The BlackRock Scenario Tester

About the analysis

The BlackRock Scenario Tester measures the potential impact of discrete market events on a portfolio using a multi-factor statistical risk model. It does this via a three-step process:

1. Determine a portfolio’s asset allocation, and translate this into underlying holdings using relevant market indexes.
2. Estimate the volatility of a portfolio of those indexes by mapping their exposure to relevant equity, fixed income and alternative risk factors.
3. Measure the historical sensitivity, looking back over 10 years of history, of a portfolio of those risk factors to changes in the market.

What do the results show?

The scenario test results show a statistical estimate of a portfolio or asset class’s (represented by market indexes) reaction to a hypothetical market event. For example, we may calculate that a portfolio could expect to return -5% in the event that the S&P 500 were to fall -15%. It is important to understand that this is just one potential outcome, as a statistical estimate is sensitive to assumptions built into our risk model. More detail on the assumptions and limitations of our risk model are below. The calculated return includes the underlying price movement in reaction to the scenario, as well as 12 months of income yield for fixed income instruments. For historical scenarios, the income yield reflects the time period of the scenario, which may be more or less than 12 months.

What is a risk factor?

A risk factor is a technical or fundamental characteristic of a security which, statistically, is able to help explain the risk / return behavior of that security.

We believe risk factors should be:

1. Interpretable: easily understandable and have a strong economic rationale
2. Explanatory: proven to explain volatility and correlation of returns
3. Consistent: significant through time across different market regimes
4. Effective: able to improve forecasts of portfolio risk and attribute portfolio performance

BlackRock’s risk model uses over 2,200 distinct risk factors across equity, fixed income, currencies and alternative investments.

Examples of risk factors include:

• Equity market risk– the inherent risk of investing in the stock market, which cannot be diversified away with stocks alone.
• Equity style risk – a collection of factors that describe the style of a stock, such as growth, value, momentum, or size.
• Interest rate risk – a fixed income risk attributed to changes in market interest rates, the dominant risk in high quality bonds like U.S. Treasury securities.
• Credit spread risk – the risk of default present in any bond not issued by the U.S. Treasury. The lower the quality of the bond, the more credit spread risk dominates.
• Many others that are subsets of the categories above, as well as other categories like foreign exchange risk, inflation risk, and alternative risk.

What assumptions are built into the risk model?

The model for this tool uses ten years of historical factor returns, derived using a statistical regression analysis, to measure the volatility of the factors we are stressing. The portfolio’s exposure to each factor is measured through currently observable characteristics of the underlying securities, for example fundamental characteristics such as financial ratios, technical analysis attributes such as price behavior or liquidity, and / or specific attributes of a given security such as yield, geographic domicile and currency exposure. The level of exposure to a factor of a given security corresponds to the where the security ranks relative to the distribution of all securities in the universe for the characteristic in question, for example whether the price to book ratio for a company is very high or very low relative to other companies. Historical covariance across the factors is taken into account to measure total estimated volatility of the overall portfolio. Certain elements of the risk model are proprietary to BlackRock.

How accurate are the results?

The scenario test results show a statistical estimate of a portfolio or asset class’s reaction to a hypothetical market event. This is done using a risk factor framework, as described above. The scenario test results are subject to uncertainty, and we can measure this uncertainty using the prediction error. The prediction error describes the range of possible outcomes around the scenario test result where 95% of possible outcomes should occur. For example, we may calculate that a portfolio could expect to return -5% in the event that the S&P 500 were to fall -15%, with a prediction error of 2%. Based on the prediction error of 2%, we would expect that 95% of the time the portfolio would return between -3% and -7% (-5% +/-2%) in the event that the S&P 500 fell 15%.

The prediction error will vary depending on the portfolio or asset class and the scenario. A larger prediction error implies that the test result may be less certain. Prediction errors tend to be larger when the underlying shock is larger; the risk factor model is less able to explain the portfolio risk (for example, in investments with a high proportion of idiosyncratic risk, like certain hedge funds); and/or the risk factors in the scenario are less correlated with the risk factors in the portfolio (for example, evaluating a U.S. equity portfolio in a scenario driven by Japanese interest rates). Conversely, prediction errors tend to be smaller when the underlying shock is smaller, the risk model is a good fit for portfolio risk, and the factors in the scenario are more correlated with the portfolio.

What are the limitations of the analysis?

The scenario tester uses market indexes to represent portfolio exposures. The actual composition of specific investments entered vary from the market indexes used, sometimes significantly. Active returns from specific investments are not included in the analysis, and could be both positive and negative. Dividend income and yield derived from securities other than bonds is not included. Fees, taxes and transaction costs are not included and would reduce the return of any portfolio.

The analysis provided is illustrative only. Neither BlackRock nor the Aladdin portfolio risk model can predict a portfolio’s risk of loss due to, among other things, changing market conditions or other unanticipated circumstances. The Aladdin portfolio risk model is based purely on assumptions made using available data and any of its predictions are subject to change.

How did we choose the size of the shocks?

The shocks are sized to approximate a one-standard deviation move in the underlying market factor. We measure the factor’s annualized standard deviation over the past ten years, and then round to the nearest whole number (or multiple of 5, for higher levels). The purpose is to scale the shocks so that they are all approximately equally likely to happen. Assuming returns are normally-distributed, a given market factor has a 16% probability of exceeding a one standard deviation move in either direction over a one year period.

How are we converting to asset class allocations?

For products, the market index (or blend of indexes) used is selected by BlackRock based on the Morningstar Category. All products in any given Morningstar Category are represented by the same market index. For asset classes, the market index used is selected by BlackRock. The list of asset classes included is limited, and is selected by BlackRock.

Indexes are unmanaged and are shown for illustrative purposes only. It is not possible to invest directly in an index.

 

USRRMH0423U/S-2842648