Market Planning

Market Planning

Central economic planning, despite its practical and theoretical flaws, received a great deal of scientific and technical attention as a discipline in the 20th Century. A great deal of technical talent was devoted to developing the tools and intellectual apparatus of this discipline. Many resources went into the creation of 4-year, 6-year and even longer term economic plans with or without the aid of computers and mathematical models. In a very real sense central economic planning remains alive and well at the world’s central banks and in their attempts to manage a number of economic factors from prices to employment through the practical tools of monetary policy and theoretical models based on the Keynesian-monetarist consensus. Variations on the Phillip’s curve, NAIRU and multiple equilibrium models remain in use, despite their utter failure as empirical or predictive guides in a complex system.

Ludwig von Mises in his 1921 book “Socialism: An Economic and Sociological Analysis” was the first to logically demolish the idea of central planning and formulate the economic calculation problem as its central flaw. His student F.A. Hayek would later delve into the reasons why complex systems and spontaneous or emergent orders are so resistant to the conceited efforts of central planners.

However the failure of central planning and the complexity of prediction in market systems is by no means a reason to abandon “planning” and the strategic formulation of long term objectives and the means to achieve them. Strategic planning of smaller scope that is related not to entire economies but to families, companies or even personal goals is a reality that we see all around us. The use of models currently used can range from the simplest Discounted Cash Flow spreadsheet to smart company data bases that put central planner’s Project Cybersyn, or the fictional Multivac to shame. As Hayek brilliantly put it “the more the state plans, the more difficult planning becomes for the individual”. Conversely the smaller the scope of the top-down central plan, the farther ahead individuals and groups can set their objectives and can plan ahead to reach them.

There is a need for a theoretical framework that will allow us to develop the discipline of long term free market planning*. Current models are antiquated and distorted by the influence of econometric simplification (masquerading as complexity) born of physics-envy that so many economists suffer under. Market planning models should look to sciences dealing with complex emergent orders: biology and meteorology. A market planning model should have the flexibility, tolerance to change and space for trial and error of a gardening project or a selective breeding program. Such a plan must be able to react to changing inputs (mainly prices) by branching off or abandoning complete modules, slowing and speeding up certain branches of the plan as circumstances change. It must have feedback loops in its decision trees and trigger levels for expansion, retrenchment or even the abandonment of the entire plan when tolerance thresholds are exceeded.

I do not hope to do more here than merely outline what I believe is a very exciting field that will  require an immense amount of work to become an useful tool for individuals, companies, families... and bring the science of economics back to its rightful place as an everyday tool rather than an academic plaything. If in the 21st Century we see only a fraction of the technical effort devoted in the 20th to central planning, to this new discipline, we can expect incredible results. Market planning software could become as ubiquitous as accounting programs currently are. Market planning algorithms will also be very useful in the creation of DAOs (decentralised autonomous organizations).


A first approach

Let me set a very simple example that by no means encompasses the wide variety of possible techniques that could be applied in the context of a market plan. Reducing the scope and complexity of goals and inputs is an absolute necessity if we want to avoid falling into the same pitfalls as the central planners. The simpler the system, the easiest it is to model. On the other hand, the more relevant factors that we leave outside the model the more vulnerable the plan is to “external shocks” as by definition anything not included in the model will be an external shock. Similarly in larger systems necessary aggregations will need to be made, unavoidably losing information in exchange for the larger view.

The most important part of the market plan and its most useful result is the decision matrix, be it a simple diagram or a complex piece of software this tool must outline the different possible scenarios and provide options and strategies to follow within the plan’s timeframe.

The database will provide the necessary data to feed the decision tree’s model. In more advanced models the database will be updated in real time. We can divide this data into several major categories:

	1. Capital and resources that are available to pursue the plan, this category includes most of what is currently available internally. In turn the capital will be subdivided into “flexible” capital stock that can be employed towards different production goals, “inflexible” capital and resources which can be repurposed with a certain loss (or additional investment) and “locked” resources which can only remain employed in their current mode of production.

	2. Market prices for possible inputs and additional resources that might be needed for production, including additional capital. Market prices and demand data for the organization’s products, as well as for potential substitutes and competing products.

3. Projections and tolerance levels for the market data in 2. and 3. helping to shape the decision matrix and adjusting the desirability of possible strategies in each possible scenario.


-------

For example, to keep things simple let us say that we have an oil producer with the following factors:

1. Capital:  oil rig (C.1.1), engineer (C.1.2), oil stock (C.1.3).
2. Prices:   oil rig (P.2.1), salary (P.2.2). Oil (P.2.3).
3. High-low scenarios in 1 year for all P values.

Our decision matrix would include a number of possible actions to perform each period:

Sell/buy oil rig (D.1), sell/buy oil (D.2), fire/hire engineer (D.3)

This would result in a number of scenarios with a growing number of potential outcomes the more periods that we consider. Once established the full decision matrix (always limited by the factors that we choose to consider and vulnerable to external shocks) the projections in 3. are applied resulting in a matrix of outcomes.

Decision Matrix	     Scenario Matrix	    Outcome matrices
DDDDD                  SSSSS                    OOOOOO
DDDDD                  SSSSS                    OOOOOO
DDDDD                  SSSSS                    OOOOOO

The utility of such a model increases the further that we project in time and as complexity increases beyond the point where traditional accounting makes results easy to see, requiring software tools. In this example a traditional balance sheet, profit and loss statement and a couple of scenario projections would be as useful as a more complicated model. However as soon as the number of elements in the decision matrix, the number of scenarios and the number of factors multiplies, the number of possible outcomes increases to such an extent that a software model and market planning tools still to be developed could be extremely useful.

Nevertheless we must warn that these tools will by nature be limited by the factors included in the model and rule breaking, disruptive innovation or other external factors can easily render such a model obsolete. Any sufficiently complex system will necessarily surpass the capacity of such models to predict potential outcomes. They will still remain very useful if they are used within their proper scope: either limiting predictions to short timeframes, or by limiting the scope of the model to few and easily controlled variables.




Félix Moreno de la Cova.
 Madrid, 2015.
* The idea of “Market planning” as a distinct technique was first expressed by Vernor Vinge.