Financial, Risk and Portfolio Analysis
“How close is your budget plan to the 'Efficient Frontier'?” That’s a question that can keep a planning team working late.
A portfolio is a set of possible funding opportunities. Portfolio Analysis is the measurement of how each portfolio performs according to many business metrics. For each of many portfolios, plot the measurements of portfolio performance in two or three dimensions. A scatter plot or cloud of points shows the tradeoffs
The Efficient Frontier is the edge of the cloud containing the best portfolios
Each of the portfolios tested can be evaluated for their range of uncertainties. The chart at left shows the NPV at different confidence levels for as a line for each portfolio. The points highlighted in yellow in the NPV vs Drill Cost chart above are highlighted in red here.
Knowing which investment opportunities are selected often in the best portfolios is a key understanding to be gained in Portfolio Analysis. The spiky chart below shows the funding level (y) of each project in the portfolio. The three projects selected above are highlighted in red below. Some projects (133, 146, 173, 185, etc) are selected in all three of the portfolios. Many other projects are selected in none (135-137, 152-171, etc).
These
charts were created with Spotfire® Decision Site using data processed
by a WiserWays WhatIF Portfolio Builder.
There are methods to “optimize” portfolios using variations of Linear Programming and Mixed Integer Linear Programing. These techniques will find theoretical best investment decisions, but invariably they tend to deliver solutions that are impractical from operational, political or other reasons that are hard to quantify. Human decision makers are can manage many unquantifiable or difficult to quantify factors that must be considered in a Portfolio Analysis. You can put great effort into defining every business factor that must be considered into an LP model.
We at WiserWays believe a wiser approach is to compare real candidate portfolios,
any of which would be considered practical to management, and explore where
they fall with respect to many efficient frontiers. LP optimization models
find portfolios exactly on the efficient frontier. But, there are countless
portfolios that are close enough to the frontier that are also better from
operational and strategy issues. It is analyzing portfolios near by the frontier
that the above graphical analysis has advantages over the black-box optimization
models.
For further demonstrations of this technique, see:
"Faster and Better Exploration at ChevronTexaco with Spotfire" by Stephen
M. Rasey, Spotfire 2002 US Users Conference, April 30, 2002, Philadelphia,
PA, http://www.spotfire.com/images/pdf/presentations2002/Stephen_Rasey_Analytics_for_Oil_Gas_breakout.pdf,