Wizard of Odds

Asset Allocation Tool

The Wizard of Odds is an asset allocation tool that enables investors to replicate a portfolio to analyze its historical performance and run Monte Carlo simulations to identify areas for improvement.

The tool is a simplified version of SOCIO, which reports on more granular data and lets investors understand the implications and possibilities of integrating RQSI’s alternative strategies alongside existing allocations.

Does your portfolio meet expectations?

The Methodology

All institutions are in a constant state of portfolio risk. While the specific sources of risk vary based on a portfolio’s allocations, instruments and managers, it is widely accepted that asset allocation is a significantly greater source of portfolio variability. For that reason, we chose to focus on asset allocation as the variable we attempt to model.

Asset definitions are not universal, so the first step in modeling the variability of asset allocation is to control for the nominal differences in asset definitions across institutions. To do so, we reviewed dozens of institutional investment policy statements to understand the entire universe of asset classes, illustrated below by three real (though anonymized) institutions.

Institution A
Large Public Endowment
Public Equity 29%
Long/Short Equity 20%
Private Equity 19%
Real Estate 5%
Resources 5%
Marketable Alternatives & Credit 12%
Fixed Income & Cash 10%
Institution B
Foundation Average
Global Equities 45%
Fixed Income 15%
Cash 5%
Hedge Funds 20%
Private Equity 10%
Real Assets 5%
Institution C
Public City Employee Pension
Broad US Equity 33%
Broad International Equity 19%
Fixed Income 26%
Real Estate 7%
Private Equity 5%
Real Return 5%
Hedge Funds 5%
Cash Equivalents 1%

The Venn diagram below shows the similarities and differences of asset class definitions across all three institutions. Our definitions of asset classes may not be all-encompassing or omniscient, but they are segmented thoughtfully enough to successfully model an institutional portfolio to uncover issues therein.

Institution A
Long/Short Equity is its own asset class in this particular IPS, which is uncommon. Upon closer examination, the components of this asset class are more similar to the more common conception of Hedge Funds. Within our tool, Long/Short is available as a subcategory under Hedge Funds.
Marketable Alternatives and Credit appears to be a reference to high yield government bonds and is not seen in any of the other IPSs we’ve evaluated. Within our tool, High Yield Bonds are available as a subcategory under Fixed Income.
Resources are given their own classification, which we have seen in other IPSs. Within our tool, Natural Resources are available as a subcategory under Real Asset.
Cash is listed as opposed to Cash Equivalents. Across the IPSs we have seen, the tie that binds appears to be the classification of T-bills as Cash. Every allocation to Cash or Cash Equivalents within the tool is assigned as T-bills.
Institution B
The IPS definitions are very generalized. Within our tool, asset subcategories exist to help make more granular determinations.
Real Estate is listed as an asset class, which we view as a subset of Real Assets.
Public Equity, Private Equity, Fixed Income, Real Assets, Cash all appear in one form another.
Cash is distinguished from Fixed Income; this distinction was kept.
Hedge Funds have their own category. Many IPSs refuse to list hedge funds and often fallaciously categorize these allocations as Real Return, however, we believe Hedge Funds deserve their own asset class.
Institution C
Real Return is distinguished from Real Estate. Upon further investigation, some of Institution C’s Real Return items are real assets like MLPs or Timber, but others are more similar to hedge fund allocations. In cases of conflicting definitions, we have to make an assumption about the breakdown of the choices.
Broad US Equity is distinguished from Broad International Equity; this distinction was kept.
The following are relevant features of IPSs that are not present in either of the institutions used in our examples, but are covered as categories or subcategories within our tool:
Private Credit (or similar) is an independent line item. The prevalence of this asset class is growing rapidly and its return profile is distinct from both Public Fixed Income and Private Equity.
Venture Capital and Leveraged Buyouts are distinguished from Private Equity in some IPSs. We don’t believe this distinction is important enough for us to make.
Emerging Market Equity is distinguished from International Equity. Though none of the three institutions shown make this distinction, many others do. We view these assets as distinct enough that they should be separated.

The Proof

The following tables and graphs present the performance figures of the three anonymized institutional portfolios compared to the predicted performance simulated by our framework. All IPS replications are reproduced in a reasonably accurate manner, though the framework will rarely ever reach 100% accuracy due to variances related to instrument, manager selection, data availability, our desire for universalizability, and changes in asset allocation that institutions make across time.

Institution A
Large Public Endowment
Annualized Return 9.7% 10.8%
Annualized Volatility 6.5% 6.3%
Sharpe Ratio 1.49 1.72
Institution B
Foundation Average
Annualized Return 6.2% 6.4%
Annualized Volatility 13.2% 14.0%
Sharpe Ratio 0.47 0.46
Institution C
Public City Employee Pension
Annualized Return 5.1% 5.2%
Annualized Volatility 7.6% 8.5%
Sharpe Ratio 0.67 0.61

Due to data availability issues, the time periods and periodicity of the analysis varies across all three institutions. Institution A starts in 2013, Institution B in 2008, and Institution C in 2014. Both Institution A and Institution B use annual returns; Institution C uses monthly returns.

While we don’t think three institution examples prove a rule, we believe that this level of precision is sufficient for gaining a picture of a portfolio with enough resolution to diagnose issues.

Once you have determined the desired asset allocation, the next step in the Wizard of Odds tool is to define the desired performance characteristics of the portfolio. While there are many variables to consider, we have chosen target return, volatility and max drawdown as key measurable factors that define how well a portfolio performs over time.

In the Wizard of Odds, you define these parameters as well as the percentile desired to see how closely the portfolio will perform against these criteria through Monte Carlo simulation, identifying the potential distribution of outcomes over the next 10 years. This functionality provides the ability to see how likely you are to achieve the portfolio's desired levels and how much this changes with different underlying asset allocations. We can also provide a more detailed summary report showing possible enhancements to the portfolio that may increase the likelihood of achieving the desired goals.

With the ability to design a detailed asset allocation and apply thoughtful portfolio performance criteria, the Wizard of Odds takes this framework and distills it into an interactive tool for helping investors uncover potential concerns within their portfolios.

For investors interested in integrating alternative strategies that complement existing allocations, our more comprehensive SOCIO application provides the same in-depth portfolio analysis as Wizard of Odds, but with more greater performance detail, backtesting of our customizable multi-asset products, and additional features built with institutional investors in mind.