Forecasting is a challenging task. We approach this endeavor with a great deal of humility.
“For superforecasters, beliefs are hypotheses to be tested, not treasures to be guarded.” -Philip E. Tetlock, Superforecasters
To form a view on asset allocation, we incorporate a forward-looking three-year view on expected returns for asset classes.
Fixed income can be simplified if we focus on bonds with longer-than-average holding period, shorter-than-average duration, and above-average credit quality. With that type of posture, current yield to maturities are a decent approximation of results.
Equities are a bit more challenging and timeline plays an important role. Here, we focus on a three-year outlook. Shorter timeframes can be heavily influenced by the emotions of the moment: sentiment and valuation. These short time frames are often referred to as a tactical view and are similar to trading strategies. Timeframes longer than three years (often five to seven years) are called a strategic view. However, projecting a strategy for an extended time such as six or seven years in the future is so difficult it is nearly useless, given the pace of change and innovation in the modern economy. Our view is that over a three-year time horizon, fundamentals are likely to be the driving force of results.
To evaluate expected returns over a three-year horizon, we utilize an Earnings and Valuation model. As the name implies, this model uses earnings and related metrics as a key input.
Key Inputs
- Corporate Profits
- Trailing and Future Consensus earnings expectations
- View of internal and external equity managers on the companies in their portfolios
- Historical and expected margins
- Sector-by-sector earnings expectations
- Outlook of U.S. economy over 3 years
Lastly, based on this research, we consider our expected earnings vs. the long-term historical trend.
Currently, we expect earnings to be slightly above the long-term trend—mostly due to higher earnings in the next 12-24 months, powered by COVID recovery in jobs and the economy, and then returning to a lower than trend rate of 5% or so. Overall, this puts us at around 7-8% for earnings expectations, which is a touch lower than the consensus view for 9-10% gains.
We then estimate the valuation level implied in current stock values and form a view of where valuations will be going forward. To do this, we also start with history—taking a long-term average of price/earnings (PE) ratios and a more recent trailing average. This provides some anchoring for where valuations ought to be. We then qualitatively assess several inputs which can move valuation higher or lower: interest rates, inflation, domestic policy, geopolitics, and sentiment. Each is evaluated in the current state of affairs, as well as our expectations for the next three years. These inputs are gauged on a scale of −10 to +10. For example, we assess inflation as a −7 (close to as bad as it can be) in the current configuration, moving to a +5 over the coming three years, as demographic and growth trends limit inflation over a longer time horizon.
The same process is followed for each item on the list—looking at historical ranges, assessing current figures, and developing an outlook. Altogether, the readings generate an adjustment to P/E ratios—up or down by a few points at maximum positive or negative for current and expected time horizons.
Overall, we evaluate the macro environment to be worse than average at present, with an expectation for improvement over our three-year time horizon. Our earnings expectations are lower than consensus estimates, though still favorable. Thus, over a three-year time frame, we look for earnings to advance at a modest rate and we expect valuations to improve from current levels. Together, this informs our view of equity gains in the coming years.
