Friday marked the single largest drop in US Equity markets since September 21st, 2011 when the Federal Open Market Committee announced the implementation of Operation Twist.

Over the weekend we've seen fairly consistent rationale that aims to add contextual insight to the event. Although spirits aren't entirely restored through this sort of information, we mend incrementally as our loss no longer remains without reason.

At Viziphi we argue, however, that although qualitative insight plays deeply to our fundamental need to understand, it poorly serves as a framework to navigate market environments visa vie portfolio construction and rebalancing. The first step, instead of exploring rational market context, should be evaluating and estimating current portfolio dynamics that incorporate the recent event in question.

For example, if a professional reads an article that states the current selloff "was due to heightened fears about the Federal Reserve raising interest rates," it's difficult to understand how that information can be used to drive allocation decisions for clients. A better way (we believe) is to first analyze portfolio dynamics, see how they have evolved given new market information, drive allocation recommendations using those analytics, and then use qualitative information to buoy your insights should you choose.

Its sounds much more complicated than it is, now that ViziPheed isn't far away. In this brief recap we leverage our new tool to demonstrate how to extract insights from our dead-simple visualizations and turn them into actionable portfolio recommendations using Friday's market correction as a case study.

Note about Asset Classes & Transparency

Asset Classes

In the visualizations below, you'll notice we intentionally use asset classes instead of tickers. In short, analytics and attribution of ETF data are highly desirable for a couple of reasons:

  1. Given the ETFs are closely screened for liquidity constraints, they can be an invaluable source of information regarding market undercurrents and dynamic factors
  2. They represent a viable investment solution -- when compared to indexes and composites, which by design are un-investable. This eliminates the chance for confounding factors that can occur when trade decisions are made on data derived from something other than the investable asset

Here are the list of tickers used to implement the broadly diversified asset classes we analyze below:

Asset Class Ticker
US Fixed Income BND
Foreign Fixed Income BWX
Commodities DBC
US Equity IWV
US REIT IYR
Foreign Equity VEU

In ViziPheed, we use transparency fill to indicate passage of time. The more transparent a color is, the further in the past it is relative to the other data points in the set. Here's a generic legend to clarify this point.


The Weighting Allocation

Instead of a bottom up (bottom-less?) method leaving the reader to interpret how their clients might be impacted by the barrage of market commentary, we start at the core: your portfolio.

The following graphic illustrates:

  • The Viziphi Moderate Aggressive model portfolio on Friday, August 21st,
  • The weights it rebalances to on a monthly basis (where the last rebalance date was August 1st, 2015), and
  • the amount of drift that took place between the two time periods

With less than a month between dates, there's very little drift. However, it gives the user a quick visual reference to understand how they are tactically weighted relative to an investment model, and the insights around the allocation recommendations in the summary will specifically leverage this information.

Moderate Aggressive Portfolio Allocation


Sources of Risk

One of the most valuable pieces of information we can understand on a day-to-day basis is the sources of risk within our portfolio. At Viziphi, we estimate a "worst case scenario," both for a portfolio and for its underlying assets (or asset classes). We believe that by using measures that quantify worst-case scenarios, investors are more mentally prepared for market dislocations which means relationships between professionals and their clients bear less stress during times of persistent decline.

Also when looking at risks of extreme events (another name for "worst case scenario"), it's important to be able to explore time variation. In this first part of the analysis, we look at three points in time:

  • Friday, August 21st, 2015
  • The prior day (August 20th, 2015)
  • One month prior (July 21st, 2015)


Extreme loss in the above chart refers to the "worst case scenario" of the asset in isolation (e.g. not taking covariation into account), based on that day's information, where a higher value indicates more risk, i.e. more loss.

With just a simple bar chart, we're able to glean a tremendous amount of information, all of which is highly useful if we were considering portfolio rebalance or reallocation.

  • Extreme risk in US Fixed Income has continued to decline, albeit marginally, over the past month
  • Foreign Equity has been, and remains, the single riskiest asset in the portfolio (when not accounting for co-movement)
  • Extreme risk in US Equity has continued to increase, and now is the second riskiest asset class behind Foreign Equity

Contribution to Extreme Loss ("CEL")

In the world of portfolio management, history is bifurcated -- the days before Modern Portfolio Theory ("MPT") and those after. Of course MPT has its shortcomings and points of contention, but never before was aggregate portfolio risk decomposed into how assets moved in tandem, also known as their covariation.

Therefore, when looking at portfolio risk, using some measure of co-movement to understand how each asset contributes to aggregate risk is not just important, it's imperative.

The chart below shows each asset class's contribution to extreme risk at the same three dates:

  • Friday, August 21st, 2015
  • The prior day (August 20th, 2015)
  • One month prior (July 21st, 2015)

Contribution to Extreme Loss


Again, the insights that can be gleaned and how they might translate to portfolio construction are all too clear:

  • Foreign Equity CEL has been on the decline and is now actually 1% less than US Equity contributions (a fact gleaned in the mouse-over functionality of the dashboard)
  • Foreign Fixed Income as a broad asset class served as a more powerful diversifier (more negative CEL) than US Fixed Income in Friday's rout
  • Commodities, although qualitatively viewed in many articles as one of the main contributors to the decline, in fact has lower CEL than it did a month ago

Summary & Implications on Investment

If you showed your client the first chart of the portfolio allocation and asset drift, you might opt to play it safe and simply rebalance at the end of the month. However, given the information we've just gleaned from a two-minute analysis on ViziPheed (our Model Portfolios come pre-built as one of the landing screens), your recommendations could easily take the following form:

  • US Fixed Income and Foreign Fixed Income, although overweight relative to the Moderate Aggressive strategic allocation are the only two assets classes in the portfolio that are dampening volatility
  • Foreign Equity, although more underweight than any other asset class, bears the highest risk of any asset class
  • US Equity, although the second most underweight asset class, has been continually increasing its risk profile, both on an absolute and contribution-based basis

Note how easily the above reasons translate to "staying put," instead of rebalancing to the strategic allocation. Instead of attempting to translate the qualitative straw-man into actionable information to communicate to your clients, financial professionals should draw insights from actual portfolio analytics on the actual portfolios they're monitoring, thereby streamlining value creation for clients and improving portfolio construction workflow.

Next Steps

The next steps, all of which would be possible on ViziPheed but would have resulted in a laboriously long post, would be:

  • Drill down into different asset classes to explore sub-class sources of risk and how they have changed over time
  • Look at analytical profiles of 9/21/2011 and this past Friday and see how similar they are
  • Create your own backtest and portfolio that showcases your own insights
 

Comment