Logging into the Transaction Center

MyTransactionCenter can be localised to customer demand

The employee account panel, showing his current balance and previous transactions

An easy wizard to assist employees in making business payments, deposits and, of cource, withdraw their earnings at the end of the month

The "My Profile" window where employees can check their organisational status and roles

First I ran a Python script that created 10′000 employee records with their corresponding accounts [download script]. Then, a second script [download] simulated 10′000 transactions between randomly chosen employees.

The time was recorded with the Unix command time. The first script took 91 seconds to complete, the second 32 seconds (translating to ~ 312 transactions / sec). In both cases the network traffic between client and server was about 100 kbit/s.

The test setup:

• The server: MySQL 5.0.34 running on an Intel P4 Linux machine
• The client: the two Python scripts were run from a P4 Linux laptop connected to the server via 100mb/s Ethernet

No blue smoke came out of the server, so I guess the test was successful

The RDBMS I chose is MySQL 5.0.32. It is open source and widely used in web applications, such as this blog (run by WordPress).

I designed a database with 3 tables:

• TABLE employees to hold the login data for each employee. In a production environment one may instead have a dedicated server to manage identities and provide log-in.
• TABLE accounts to store the transactional accounts of the employees.
• TABLE transactions to log all transactions that are being made

The more important data fields of the three SQL tables. The foreign key relations are highlighted

The exact SQL spec of the three tables looks like this:

CREATE TABLE employees (
        id MEDIUMINT UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY,
        name VARCHAR(36) NOT NULL,
        login VARCHAR(12) NOT NULL,
        password CHAR(40) NOT NULL,
        status ENUM ('active', 'departed') NOT NULL DEFAULT 'active'
)ENGINE=INNODB DEFAULT CHARSET=utf8;

CREATE TABLE accounts (
        id MEDIUMINT UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY,
        holder MEDIUMINT UNSIGNED NOT NULL,
        balance DECIMAL(9,2) NOT NULL,
        created DATETIME NOT NULL,
        status ENUM('active', 'suspended', 'closed') NOT NULL DEFAULT 'active',
        INDEX (holder),
        FOREIGN KEY (holder) REFERENCES employees(id)
                ON DELETE RESTRICT
                ON UPDATE RESTRICT
)ENGINE=INNODB DEFAULT CHARSET=utf8;

CREATE TABLE transactions (
        id INT UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY,
        date TIMESTAMP,
        sending_account MEDIUMINT UNSIGNED NOT NULL,
        receiving_account MEDIUMINT UNSIGNED NOT NULL,
        amount DECIMAL(9,2) NOT NULL,
        reference VARCHAR(255),
        INDEX (sending_account, receiving_account),
        FOREIGN KEY (sending_account) REFERENCES accounts(id)
                ON DELETE RESTRICT
                ON UPDATE RESTRICT,
        FOREIGN KEY (receiving_account) REFERENCES accounts(id)
                ON DELETE RESTRICT
                ON UPDATE RESTRICT
)ENGINE=INNODB DEFAULT CHARSET=utf8;

Two notes regarding the SQL description of the tables:

• The foreign key relations have ON DELETE RESTRICT and ON UPDATE RESTRICT constraints in order to preserve the integrity of the records. For example, when an employee leaves the company his account data must not be deleted, otherwise the other employees may receive an incomplete account statement of their previous transactions. Instead, the status of the leaving employee record should be changed to ‘departed’ and his account flagged as ‘closed’.
• The employee.id and accounts.id fields are defined as UNSIGNED MEDIUMINT to provide ID numbers for up to 16′777′215 records. The transactions.id field is an UNSIGNED INT which can provide up to 4′294′967′295 transaction IDs. I assume this number would suffice even to a large company for at least several years: 100′000 employees x 10 transactions daily x 365 = 365 mio. transactions p.a. If not, the integer size can be increased, but this won’t be the only IT issue then

I decided to implement the transactional business logic as stored procedures within the database. The idea is to keep the ‘transactional engine’ as compact as possible by having the data and the logic close to one another. Again, there is nothing fancy to the actual transaction rules, just simple common sense code that says, for example, that before making a payment the program should check whether the employee has enough money in his account.

