2/2/2026, 9:00am
What do the 12 strategies we developed this summer and blogged about here say as we head into February 2026?
Before we share that let us first note that the strategies utilize volatility constraints ranging from 10% to 20%, and ticker concentration constraints ranging from 3% to 10%. VecViz has no idea where your risk tolerance fits along that spectrum (or beyond it). For that and many other reasons, especially the fact that VecViz is not an SEC registered Investment Advisor, and also the fact that VecViz staff has exposure in their personal investment portfolios to some of the tickers discussed below, this is not Investment Advice or and Investment Recommendation. See our terms and disclosures (linked to at the bottom of this page), for more on this important issue.
As you may recall from the blog, we utilize mean variance optimization (MVO) for both the Vector Model and Sigma. The inputs to that analysis are expected return, ticker level volatility, and ticker correlation.
- For the Vector Model we utilize a VaR Breakage Regime based measure of expected return, the Vector Model 99th % tile ticker level VaR for ticker volatilty (scaled to a ~ 1 sigma annualized basis), and for correlation we have two techniques: one that considers VecEvent similarity (S10), and one that looks at the correlation of each ticker’s many VecViz analytic attributes, standardized for comparability. We call this latter method VecViz “fingerprint” correlation (S14). Note that we have not yet updated our VecEvent correlation metrics for the recently updated, Gemini 3.0 Pro based VecEvents currently embedded in our Dashboards.
- For Sigma (S1) we base all inputs on the trailing 252 trading days, with the expected return set simply to the average daily return over that period, the ticker volatility set simply to the annualized standard deviation of daily returns over that period, and the correlation set to the Pearson correlation of daily returns over that period.
- Both the Vector Model and Sigma MVO processes are subject to the same constraints on expected portfolio volatility (10%, 15%, 20%, averaging 15%)1 and max ticker weighting (3%, 6%, 10%, averaging 6.33%). They were also both limited to VecViz’s ticker coverage universe to draw tickers from (see our dashboards for the full list).
We present below the top 30 exposure for the VecViz average portfolio on the left (in blue), with corresponding Sigma exposure listed alongside them. On the right we present the top 30 Sigma exposures, with corresponding Vector Model exposures alongside them, as well.
Biggest Vector Model vs Sigma ticker weighting differences:
The Sigma portfolios are more exposure to Gold and related names and the Vector Model portfolio has more exposure to tech (defined broadly, including Apple, MSI, GOOGL, WDC, CDNS, MSFT, META, etc).
We will aim to refresh these weights monthly.