Saturday, February 22, 2014

Value cycle and value equation


Creative Commons (BY NC CA) licence granted by the author(s)

Disclaimer: This blog entry reflects the thoughts of the author and does not speak on behalf of the Sensorica community. Further, the work is built on the work of the Sensorica community on value equation. Moreover, the author has many views on the value equation and this blog represents only of the many perspectives. Lastly, the author assumes that the reader is familiar with concepts of Open-value network.
The current capitalistic economic model was designed in the industrial era to reflect the thoughts, culture, technology, knowledge and processes of that era.  In fact, our current economic model has been optimized to reflect the technological (information processing) capacity of industrial era. The era of internet, however, requires a new economic model and new efficiency mechanisms. In order to understand the notion of value equation, it is important to understand the value cycle and the efficiency mechanisms of the current economy.


Value cycle

Value cycle refers to the processes of how value is created, exchange, distribution and accumulated in the economic system.

Fig: Value Cycle

Value creation refers to what we call "use-value" in economics.  People contribute in the value creation process by providing time, ideas, financial capital, labour, etc and value-streams (or relationship between contributions) are built during the value creation process, which often involves a (large) number of people, in order to satisfy present and/or future needs and wants of the market. Once sufficient value is created, it becomes Goods, Product and/or Services including knowledge (GPS) that can be exchanged for another GPS, which has undergone a value creation process by a (large) number of people.  The issue is that the value exchange among GPS becomes a matching problem.  That is, 5 apples = 3 oranges = 7 bananas = … varied by personal subjectivity; this is why money exists to simplify and create efficiencies in the value cycle. Afterwards, the exchanged GPS (ex: money, apples, etc) needs to be distributed (reward, salary, labour hours, etc) among a (large) number of people. Lastly, exchanged GPS or value is stored and re-used for future value creation processes (ex: savings, seeds, etc).

Currencies exist in order to create efficiencies in the market place by creating units for the value cycle. Conceptually, currency could be thought of as a standard for the unit of value.  Perhaps not so surprisingly, money becomes the de-facto currency since it already exists as units but at the end of the day, money is nothing more than a solution for the matching problem. In its current state money solves two crucial matching problems - value exchange problem (5 apples = 3 oranges = 7 bananas = $10) and value distribution ("equitable" reward) problem (or salary/pay in simple words). In a free-market system, this is a matching problem in a sense that the basic units exists and the people can do the matching ; hence, 1 apple = $2, 1 orange = $1, 1 banana = $0.5, etc. (exchange process), similarly, 1 hour work of engineer = $60, etc. (distribution process); whereas, in a controlled market, the government does the matching (subsidies, fixed income, etc). In reality, all systems are a combination of the two.

There are three major problems within the current industrial era based value cycle that would need to be addressed within the internet era. First, reserves (accumulation) process becomes value creation process by the principles of interests (money makes money); this makes money the de-facto currency. Second, only money is used for motivation during the value creation process by influencing the value distribution process, even though, research shows that money is a negative motivator (hygiene factor) - that is, without money people still work but with money people may or may not work. Third, value creation process can involve thousands of people (for example, open-source projects) but exchange value (including reward and money) can only be distributed to a small subset of participants in the value creation process. This phenomenon is observed because accounting during the value creation process and valuation during the value distribution process are optimized for the industrial era by reliance on extreme human intervention and not for internet era.

The key idea behind value equation is to reformulate the value distribution problem to a matching problem and disconnect money (or exchange value) from the process of value distribution. Even though, money could be the reward to be distributed, it is not the only basis of the accounting. In this way, value equation and accounting can provide a solution to the matching problem of value distribution in the internet era. The value equation, however, does not solve the value exchange or accumulation problem; although, the ideas in this blog could be extended to those problem sets.

