Asset Allocation Models

Our asset allocation models use investment asset class return assumptions with asymmetric (non-normal) probability distributions. Since the long-term expected return of equities is near 10% and the standard deviation is near +/- 16%, one might expect that 96% of the time returns should fall between -22% and + 42%. While the S&P 500 came close to the top of that range in 1997 it also fell to the low end in 2002. Interestingly the S&P 500 managed to come close to the top once again in 2013, while more importantly dropping significantly below the ‘expected’ low return in 2008. Our models, like the actual stock market, include a higher than 2% chance that the stock market could fall more than 2 standard deviations.

The goal of these simulation models is to seek an asset allocation which produces the highest total return with the lowest likelihood of producing a negative return by simulating uncertainty in the financial markets. Although this analysis is being used to show how pension plans and foundation endowments can better address their cash flow requirements by being invested in a diversified portfolio of stocks, bonds, and alternative investments, these lessons can also be used by individual investors. Our analysis shows us what could happen to an investment portfolio over a 5 year (or longer) time horizon in the greater than 1 in 50 chance that negative investment returns during some part of that 5 year period were worse than even the “Worst Case Scenario” of a typical analysis. What we find is an asset allocation properly balanced between stocks, bonds, and alternative investments should produce better returns with significantly reduced overall risk in an investment portfolio compared to a more conservative approach or the typical 60% stocks and 40% bonds portfolio. The use of Dynamic Rebalancing can improve the risk adjusted returns even more.

Using the “building blocks” approach, expected return characteristics for each asset class are developed, based on its historical pattern of returns relative to the risk free rate. On the other hand, we believe that there is strong logic to support estimations of future standard deviations and cross-correlations by basing those estimates upon their long term histories. Individual asset class return assumptions are thus driven by the current risk free rate, as well as the projected risk free rate of return over the investment time horizon. For example, our model’s recent “starting block” for constructing expected returns used a 2-3% risk free rate, determined as of the beginning of 2008. By comparison, the long term historical risk-free rate has been 6-7%.In summary, our investment asset class assumption sets have a strong forecasting bias in favor of assumed central tendency, adjusted for the current cyclical stage at the time a portfolio model is created. The result is an internally consistent set of assumptions that run in a real-world modeling context to produce forecasted outcomes that are not only robust, but also practical for an average client decision-maker to understand and consider.

For more details on our Asset Allocation Process, read how we use @Risk for Asset Allocation Modeling.

The traditional approach to asset allocation has been to use Efficient Frontier models that seek to find the optimal portfolio mix that has the lowest possible level of risk (standard deviation) for its level of return (mean). When all asset class returns are assumed to follow the normal distribution, an efficient frontier model will yield risk vs. return results that are consistent with our simulation model over a one-year time frame. While this approach provides useful results, it still leaves many questions unanswered. This is where tools like @Risk and Monte Carlo simulation modeling come into play.

An obvious advantage to using @Risk and simulation modeling in preference to the simpler optimization techniques of the efficient frontier model is that we can realistically incorporate the effect of time horizons longer than one year. This is important because: (1) If volatile assets (e.g., stocks) are included in the model, then one-year modeling output has little chance of realism; (2) Consequently, users of modeling output that includes stocks and other volatile assets are forced to consider time horizons of at least 3-5 years; (3) @Risk’s probabilistic modeling output demonstrates to model users how multi-year time horizons increase the beneficial diversification effects of adding asset classes to the portfolio mix.