While various risk parameters are continuously monitored by risk managers, one is not sure if there is any better way of handling all the risk parameters and how to prioritize them in their daily monitoring. How should we relate all these risk parameters during different market condition and bring out any adverse risk in the portfolio?

PERSPECTIVE FROM LITERATURE

In order to understand the risk in the portfolio, one need to understand the various external factors at play and relationship of portfolio with these external factors. This relationship with external factors is computed as beta. However, there are various external factors against which one would like to understand the portfolio.

This is where Arbitrage Pricing Model comes handy compared to CAPM model. A multi beta model also have high explanatory power to describe the future returns of the portfolio. But again, this may be insufficient as explained by Roll et al, 1980. So, one need to find a set of factors more relevant to current market condition and portfolio.??This can be difficult for a market risk practiser as he would have to take many factors into consideration. I agree with (G.C. Heywood et al, 2003, p4), that “the problems and challenges of portfolio construction and risk management are every bit as much a ‘science’ as they are an ‘art’. It is highly desirable to differentiate the intended and unintended bets taken by a fund manager. G.C. Heywood et al, 2003 explains how cluster analysis helps in understanding changes in the risk structure of the market. The paper further explains how to decompose risks like tracking error and how objectively to look at covariance terms. The kind of changes in structure of risk of the market over time and relating to portfolio is very vital to narrow down the risk indicators that one should investigate. This also helps in understanding the causality of changes in risk.

While Standard deviation and value-at-risk are two risk parameters widely used by risk practitioners globally, they serve two different purpose and hence their usage is very different. Apart from these market risk measures, credit risk in market and portfolio cannot be overlooked. As mentioned by Robert and Stuart, (The intersection of market and credit risk ,2000) “Empirically returns on high yield bonds have a higher correlation with equity index returns and a lower correlation with Treasury bond index returns than do low yield bonds”. Also, there could be short term market scenarios where low yield bonds also could have high correlation with index returns. Hence, while accessing risk of portfolio, assessing market scenario, credit scenario and liquidity scenario are vital parameters that need to be understood and modelled.

Portfolio characteristics are measured against their benchmark and so is volatility of portfolio. Benchmark volatility can be computed using benchmark returns. This may not appropriately capture the risk across stocks and sectors. This is because index construction is based on market cap of stocks in the portfolio and hence securities with higher market cap has more influence on the return of the index and volatility. A better way to understand the market volatility is to understand the dispersion of individual stocks and then combining them to understand the risk structure changes. This is where the cross-sectional volatility (CSV) of stocks returns helps in understanding the risk in the market. In paper by?St?ckl et al, 2016,?which adds to work of Maio (2016) where he finds that cross-sectional volatility of stocks helps in forecasting decline in equity premium and further observes that cross-sectional skewness spans the predictive quality of cross-sectional volatility over short-forecasting horizons and cross-sectional kurtosis significantly contributes to long-horizon forecasting of 12 months and above.?

It is important for Risk Analyst to understand if the market is in a ‘risk on’ or in a ‘risk off’ mode. Analyst looks at various parameters including gold price movement to understand the same.

Structuring and understanding all these and many other parameters in different market scenarios helps in better analysis of portfolios.

For analysing risk, one can opt for a top-down or bottom-up approach. Looking into macro-economic variables and relating to portfolio is a top-down approach. While macroeconomic variables are easy to understand, they are published with a lag. However, there are various high frequency indicators which can be observed to help understand or predict the macroeconomic condition. Nowcasting methodology, as mentioned by Banbura et al, 2010 in their paper “Nowcasting” is one such methodology used for explaining changes in macroeconomic conditions. Also during analysis, one need to keep in mind that macroeconomic variables may not be able to explain the short-term security variance but has high explanatory power when horizon is extended to one year as explained by K. K. Guru-Gharan et al, 2009.

While it might be difficult to predict the macro scenario, a scenario dashboard is quite helpful to explain how portfolio may be impacted during different market condition and might be easy to bring out the inherent risk in the portfolio. One can also assign probabilities to these events.

Since macroeconomic risk impacts all businesses, it is classified as systematic risk. A simple measure of Beta is used widely to capture the systematic risk in one’s portfolio. There are various macroeconomic variables which can be used to measure systematic risk and understand its impact on portfolios but it’s difficult to understand and analyse unsystematic risk. Though it can be diversified but understanding and monitoring unsystematic risk component is hard at security level. One way could be monitoring the diversification benefit that one gets because of having diversified stocks in the portfolio. This is the covariance term while computing bottom-up portfolio volatility.?