Side notes:  Unitized (unit based) currency made sense when we did not have the technology to be able to "solve" complex NP-hard or NP-complete problems.  Perhaps, we still do not have the technology and mechanism (data) to "solve" the matching problem for a larger problem (ex: marketplace) but we do have the technology to "solve" matching problem for the value distribution process.  Although, technically speaking, we do not know how to "solve" NP-complete problems efficiently, that is, calculate the optimal solution. However, we do know how to approximate them and for a certain class of NP problems, we understand the range of error of the approximation.

Value equation

As mentioned above, value distribution process currently (in traditional organizations) uses money. Money, again, is a solution to the matching problem for the value exchange process that is also used within value distribution processes (ex: salary). Nevertheless, organizations also use stocks in the value distribution process. Stocks are unitized solution to the matching problem for the distribution process similar to as money is to value exchange process. Similarly, equity represents a non-unitized solution to the matching problem for the value distribution process. Value equation, simply put, is a way to decide how many stocks or how much equity to issue to each contributor to the project. That is, equity or stock may or may not have any market value or exchange value, similarly value equation is simply an agreement on the rules for the distribution of exchange value.

Conceptually, value equation could be thought of as an algorithm to solve the matching problem of how many stocks to generate or how to create equity in order to match the rewards to contributors as per the ethos of the system. The current societal approach to problem of value distribution today is that we create units and then hope that the solution to the problem is optimized in such a way that all parties would be satisfied. This is a challenging problem since we are relying to humans to solve a matching problem.  Matching problems are NP-complete or complex and hence, the optimal solution or even approximation thereof is not possible due to human factors (we cannot possibly track and compute this information in our heads). An optimal solution to the value distribution matching problem would accurately map the contributions to rewards but there is no such thing as the "true" value of contribution. This requires a complex accounting system to track the contributions and rewards as well as algorithm to perform the mapping. 

Value and contributions are subjective unless we analyze entropy (ecological economics deals with this to introduce objectivity) and/or information (which I argue is also subjective).  In comparing value distribution process to value exchange process, the question is similar to that of how many apples for oranges?  It depends on who we ask but we may get a median or mode types of responses to determine the "democratic" subjectivity but this subjectivity changes with time, advertisement, scarcity, etc.

Determining a value equation thus is a subjective issue determined by the ethos of the system. Value equation could vary from a capitalism (completely free-market) to communism (complete equality) to time-based system, and mathematics based (machine learning and collective intelligence). Similarly, the governance of value equation or governance equation (decision on value equation itself) would also vary from democracy (a representative decides on value equation, direct democracy (every participant decides on value equation) and liquid democracy (trust based representative decide on value equation) to meritocracy, dictatorship, kingship, etc.

One of the major advantages of value equation is the flexibility in defining the system based on ethos. Whereas the industrialized system enforced a certain ethos (capitalism vs. communism), the internet era allows for choice of ethos. Nevertheless, there are certain best practices that could be adopted to ensure longer-term continuity of the system.  The following section would provide one of my thoughts on how to design such a system (or value equation).

Note: rewards in our present society refer to money (and/or reputation at times, etc) but in future, rewards could be vary varied. For example, barter system (1-1 trading) and network barter system (many-to-many trading) is an interesting way to solve the matching problem of value exchange.  EconomyApp is working on the network barter system but it uses money as units in the barter system. It would be interesting in the future to link the value exchange and distribution matching problems. I think this would give rise to a completely different currency system.

 "Solving" for Value equation

In our approach in traditional economics, we generate money and then solve for equitability or satisfaction. That is, money is created and then we use the income distribution as a measure to understand the health of economic system. At Sensorica, we have been using a similar approach in the value equation design process: create the equation and then verify to see if people are satisfied and perhaps negotiation could be used after the fact for satisfaction. In this blog, I would like to present a different approach.

Key idea behind this new approach for value equation is to fix the income distribution first (how poor should be the poor and how rich should be the rich) and then place people ("high or low in the food chain") based on their contribution. Hence, using the placement of people with respect to their contribution as health check rather than the distribution of income. In other words, the income distribution (or relative inequality) is fixed.

