On New Year’s Eve we resolved to fill a gap universally recognized by global macro investors, if not by name-brand index providers (Dow Jones, FTSE, Standard & Poor’s, Barclays, and especially MSCI).

Today, we introduce a global, multi-asset category tracker. The Stracia™ Globally Optimized Capital Markets Index™ (GOCMX) measures the blended performance of equities, fixed-income securities, commodities, real estate, and currencies on a global basis.

Tandem Index Mundanus

The GOCMX is not only the world’s only index of the global capital markets, but as far as we know, the first and only optimized index of any kind, ever.

The GOCMX took months to craft: We completed the design and specification work last December, having already begun to collect and validate the underlying data, while simultaneously designing the algorithm by which the index obtains optimal allocation solutions.

We wrote the software that implements the optimizer algorithm in-house, as well as all of the supporting modules — data collectors and aggregators, time-series rationalizers, the complete statistical package. Then we designed the graphics package, including those exhibits that appear in this post and which we will update from time to time.

GOCMX Returns

GOCMX Volatility

GOCMX Reward-to-Risk

Pricing

Access to the data itself, as well as certain graphical exhibits, are available for non-commercial use according to the following pricing plans:

• Weekly updates: $80 per year (updates published each weekend)
• Monthly updates: $60 per year (updates last weekend of every month)
• Quarterly updates: $40 per year (updates last weekend of every quarter)

Send us an e-mail if you are interested in one of these plans. We can deliver the data by e-mail or make it available for download from the site, in Excel 2003 or 2007 formats, and/or CSV.

Each plan includes not just the headline index data, but the specific allocations to each asset class over the life of the index (the five sub-indexes; see graphical exhibits at bottom). We will probably raise prices in the future, but any customer who signs up now can lock in these rates for one year.

The Data: Global Coverage

Time-series data has been obtained from a variety of sources representing the global asset categories referenced above. Not only is the reliability of the time series itself an important consideration in the selection of data inputs, but the ability to obtain the data for multiple pricing sources and/or from various distribution channels is of profound concern where international securities are involved.

We believe that the data sources used in the construction of the GOCMX represent 98% to 99% of publicly traded global stocks; 94% to 95% of public fixed-rate, investment-grade U.S., European, and Asian debt; and a broad exposure to oil & gas futures, precious and industrial metals, and agricultural commodities (grains, livestock, etc.), global (especially developed-world), size-variable real estate securities, and hard currencies.

The GOCMX has been indexed to a value of 100 beginning on 3 January 2000 (the first trading day of that year).

The Constraints

The GOCMX algorithm performs a constrained optimization function, the goal of which is to maximize the return-to-volatility ratio of the five asset-class portfolio, subject to certain constraints. First, of course, the resulting allocations must sum to 100% (we are toying with the idea of offering an optimized long/short index in the future). But if you leave the problem further unconstrained, you will quickly find yourself staring at a model that would have recommended a 47% allocation to real estate in 2004, for example. (This is not a hypothetical example. As a feasibility test, we ran the model unconstrained.) Such a recommendation stares back at you, straightfaced and unblinking, and silently dares you to “drop trow.”

So in investment allocation, constraints come with the territory. But they are not without problems: our contraints may be different than yours, and neither of us may have the right ones. We built the following parameters into our algorithm:

What this means is that the model is constrained against, prevented from, ever recommending a portfolio allocation strategy that would have us more than 65% invested in stocks, or let us have less than a 15% allocation to bonds, and so on. We based these parameters on a survey of the literature and of wealth-advisory practitioners’ actual recommendations, our own experience, and common sense.

Should the maximum allocation to bonds be set at 35%? Or the minimum allocation to commodities at 8%? Perhaps. We could count these angels all day and crack the egg on either side. But we feel the constraints as set capture the maximum and minimum tolerances that a sophisticated investor would accept, given investment conditions that would result in allocating to these boundaries.

A Rolling Picture of Optimallocation

A kineograph or “flip book” is a collection of photographs that, when viewed in quick succession, one after the other, create the illusion of a moving image.
Flip bookAs everyone knows, a movie or “motion picture” operates by the same principle: a collection of still frames give way one to the next such that we mistake the succession for motion itself. Maybe you have seen drop-card viewers in the old penny arcades. The Film Forum, in SoHo, has one (inset). You drop your quarter in and an electronic “flicker book,” as Baines calls it, treats you to an optical illusion created by vision persistence, sometimes in 3-D.

By decomposing the GOCMX to its constituents — the five primary asset classes — we can achieve a similar effect but for allocation strategy. It is like a moving snapshot that captures the ever-changing trade-offs, the precise debits and credits among asset classes that yield, empirically, the greatest risk-adjusted return possible at every point in time. We think that the ability to see how optimal allocations evolve over time, as in the exhibits below, is so much more valuable than the latest “trust-me” recommendation or periodic advisory, compliments of an individual investor’s brokerage firm or mutual-fund company, to rebalance.Those are mere snapshots, and as such cannot provide the depth of field and focus of a continuous, objective, empirical time series, much less one that works to maximize reward per unit of risk. And our metrics come with historical, risk-adjusted performance records (see above) to boot.

GOCMX Equity

GOCMX Bonds

GOCMX Commodities

GOCMX Real estate

GOCMX Currencies

And the index is always right — if we define what’s “right” as that set of allocation decisions which, with the benefit of hindsight, would have resulted in the lowest risk-to-reward ratio over the previous twelve months, up to and including the date of calculation. Hazzah.

We will explore various aspects of the index in due course. Global investment allocation is the primary objective of this site, and the GOCMX is as powerful a tool for exploring that objective as we can currently envision.

In the meantime, do you have a question about the GOCMX? Would you like more information about accessing the data, or on the pricing plans? Is there an angle from which you’d like to explore allocation theory, and think our index can shed some light? Send us an e-mail. ♦