Here I’ll list only the make_payment() procedure, the rest are more or less of the INSERT/SELECT/UPDATE type of code. Expectedly, the procedure has to execute the transaction statements as an “atom” and makes use of the standard COMMIT and ROLLBACK commands.

# Make a payment
# returns SELECT as return_code:
#        0 on success,
#       -1 on insufficient balance,
#       -2 on missing/inactive sender account,
#       -3 on inactive recipient account
CREATE PROCEDURE make_payment (
                IN sending_account MEDIUMINT,
                IN receiving_account MEDIUMINT,
                IN amount DECIMAL(9,2),
                IN reference VARCHAR(255)
        )
        proc: BEGIN

                DECLARE status_result TINYINT SIGNED;
                DECLARE sender_balance DECIMAL(9,2) SIGNED;

                START TRANSACTION;

                # Check if sending account is active/exists
                SELECT COUNT(status) INTO status_result FROM accounts
                        WHERE accounts.id=sending_account AND
                        accounts.status='active';
                IF (status_result != 1) THEN
                        ROLLBACK;
                        SELECT -2 AS return_code;
                        LEAVE proc;
                END IF;

                # Check if receiving account is active/exists
                SELECT COUNT(status) INTO status_result FROM accounts
                        WHERE accounts.id=receiving_account
                        AND accounts.status='active';
                IF (status_result != 1) THEN
                        ROLLBACK;
                        SELECT -3 AS return_code;
                        LEAVE proc;
                END IF;

                # Check if sending account has enough money for the payment
                SELECT balance INTO sender_balance FROM accounts
                        WHERE accounts.id=sending_account LIMIT 1;
                IF (sender_balance - amount < 0) THEN
                                ROLLBACK;
                                SELECT -1 AS return_code;
                                LEAVE proc;
                END IF;

                # Proceed with transaction
                INSERT INTO transactions SET
                        sending_account=sending_account,
                        receiving_account=receiving_account,
                        amount=amount,
                        reference=reference;
                UPDATE accounts SET balance=balance-amount
                        WHERE id=sending_account;
                UPDATE accounts SET balance=balance+amount
                        WHERE id=receiving_account;
                COMMIT;
                SELECT 0 AS return_code;
        END;

This SQL code of course isn’t final and would probably evolve further as the software development goes on.

Finally, if this programming code hasn’t scared you, you may want to have a look at these database software jokes

More often that not I’m beginning to see people use the word “economic” in the context of shortsighted, unsustainable and unecological.

Can there really be a conflict between our endeavour for prosperity and the wellbeing of our planet and children?

Or is it the ineptness of some of our policy makers and business leaders at economic thinking that has given the purpose of this highly reasonable science a bad name?

Theoretically, you can get your transaction company going with a simple paper register. Transactions can certainly be administrated the old-fashioned way, using pen and paper. I suppose that for initial adoption experiments on a small scale this might actually be a good idea. Computers, however, have a number of advantages and anyone who is serious about implementing the Transaction Company concept cannot really ignore them.

The following diagram outlines an IT system for internal transactions. I arranged the functionality into modules, so that a company can start with a minimal system and extend it according to its needs. Those of you who are familiar with IT would recognise a system based on a common multi-tier architecture.

At the bottom, coloured in blue, is the server. It has two components:

• A database to store the transactional accounts of the employees and a register of all transactions — pending and completed
• Software that defines the corporate rules of internal transacting, along with other related business logic

At the top we have a set of six client modules that connect to the server. They should be more or less independent from one another and a company can use one or more of them, according to its requirements:

