Risk Management Framework

080730-N-5277R-003

With ever evolving markets, a risk management framework which can match a desired risk profile with a risk solution has become a necessity. The Orpheus patented risk management framework is a data innovation methodology that offers a risk management framework to rank, select, identify, and allocate risk.

Why do we need a risk management framework?

Post 2008 volatility is the new normal.
Capital conservation is essential now.
Drawdown comfort needs to be redefined.
Need for scalable and customizable systems.
Need to sift through deluge of information.
Need for peer outperformance, beating respective benchmarks.

Selection from a “Universe”

Before allocation we need systems that can assist in selection.
Selection can be for manager, asset, region, risk, profile, strategy, other constraints (example liquidity).
Selection can also be for component ETFs from a large choice of passive Universe ETFs based on S&PDJIndices1.
Owing to the large available choice, the idea of “what is my universe?” is unclear.
And even if we know what our universe2 is (e.g. DJ Sector Indices) re-characterizing the universe can improve selection.
Today popularity of a universe is more important than the statistical character of the universe.
Example an absolute performance ranking can be strikingly different from a relative strength3 ranking on a universe. (Illustration from Jan 2012).
The size of the universe can also change the ranking of respective components.
This means a new universe could mean different rankings and potentially a different selection.
Hence it becomes paramount to define universe based on client utility before asset selection.

1

Mean Reversion: Reality and Failure

Though mean reversion4 failure is a market reality, the investment management business relies on mean reversion directly (reversion strategy), indirectly (momentum). Behavioral finance showcased that a worst losers portfolio outperformed the best winners portfolio5. The subject used mean reversion to challenge conventional economics and prove markets were not random. The statistics prove that 90% of fund managers fail to beat the market consistently6. Hence understanding mean reversion, its reality and failures is important for investment managers.

6

Data Innovation: Extreme Reversion

The Orpheus Risk Management Framework (ORMF) is based on the ideas of Extreme Reversion7 and Dynamic mean8 .
Mean reversion is based on the idea of a statistical mean9 , which has a few disadvantages.
It is sensitive to extreme values; is suitable for a certain time series; and works when allocation weights are equal .
However market performance varies over multiple time periods10  and investing also requires unequal weights.
Dynamic mean accommodates for all the weaknesses of an absolute average mean (Statistical Mean).
Dynamic mean based on group (assets), which can  be large, small, equal (unequal) weighted, with (without) outliers11.
Dynamic mean is applicable to multiple investment time frames.
Dynamic mean can change as the defined group (universe or sub-universe) changes or is redefined.
Markets are like extreme groups regressing to a dynamic mean, rather than moving back to an absolute mean11 .
Extreme reversion looks at markets like a chaos12  systems, as a flock of birds13 that keeps moving.
In this Orpheus ranking innovation a percentile ranking from 0 to 100 is assigned to an asset in a group14.
And just like chaos systems these rankings are dynamic and create various seasonal patterns of growth and decay.

Momentum vs. Reversion

Momentum15 and Reversion are two key investing styles which suggests investing in either winners or in losers16
The ORMF Extreme Reversion data innovation approach combines momentum and reversion
Reversion is used for selection and risk management while momentum is used for growth.
Below we have illustrated17 how reversion and momentum based investing approaches can overlap.

5

Filters and Customizations

Rankings are filtered for reversion ideas
These reversion ideas are filtered for potential signals.
These signals are filtered for noise and quality.
The selection is finally used for allocation.
Volume, Returns, Risk and even Fundamentals can be used as filters.
The selected allocation models are the Orpheus Risk Management Indices ORMI ® 

4

Risk Management Styles

ACTIVE STYLE is a selection system for periodical entry and exits like the ORMI US 30, ORMI Toronto 15 or ORMI Fixed Income 10. The difference between the various selection models is the underlying universe. One may select components from the S&P 500 Equity SPDRS while the other may look at Fixed Income ETFS from North American Region. Active styles are cash conserving, absolute return Indexed models. They actively enter and exit a position and go cash if needed.

THE WORST 20 STYLE is about selecting the worst components from top 100 Universe (USA, Canada, SPDRS, ETS, S&PDJ Indices etc.). This is a quarterly rebalanced portfolio and is more about relative performance vs. the underlying top 100. This style is not a cash conserving absolute return model, but about beating its respective peer universe. Because of the idea of negative outliers outperforming, the worst 20 style outperforms the universe.

THE EXTREME REVERSION STYLE selects and allocates more in the negative outliers and less in the best outliers among a selected universe. For example DJ Sector Indices like Banking, Auto, Energy, Health Care etc.

THE RELATIVE PERFORMANCE STYLE selects and allocates proportionally more on relatively stronger components and less on weaker components in the respective universe.

THE TACTICAL STYLE works around a macro portfolio that needs to allocate across various asset classes like Fixed Income, Commodity, Forex or Equity. Tactical styles can work around a group risk level.

2

(1) http://www.spindices.com
(2) A broad group of assets used for selection.
(3) Based on price performance in Jan 2013.
(4) Mean Reversion

(5) DeBondt and Thaler 1981
(6) The Selection and Termination of Investment Management Firms by Plan Sponsors

(7) Extreme Reversion is a term that explains mean reversion tendency in group extremes or outliers.
(8) Dynamic Mean is an Orpheus Data Innovation, which assigns a percentile mean value which changes with the movement in the group.
(9) Absolute or statistical Average

(10) Multiple time periods example is monthly, quarterly, annual performance.
(11) Outliers are extreme values prone to reversion

(12) Chaos

(13) Flock of birds is a dynamical system.
(14) Mean Reversion Indexing

(15) Momentum Investing

(16) Running Winners Cutting Losers

(17) Worst 20, Relative Performance, Extreme Reversion are ORMF Styles of Selection.

3

 

7

 

8

 

9

10

 


Post a comment