Risk premium forecast: main asset classes October 4, 2021


The Global Markets Index (GMI) fell in September compared to the , dipping to 5.9% and reversing an upward trend in recent months. The new estimate reflects a long-term forecast of performance relative to the “risk-free” rate, using a risk-based model (details below).

GMI is an unmanaged, market-weighted portfolio that holds everything (except cash) and represents a theoretical benchmark of the optimal portfolio for the average investor with an infinite time horizon. In other words, the GMI is useful as a starting point for research on asset allocation and portfolio design. Indeed, GMI’s history suggests that the performance of this passive benchmark is competitive with active asset allocation strategies on the whole, particularly after adjusting for risk, trading costs and taxes.

Using short-term and medium-term reverting market factors (defined below) to adjust the forecast reduces GMI’s ex ante risk premium to an annualized rate of 5.2%.

Expected risk premiums vs followers

All forecasts are expected to be flawed to some extent, although GMI’s projections are likely to be somewhat more reliable than the estimates for individual asset classes shown in the table above. Forecasts for market components are subject to greater uncertainty compared to aggregation forecasts, a process that can cancel out some of the errors over time.

For a bit of a historical perspective on how GMI’s realized risk premium has fluctuated, consider the results on a 10-year annualized basis. The chart below compares GMI’s risk premiums against the equivalent for US stocks (Russell 3000) and US bonds (Bloomberg Aggregate Bond) up to last month.

GMI’s current 10-year performance (red line) is currently 8.4%. This is close to the previous peak of a few years ago. It is also well above the current long-term projection, suggesting that investors should manage expectations for multi-asset class portfolios lower from the all-time high of the past decade.

Annualized historical risk premiums over 10 rolling years

Annualized historical risk premiums over 10 rolling years

Now let’s review the methodology and rationale for the above estimates. The basic idea is to reverse engineer the expected return, based on risk assumptions. Rather than trying to predict return directly, this approach relies on the moderately more reliable model of using risk measures to estimate the performance of asset classes.

The process is relatively robust in that it is slightly easier to forecast risk than to project return. With the necessary data in hand, we can calculate the implicit risk premiums with the following inputs:

  • an estimate of the expected market price of GMI risk, defined as the Sharpe ratio, which is the ratio of risk premiums to volatility (standard deviation)
  • the expected volatility (standard deviation) of each asset
  • the expected correlation for each asset with the overall portfolio (GMI)

The estimates are taken from the history since the end of 1997 and are presented as a first approximation to model the future. The projected premium for each asset class is calculated as the product of the three entries above. GMI’s ex ante risk premiums are calculated as the market value weighted sum of individual projections for asset classes.

The framework for estimating equilibrium returns was first described in a 1974 article by Professor Bill Sharpe. For a more practical summary, see Gary Brinson’s explanation of the process in Chap. 3 of the portable MBA in investment. I also review the model in my book Dynamic Asset Allocation.

Here’s how Robert Litterman explains the concept of estimating equilibrium risk premia in Modern Investment Management: An Equilibrium Approach:

“We don’t need to assume that markets are always in equilibrium to find a useful equilibrium approach. On the contrary, we view the world as a complex and highly random system in which there is a constant barrage of new data. and shocks on existing valuations which as often as not move the system away from equilibrium.

“However, while we expect these shocks to constantly create deviations from the equilibrium of financial markets, and we recognize that frictions prevent these deviations from immediately disappearing, we also assume that these deviations represent opportunities.

“Savvy investors who try to take advantage of these opportunities take actions that create the forces that continually pull the system back into equilibrium. Thus, we view financial markets as having a center of gravity defined by the supply balance. and demand.

“Understanding the nature of this equilibrium helps us understand the financial markets because they are constantly shocked and then pushed back into this equilibrium.”

The adjusted risk premia estimates in the table above reflect changes based on two factors: short-term dynamics and long-term average reversion. Momentum is defined here as the current price relative to the moving average of the past 10 months. The average reversion factor is estimated as the current price relative to the moving average of the last 36 months.

The raw risk premium estimates are adjusted based on current prices relative to 10-month and 36-month moving averages. If current prices are higher (lower) than moving averages, estimates of unadjusted risk premiums are lowered (raised). The adjustment formula simply takes the inverse of the average of the current price at the two moving averages as a signal to change the projections.

For example: if the current price of an asset class is 10% above its 10-month moving average and 20% above its 36-month moving average, the risk premium estimate is not adjusted is reduced by 15% (the average of 10% and 20%).

What can you do with the forecasts in the table above? You can start by asking yourself whether the expected risk premiums are satisfactory… or not. If the estimates are below your required return, you may want to consider creating a higher rate of performance by customizing the asset allocation and rebalancing rules.

Keep in mind that GMI’s gross implicit risk premiums are based on a weighted combination of the unmanaged market value of major asset classes. In theory, this is the optimal asset allocation for the average investor with an infinite time horizon. Unless you are a foundation or a pension fund, this time horizon assumption is not practical and therefore there are reasonable arguments for a) changing Mr. Market’s asset allocation to suit your needs. special needs and your risk budget; and b) add a rebalancing component to your investment strategy.

You can also estimate risk premiums with alternative methodologies for additional information about the near-term future (a great resource on this topic: Antti Ilmanen’s expected returns).

For example, let’s say you have confidence in the Dividend Discounting Model (DDM) to predict the performance of the stock market over the next 3-5 years. After calculating the numbers, you find that DDM tells you that the expected performance of the stock market will differ significantly from the estimate based on long-term equilibrium. In this case, you have some tactical information to consider.

Also keep in mind that combining forecasts through multiple models can provide a more reliable set of forecasts than estimates from any model. Indeed, a number of studies published over the years document that combined forecasts tend to be more robust than projections from a single model.

What you can’t do is spill blood from a stone. No one really knows what the risk premiums will be in the months and years to come, so relying solely on forecasts (especially for the short-term future) is a problem. In other words, you need to deviate from Mr. Market’s asset allocation with care, thought, and for reasons other than assuming that you are smarter than everyone else (i.e. the market).


Leave A Reply