• A desktop client: this, together with the server, comprises the minimal software system. It allows employees to log onto their transactional accounts, check their balance, and send and receive money. The client can be implemented using standard web technology, so that people can access the system through a web browser
• A set of three modules useful to larger companies (e.g. 50+ employees) and/or to businesses spread over multiple locations:
  • A market forum that allows employees to post, search and negotiate/bid for projects. Users may also have the ability to browse the transactional histories of their colleagues and to collect comments/rating points from their transaction partners (like the stars on eBay). The substance of the market forum would not differ much from the social networking sites on the Internet.
  • A module to visualise the network of transactions across the enterprise: with varying granularity (e.g. individual or departmental level), aspects such as connection structure, money flow and value creation. The idea of this tool is to assist decision making, so that managers/employees can spot profitable/active areas and direct their attention there.
  • A banking interface to allow the transactional accounts to be hosted with a financial institution. Employees may then earn an interest rate on their balance or be provided with access to credit facilities. I think this could turn into an interesting new service for banks.
• A set of two modules to ease transacting and make it more pervasive across the enterprise:
  • An interface to allow transactions to be automated. This could be a useful feature in manufacturing to make automatic payments when a part on the production line passes from one stage to the next
  • A mobile interface that employees can use to make transactions using their smartphones.

Yes, the graph is backed by concrete statistical data. But these numbers, by themselves, speak nothing of their cause.

Were changes in fiscal policy the primary cause behind the divergence of incomes? Taxation does affect income: directly, through policies for its redistribution from the richer to the poorer, and indirectly, by stimulating the economy. The first mechanism, however, doesn’t address the economic problem that led to the income inequality in the first place, it is only meant as a relief after the fact, while the second has too broad an effect.

I would rather propose to look elsewhere for the cause. To me the government has very little to do with both the problem at hand and the solution. Those of you who have read my previous article “Held Together by Faith” would know where my argument is heading to.

Not only in the US, but also here in Europe politicians are talking about a growing income disparity and a falling behind of the middle class. Some attribute this to deficits in education, others to the increased competition that arrived with the globalisation.

I say the principal cause lies elsewhere: corporations have been stagnating with respect to the market and globalisation has come to expose this, revealing millions of employees caught in workplaces where they fail to develop their economic potential.

Who makes up the top 10% of the income bracket? Mostly investors in the stock market, entrepreneurs, independent professionals, people with own businesses. What is common to all of them? Their greater engagement and activity in the open market where they enjoy a higher degree of economic potentials – compared to the workplaces within companies where one generally has as an employee a much more constrained playing field.

This economic divide increased and became more apparent over the past 10 – 15 years. After the end of the Cold War the market capitalist system underwent a significant boom. It expanded to the countries of the former Eastern Bloc and to China where hundreds of millions of people live. In the Western world it also went through developments marked by the liberalisation of many sectors, the falling of various protectionist barriers to trade, the adoption of electronic communications and trading, the privatisation of major utility companies across Europe and the introduction of the Euro. Markets became global.

Therefore, it is no surprise that to the entrepreneurs in the market globalisation came like manna from heaven: access to more capital and larger markets, new possibilities for networking and investment, exchange of ideas and know-how on a broader scale.

Corporations, on the other hand, developed rather sparsely over the same period. Most internal structural shifts, however heroic, were limited to the accommodation of computer technology. Still, over the past 10 – 15 years an array of companies underwent huge expansion and posted record profits. You may wonder, “how come?” The answer is simple: the primary cause for this spectacular growth of corporations didn’t originate from within, but was due to the improved conditions in the market environment: more people to sell to, more sources of finance, access to cheaper labour.

So, in a way globalisation both exposes and masks the growing gap between the market and the corporation. On one hand it presented companies with “free” room to expand, on the other, it exposed the fact that many employees behind corporate walls are falling behind. I suspect businesses would have faced greater pressures to reinvent themselves had the spread of the market capitalist system not provided new pastures to compensate for that.

Sooner than later, however, the world will run out of countries with non-market economies as the last stubborn supporters of “communism” succumb to capitalism; the population growth will slowly begin to level off. At last then the corporations will notice that they’ll have to look inward if they are to avoid stagnation, that we cannot have sustainable growth – of corporate profits, of personal income, if the economic potential of employees is being depressed.

There aren’t many options to close the gap, in fact I see just one – companies becoming more like their environment by adapting mechanisms from it: allowing monetary relations between employees, decentralising decision making, trimming frustrating bureaucracy. Yes, there will be initial suspicion and resistance to these changes, but in the end most people will like their new transformed workplaces – the possibilities for unrestricted interaction with colleagues, the more personal touch of relations in the office, the simple money rules, the ability to earn more.