The placement of people on the income distribution, however, is probabilistic (as per contribution) in the algorithm since the algorithm is stochastic in nature hence it would reduce any "corruption" or human negotiation errors.  Also, since it is stochastic (random at times), the solution generated may be more acceptable due to the phenomenon of procedural justice. From a motivation perspective, the algorithm is designed to ensure continuity of value creation process since it provides incentives to work hard (higher probability of being higher in the "food" chain) while keeping the income spread low and standard of living relatively decent at the bottom since anyone could end up at the bottom (but most likely those who don't contribute much would end up at the bottom).  

Premise behind the Value equation algorithm

(Some raw ideas for now)

Relative background needed

1) Process fairness
2) Cooperative vs. competitive games
3) Risk-based Game theory
4) Auction games
5) Scale-free networks
6) Statistics (for based lottery games)

(the talk talks about co-operative games a lot)
(most interesting part ~39 minutes on process fairness within cement delivery industry)  <-- there is a lot of other literature on process fairness

3) See the paper by Lara Buchak - "Risk aversion and rationality", July 2009

(although, I am not sure which type of auction game to play -- which bidding system)

5) Scale-free networks (wikipedia) - Better source of information: scientific america - scale free network (2003) article

(Again, lots more interesting papers exist on scale-free networks)

6) Lottery mathematics or stochastic weighted matching (for mathematicians)


Basic Idea:
- People get equity based on chance (probability is based on their contribution, the higher the contribution, the higher the probability of getting higher equity)
Why:
- Process could be more important than meritocracy or equality
- People are willing to agree upon a chance based system as long as the process is fair
- Because of chance, people have an incentive to reduce their risk and collaborate to lower the income differentiation (conceptually, increase the minimum wage and reduce the maximum wage)
- People would continue to play the game (as long as the effort generates exchange value) 

Refinement:

Assumption:
- Value accounting system exists and people insert contribution data (contribution is everything - time, money, ideas, labour, material, etc)

Auction:
- People bid to decide how much should the lowest worker get paid as compared to the highest worker (in ratio) ex: 1:12 (they tried this in Switzerland), 1:100, etc
- Auction can be a multiple round auction (for fixed number of rounds of until the system stabilizes) 
- Auction can be repeated with time (every months, every year, etc) at fixed or random interval
- Auction can be repeated when after the value exchange process has begun.
Nuance: In a unitized based system (ex: money based currency such as dollars), people may also bid on minimum wage.

At the conclusion of each auction:
- The system creates the income distribution function that follows either normal distribution or power-law distribution 

Nuance - 
- It would interesting to experiment to see which combination is better for distribution between governance equation and valuation equation.  Example: governance equation could follow normal distribution (most people have the same voting power), whereas, value equation could follow power-distribution (high contributors can a lot more than low contributors)
- it would be interesting to experiment ask people to bid between normal and power-law distribution

Next, the system picks people to win the "lottery" every, whereas, the lottery is where you get placed in the income distribution (or percentile). Your probability of winning a higher percentile lottery would need to proportional to your contribution.  

Nuance
- Contribution could be peer valuated or weighted.  
- People may or may not assign a decay function for your contribution (ex: depreciation of physical goods)

Lastly, the exchange value is distributed as per the distribution function and your placement on the distribution function.

Note: the auction could be conducted as a form of series of questionnaires  (mandatory or not) or direct democracy or liquid democracy, governance equation, etc.  For this game, I assume there is 100% submission and that the bidders understand the game (perfectly knowledgeable, participative, and rational bidders)

2 comments:

  1. Thanks Yasir for starting this. I validate your post. I think the intention behind it is very good and I think you are going in an interesting direction.

    ReplyDelete
  2. By Bob Haugen (who couldn't comment due to technical issues):

    Opposing opinion: the value equation should be transparent and predictable. No chance.
    It should be simple enough to be posted on the door. When people join Sensorica, they should know what to expect. It should be like a contract. If the value equation is changed, it should be by a democratic process involving all contributors.

    ReplyDelete