How to Measure Investment Risk
KEY POINTS
Beta, alpha, R-squared, standard deviation, and the Sharpe ratio are fundamental tools utilized to measure investment risk
These tools incorporate risk and return and facilitate critical comparison of managers
Statistical based tools have inherent limitations and potential biases
What is investment risk?
We invite you to read our article Types of Investment Risk - A Universal Guide to familiarize yourself with the various risks impacting your investment plan. These risks fall into two categories: systematic and unsystematic.
Systematic risks are non-specific (e.g., interest rates, inflation), affect the entire market, and cannot be reduced through diversification. However, such risks may be mitigated through hedging.
Unsystematic risks are specific (e.g., credit/default, liquidity, legal, regulatory, concentration), do not affect the entire market, and can be reduced through diversification.
What is risk management?
It is a process whereby you identify and measure risk to ensure the risk taken is within your risk tolerance range. You may have multiple risk tolerance ranges for different investment goals and related time horizons.
For example, you may have the following savings goals: 5 years buy a house, 15 years fund children’s post-secondary education, 40 years retire. Your risk tolerance typically increases as your time horizon increases. Therefore, your retirement portfolio may have a higher risk profile than your housing fund.
Financial models attempt to measure market risk and its impact on real world scenarios. Market risk includes changes in stock prices, interest rates, exchange rates, inflation, and commodity prices. These financial models do not capture credit risk (e.g., counterparty fails to pay interest on their bond) or business risk (e.g., strategic planning, operational).
Effective risk measurement allows a portfolio manager to estimate the impact on an investment given a change in a specific risk factor. Such impact may include the size and frequency of a potential gain or loss.
To be clear, financial models are based on historical data and not all models are built equally. Therefore, such results must be interpreted by seasoned investment professionals.
For example, Spurious Correlations illustrates that the marriage rate in Kentucky is highly correlated with the number of people who drowned after falling out of a fishing boat. Such data may be correlated, but the events themselves are obviously unrelated. This example has nothing to do with investing, but it illustrates an inherent risk in building financial models and interpreting their results.
We will now focus on the key statistical measurements you can utilize when measuring risk.
5 Tools to Measure Investment Risk
1. Beta (β)
Beta measures systematic risk. The broad market beta (e.g., average of all the individual security betas in the S&P 500) equals 1. An individual security’s beta is measured in relation to the broad market beta.
If your security has a beta of 1.5, such would infer it historically moves 1.5 times as much as the S&P 500. Therefore, if the S&P 500 were to increase/decrease 10% the security with 1.5 beta would historically increase/decrease 15%. Similarly, if the security’s beta was 0.5, it would move half as much as the market.
Let’s look at a random sample of securities and their respective betas to gain some perspective. The following betas were accurate at the time of this writing and are subject to change:
This information may be available from your financial institution, and many places on-line. For example, Norwegian Cruise Line Holdings Ltd beta can be found here on CNBC .
We will highlight two limitations of utilizing betas.
First, it is unlikely for a company to maintain a steady growth rate throughout its existence. Therefore, the historical data utilized to calculate the beta will have a blend of performance that is not indicative of its current or expected rates.
Second, based on the first limitation, different time periods will exhibit varying degrees of correlation to the S&P 500. Therefore, a security’s beta could be different depending on the time period utilized. Due to this reason, it is possible to obtain different beta values for the same security from multiple websites / financial institutions.
2. Alpha (α)
Alpha measures an investment manager’s ability to generate returns in excess of the market. Alpha can be positive or negative as managers can outperform and underperform the market.
Modern portfolio theory incorporates the efficient market hypothesis. This assumes all security prices (e.g., stocks, bonds, options, futures, forwards, private equity, hedge funds) reflect all known information at any given point in time. If a security becomes mispriced, it is assumed the related arbitrage opportunity will be quickly taken advantage of and the security price will be correctly adjusted by market forces.
Based on such theory, it should be virtually impossible for an investor or portfolio manager to outperform the market over an extended period. You will often hear it referenced that very few investment managers can outperform the market for three consecutive years. Such becomes even rarer when you factor in management fees, performance fees, and taxes.
This is where we see the distinction between beta and alpha.
Beta can be earned through passive investing. For example, you can earn the market return of technology stocks by investing in a diversified exchange traded fund comprised of technology stocks.
Alpha requires active trading as it measures returns in excess of the market. As passive investing will generate beta neutral market returns, it follows that active trading would be required to achieve returns in excess of the market.
A key limitation to utilizing alpha is the potential for the incorrect benchmark to be utilized. For example, it would be inappropriate to utilize the S&P 500 as a benchmark if the investment manager’s mandate is to actively trade bank stocks. Banking is an industry within the S&P 500.
3. R-squared (R²)
R-squared measures the degree of a fund or security’s performance is attributable to a benchmark index. It is measured in a range of 0 to 1 and often expressed as a percentage (e.g., 10%, 60%, 90%).
A high R-squared value indicates an increased correlation with the benchmark. Investors must consider the value added by active managers if their fund has a very high R-squared indicating much of their performance is attributable to the benchmark. It is highly probable that such benchmark returns can be obtained through an allocation to less expensive exchange traded funds.
R-squared can be viewed in relation to beta. For example, a fund with a high R-squared and high Beta may produce returns that exceed the benchmark. If such excess returns were generated, you would expect a positive alpha measurement.
4. Standard Deviation (σ)
Standard deviation is a statistical measurement of volatility. Such volatility equates to total risk which includes systematic risk and unsystematic risk. It measures the spread of a security’s price from its average price. When the security’s price experiences large movements up and down it will have a higher standard deviation than if its prices experienced more modest swings.
Standard deviation can be utilized to measure the volatility of any single security, portfolio, and market index. Higher volatility equates to higher risk.
You can calculate the standard deviation in Excel utilizing the STDEV formula. Your brokerage firm should provide access to historical stock quotes necessary for such calculations. Alternatively, there are free on-line sources available (e.g., Yahoo Finance, Nasdaq).
A low standard deviation does not necessarily equate to better returns. Remember, standard deviation only measures volatility. Let’s compare two securities to illustrate the point.
Security A has a consistent monthly return of -1%. Security B has alternating monthly returns of -1% and 2%. Security B has a higher standard deviation and cumulative return than Security A.
The investor would have experienced poor performance if the investment decision was based solely on a low standard deviation. Expected return is key part of the decision-making process. Investors will generally expect higher returns to compensate for higher risk.
Like beta above, standard deviation can be impacted by the historical data set utilized (e.g., 3 years of return data versus utilizing 10 years of return data) and how such data periods align with current and future periods.
The model is also susceptible to significant outliers. For example, the terrorist attacks on September 11, 2001 created varying degrees of shock to world markets. These shocks translated into outliers with a statistically significant impact on volatility. These outliers have an outsized impact on standard deviation.
5. Sharpe Ratio
The Sharpe ratio can be utilized to assess an investment manager’s skill by quantifying the relationship between risk and reward. It divides the return realized in excess of the risk-free rate of return by the security/portfolio’s standard deviation.
It normalizes returns for a given level of risk. Investors can compare Sharpe ratios to determine how well they are being compensated for every unit of risk that the take.
The Sharpe ratio is prone to the same disadvantages noted under the Standard Deviation section above.
Measuring Investment Risk – Bottom Line
The above tools are useful when measuring risk, comparing investment managers, and comparing investments against one another. We recommend you consider the types of advisors you need to engage and apply the steps in our definitive guide when choosing such professional advisors to ensure your investment plan uses these tools appropriately.