This blog’s aim is to walk you through this new and transformed company. Stay tuned!

I thought about your idea during my bus ride to work this morning. I asked myself: How can the bus driver participate in the transaction company? Should he be able to reward the technician that maintains the bus for him? Should he give him 10 Euros every time he feels like giving it him? How does he determine this value anyway. How do you determine the [credits amount] he can hand out during a period of time? [...] How do you know that employees know how to value the services of the other?

Transactions are not this complicated. Actually, they can be quite intuitive and easy to grasp.

Let’s take Michel’s example of the bus driver and create a hypothetic bus company built around internal transactions. I chose it to have five people, which should suffice to illustrate the concept:

	The boss of the bus company
	Three bus drivers
	A technician

A company normally requires material assets and people in various functions in order to operate. In our bus company these are arranged as follows:

• The boss provides the material assets. He obtained a bank loan to buy three buses and a bus garage to park them overnight. He is also responsible for the marketing and for obtaining the required bus line licenses from the city council.
• Three drivers for each bus. They also sell the tickets to their passengers.
• A technician who takes care of the buses’ maintenance. His repair shop is located in the bus garage.

How do you turn the bus company into a Transaction Company?

The recipe: Give our five little heroes the possibility to establish monetary relations with one another in order to obtain the required inputs for their jobs and then to gain from what they produce/provide. The concept does not prescribe anything more specific because the exact terms for a given transaction are best left to the interested parties to define.

The recipe for The Transaction Company is simple

A word of caution: Driven by a desire to achieve greater control, the boss (or anyone else in the company) has to be careful not to distort the recipe by insisting on transactions with employees with whom one actually has no direct work relations. This is most likely to be counter-productive, resulting in moral hazard and perverse incentives — precisely the problem, common to many conventional organisations, that The Transaction Company aims to solve. This is an important issue which I would like to write about in another article.

Let’s now apply the recipe to the bus company, together with some economic common sense in order to anticipate how the five heroes might set up the transactional relations among themselves. Here is one possible scenario:

For the three buses that the drivers need for their job, the boss agrees with them on a lease fee to be paid to him monthly. In addition, a percentage of the ticket sales is agreed to go to the boss for his administrative and marketing work. Finally, the drivers agree with the technician on a pay for his maintenance services. He, however, needs some space in the bus garage to set up his repair shop, and for that he negotiates a monthly rent with the boss.

To sum up, for any one of our heroes to identify their possible transactional relations, they simply have to ask themselves two questions:

1. “Who do I need for my job?”
2. “Who needs me?”

The following interactive diagram maps the internal transactions (among our five little heroes) along with the external ones (with the customers and the bank). The internal transactions of a company cannot be regarded in isolation from the market environment. This is a great feature of The Transaction Company that adds to the understanding of the business in its financial relations with customers, suppliers and creditors.

Now that we know how the transacting partners in the firm identify one another, the next logical step is to look into the valuation of the actual transactions. But before we discuss this, let’s remember how personal income is derived. The Transaction Company says that this is simply the difference between the inbound (money received) and outbound transactions (money paid) for a person over a period of time (here I ignore the special case when shareholder dividend is paid):

• For each bus driver, this is the money collected from the ticket sales after the bus lease, the sales percentage and the maintenance service have been paid for.
• For the technician this is the money received from the bus drivers for his maintenance services minus the amount he paid to the boss to have his repair shop in the garage.
• For the boss this is the money received from the three bus drivers and the technician minus the loan instalment owed to the bank.

Getting back to the valuation of transactions, there is a question that I often get bombarded with: “But how would the person X know how much a particular transaction is worth?” I have answered this in a previous post, but now I’ll explain it again, using the bus company example at hand.

As I mentioned above, the mantra of The Transaction Company is “give people a framework for monetary transactions, the rest they shall sort out themselves”. This also includes the pricing of transactions. Valuation is subjective, therefore I see no point in attempting to impose some “objective” rules to it.