With Arbitrage Pricing Theory, a linear relationship was shown between security return and various market factors or systematic factors as mentioned by Connor et al, 1993. These market factors are easy to observe, measure and understand on a real time basis. This helped creating various multi-factor model like Fundamental factor model (where sectors and style like size and value factors are used to understand the return of the portfolio), Macro-economic model (where macroeconomic factors like inflation etc are used to understand the return of portfolio), Statistical Model (where security returns are decomposed using methodologies like principal component analysis to understand the return of portfolio) and various hybrid models (which is a mix of above three models).

Each of these models is used by different market participants to access risk in their portfolio. While asset owners have a longer horizon and might have investments across regions and asset classes, macro-economic models might be better to analyse and comprehend the risk. Similarly, fundamental factor model might be helpful for a fundamental fund house who believes in bottom up approach of investing.

Apart from Macro factors, one also need to decompose the portfolio risk which can be a bottom-up approach. A simple way to understand the portfolio risk is to understand the concentration bets that the fund manager is taking. Simply looking into sectoral weights explains the sectoral tilt of the portfolio. Similarly, one can investigate various style bets that the fund manager is taking by grouping the portfolio by market cap, PE ratio buckets etc.

For security liquidity measure, impact cost is a good indicator and need to be monitored against redemptions or liability side concentration and could be integrated with other measures of risk. Liquidity Value-at-Risk is a good measure to combine the liquidity risk and tail risk of securities.

Conclusion

Risk analysts need to understand the various sources of risk and should communicate this risk differently to different stakeholders. For example, a portfolio manager might need a deep dive of his portfolio but marketing team might need to understand the key differentiator of the portfolio compared to peers to pitch the investment risk to investors. Board members may be more interested from regulatory and macro risk perspective. An efficient risk dashboard can be created keeping the stakeholder interest in perspective.?

So, studying various risk parameters and their relationship in the economy could help create dashboard of early warning signals which in turn need be related to the portfolio to create informative dashboard about portfolio risk. Also, bottom-up portfolio analysis helps understanding the various exposures portfolio has and this in turn could be cross-checked with various early warning parameters. This could help risk manager have comprehensive view on the subject and help communicate the same to different stakeholders.

References

By Heywood, G. C. , Marsland, J. R. and Morrison, G. M. (2003). Practical Risk Management For Equity Portfolio Managers. Presented to Institute of Actuaries, 28 April 2003, https://www.actuaries.org.uk/system/files/documents/pdf/sm20030428.pdf?

Guru-Gharan, K. K., Rahman, M., & Parayitam, S. (2009).?Influences Of Selected Macroeconomic Variables On U.S. Stock Market Returns And Their Predictability Over Varying Time Horizons.?Academy of Accounting and Financial Studies Journal,?13(1), 13–31. Retrieved September 20, 2021, from?https://www.proquest.com/openview/78d977f7f35a26c76db3fcf28621d068/1?pq-origsite=gscholar&cbl=29414.?

Roll, Richard, and Stephen A. Ross. “An Empirical Investigation of the Arbitrage Pricing Theory.”?The Journal of Finance, vol. 35, no. 5, [American Finance Association, Wiley], 1980, pp. 1073–103,?https://doi.org/10.2307/2327087.

Jarrow, Robert & Turnbull, Stuart. (2000). The Intersection of Market and Credit Risk. Journal of Banking & Finance. 24. 271-299. 10.1016/S0378-4266(99)00060-6.

St?ckl, Sebastian and Kaiser, Lars, Higher Moments Matter! Cross-Sectional (Higher) Moments and the Predictability of Stock Returns (November 9, 2016). Available at SSRN:?https://ssrn.com/abstract=2747627?or?https://dx.doi.org/10.2139/ssrn.2747627

Banbura, Marta and Giannone, Domenico and Reichlin, Lucrezia, Nowcasting (November 30, 2010). ECB Working Paper No. 1275, Available at SSRN:?https://ssrn.com/abstract=1717887

Connor, Gregory and Korajczyk, Robert A., The Arbitrage Pricing Theory and Multifactor Models of Asset Returns (September 30, 1993). Handbooks in Operations Research and Management Science, Vol. 9, Page 1, Available at SSRN:?https://ssrn.com/abstract=1441422