Full Suite of New Combination Methodologies Released on C8 Studio
Our new suite of combination methodologies has been released.  One of the most powerful advantages of using the C8 Direct Indexing platform, C8 Studio, is the ability to combine and weight indices from multiple sources.

Importantly, the methodologies can also be used to solve other significant portfolio issues. For example, an investor has multiple indices to choose from, so just as important is the decision as to which index to select. C8 also assists with solving this problem.

The C8 platform has now been extended to include all of the following portfolio solutions:

  • User Defined Weights – investor sets their own weights for each index.
  • Risk Parity – a volatility-normalised portfolio where each component is given an equal risk-adjusted weight.
  • Clustered Risk Parity –  an unsupervised machine learning technique called Agglomerative Hierarchical Clustering (AHC) to find natural groupings in the constituent indices where each group contributes equally to the overall portfolio risk. This has the advantage over pure risk parity in that by grouping correlated assets together, risk is then more evenly distributed.
  • Tactical-Asset-Allocation – see below
  • Mean-Variance – a classic Markovitz solution that maximises risk-adjusted return by allocating more capital to outperforming indices.
  • Mixed-Integer Optimisation – see below
  • Principal Components Portfolio – classic Principal Component Analysis (PCA) which derives uncorrelated variables to define a time series. These are then combined to construct an optimal portfolio
  • Performance Factor Portfolio – see below

Tactical Asset Allocation

US Allocation Diversified Combination

On C8 Studio, there are a number of good quality US allocation programmes which aim to maximise returns, relative to risk, by switching between equities and bonds (as well as other assets).  C8 Studio gives investors the ability to easily combine a number of these disparate indices together to generate consistent and robust returns.  In this example, we combine the following indices:

  • Ned Davis Research – Smart Sector Model – weights US equity sectors using the NDR proprietary sector model, with exposure reduced in market declines.
  • JP Omega – US Cross Asset Index – uses a proprietary Johnson-Omega methodology to allocate across US equity sectors, bonds, oil, gold and cash.
  • Stargazer Asset Management – Dynamic Allocation Index – a proprietary methodology that allocates between any combination of S&P 500, US bonds and gold.
  • Alpha Vee – Risk Managed Allocation US Equity and Treasury Bond Index – which uses a proprietary methodology to select their 50 top-ranked equities along with an allocation between 20-100% in US Treasury ETFs according to their fundamental indicators
  • Smooth Sailing – US Stock and Bond Rotation Index – systematic tactical rotation between US government bonds and various equity index ETFs.

Please see C8 Studio for full details on each index (or please request login details).

The indices are then combined using Tactical Asset Allocation, a C8 proprietary method, in which, unlike mean-variance, the allocation programme gives most weight to indices that performing in line with their average. That is, the methodology underweights underperforming indices (similar to mean-variance) but, importantly, also underweights indices outperforming their typical performance. Classic mean-variance optimisation will always give the largest weight to an index just before it mean reverts.  With Tactical Asset Allocation, the weight will be reducing ahead of any correction to mean.

Whilst each index in the combination has many strengths, we believe that using them with C8’s combinational algorithms can bring added value.

Mixed Integer Optimisation

ESG Diversified Combination

The C8 Studio works well with ESG indices, as direct indexing provides full transparency of holdings pre-execution (so no unexpected surprises), customisation (ESG mandates can differ considerably) and voting rights are retained, so investors can vote their shares in line with their ESG principles. We have a number of innovative ESG providers on C8 Studio:

  • Entelligent’s Smart Climate Index which estimates the sensitivity of companies to climate change (also discussed in the March Update).
  • TSC’s US and Euro Water Security indices, that focus on a company’s water footprint, reducing the investment risk from water insecurity.
  • iClima’s indices that focus on CO2 reduction with their Decarbonisation and Renewable Energy indices.
  • LGBTQ100 focusing on the Social (S) aspect of ESG, that weights stocks by surveys of companies’ social attitudes.
  • C8/ Greenblue ‘Good Governance’ L/S index which uses natural language processing to assess governance standards of US stocks.
  • C8/SEB Sustainability L/S index which focuses on sustainability criteria for European stocks.

We use Mixed-Integer Optimisation to build an ESG portfolio across the above indices. This algorithm looks to chose between any 4 to 6 indices out of the 8 at any one time. We have set the rebalance to every three months using the previous 2 years returns (so all return streams are out-of-sample).  The performance of the combination can be seen below:

Performance Factor Portfolio

Global Equity Ex US Combination

Another good example is of a larger combination problem, taking 21 non-US country market-cap weighted equity indices from Solactive, then selecting between 5 and 10 countries to track for the next year. The Performance Factor Algorithm takes a different approach to Mixed Integer Optimisation, which uses mean-variance and investigates the best overall combination on that basis. With Performance Factor, each index is scored by its performance using classic metrics such as Sharpe Ratio (average return relative to volatility), Sortino Ratio (average downside return to volatility), Calmar Ratio (average return to max drawdown), as well as other more esoteric performance measures.

This algorithm creates a portfolio that, in terms of Sharpe, performs as well as the best performing equity market, but also substantially reduces the complexity of tracking 21 equity markets.

Walk-Forward Analysis

One key element in all the combination methods discussed above is that these were all used in conjunction with ‘Walk-Forward’ analysis.   In C8 Studio:

This gives a more realistic view of potential real-life performance, as the performance is calculated ‘out-of-sample’ ie avoiding the risk of ‘curve fitting’ (where the portfolio selection and weights is based on a known past performance history).  The figure below illustrates how this works, only the out-of-sample periods (blue) are used in performance measurement:

C8 Continues to Expand

David Jervis joins as Global Client Officer for C8 Technologies to take responsibility for the firm’s client relationships, index and investor partners alike.

David has worked in global capital markets for the past thirty-five years, his prior role being Head of Senior International Relationship Management at Bank of America.

Previous management roles for Bank of America and Merrill Lynch over twenty-eight years included Head of Wealth Management in EMEA, Head of Equity Sales in EMEA, Chair of Merrill Lynch Capital Markets in France, Head of CEEMEA Equity Sales and Co-Head of Japanese Equity Sales in Tokyo.

Before that, David worked at Baring Securities, Kleinwort Benson and New Japan Securities, always on institutional investor clients.

Thanks for your interest,

The C8 Team

Other posts

NDR Fixed Income Allocation Strategy January 2024 Update

BY BRIAN SANBORN
The NDR Fixed Income Allocation Strategy, Positioning Update Read more →

NDR Dynamic Allocation Strategy October 2023 Update

BY BRIAN SANBORN
Dynamic Allocation Strategy, indicators, weightings update Read more →

NDR Dynamic Allocation Strategy April 2024 Update

BY BRIAN SANBORN
Dynamic Allocation Strategy, indicators, weightings update Read more →
Back to all posts →