I can’t tell for sure what strategy a particular person would choose to determine and negotiate transaction value, but here is one possible to start with. For example, it is not hard to imagine that many people would aim to earn at least the typical average for their profession (this kind of strategy is called satisficing). So, let’s assume that one of our bus drivers wants to earn at least as much as his peers, e.g. 2000 Euros per month. In order to net this amount, when negotiating the terms of his transactional relations he would have to achieve such a balance between received and spent money, so that at the end of the month he is left with minimum 2000 Euros. As a simple equation:

Sum of received money - Sum of spent money ≥ € 2000

or more specifically

Ticket sales -  Bus lease - Sales percentage - Bus maintenance ≥ € 2000

The bus driver may then notice that of the four factors affecting his pay, some are more variable than others, or that some he can influence more easily. For example, the boss may agree on better terms of the bus lease, while the technician may not accept to work for less because his skills are in high demand; or he may see a chance to increase his passengers by adjusting the timetable. In any case, our bus driver has a clear picture of the factors that make up his personal income and a direct mechanism to act upon them. Employees in conventional companies do not enjoy this kind of transparency and flexibility.

Regardless with which strategies our five small heroes set out initially, in the end they will balance out. And because the bus company is part of a larger economic system, the transaction values will have to balance not only internally, but also with respect to the market. This balance, of course, will be dynamic; it will shift as internal and external business conditions change.

Mathematically, you can represent the problem of transaction value balance as a system of equations. For our bus company of five people we will have five such equations, that is, one per person.

If we assume the following

• The three bus drivers aim to earn at least € 2000, the technician at least € 4000, and the boss at least € 7000
• The bus lease (BL) and ticket sales percentage (SP) is the same for all drivers
• TS_1,2,3 signify the ticket sales for the corresponding driver
• BM_1,2,3 signify the bus maintenance outlays for the corresponding driver
• RS is the rent paid by the technician for his repair shop
• LR is the loan repayment owed by the boss to the bank

the resulting system of equations will look like this:

TS1 - BL - (TS1 x SP) - BM1 ≥ € 2000

TS2 - BL - (TS2 x SP) - BM2 ≥ € 2000

TS3 - BL - (TS3 x SP) - BM3 ≥ € 2000pre>
BM1 + BM2 + BM3 - RS ≥ € 4000
3 x BL + (TS1 + TS2 + TS3) x SP + RS - LR ≥ € 7000

Now, I didn’t mean to scare anyone with this equation. I just wanted to demonstrate that transactions cannot have arbitrary values, but values that are balanced by the interplay of the people’s economic relations within the company and the greater market environment.

I hope this argument is going to calm Michel’s concerns that people in The Transaction Company may turn irrational (in the economic sense) and, for instance, start awarding each other exorbitant sums of money. Michel, money always has to come from somewhere in the end! The fact that The Transaction Company doesn’t have a centralised remuneration system doesn’t mean that people would have an incentive to do crazy things. Actually, it is quite the opposite — we have a robust system with transparent economic mechanisms that emphasise self-responsibility.

In a recent interview, Alan Greenspan, former chairman of the Fed, shared his opinion that the expansion of the market capitalist system (referring mostly to the former Eastern Bloc and China) is due to the spread of the general belief among people that it leads to material prosperity. He, however, noted that if the currently increasing gaps in income, health-care and education continue to widen, people may lose faith in the system and its existence may become threatened:

Photo Trackrecord

“You cannot have a market capitalist system if there is a significant view within the population that its rewards are unjustly distributed.”

– Alan Greenspan on Charlie Rose’s TV Show, 21 September 2007

As I have previously said, the market capitalist system highlights the solution; it is not a cause of the problem. Increasing taxation or attempting to appropriate some of the core market functions only distorts the situation; it doesn’t bring us further.

The market system has been so successful because it enables participants to interact with one another efficiently, stimulating productive economic activity that generates material well-being. Entrepreneurs, that is, people with extensive direct symmetrical presence in the market, note — both as consumers and producers of value, have prospered tremendously. Starting your own business is widely regarded as the way to go if you want to become rich.

Contrast this with the general perception about companies: how many people truly believe that they can earn riches by becoming employees? Not many!

The real ‘disparity’ here is that employees have a one-sided presence in the market - only as consumers of goods/services, never as producers. In this they miss a lot. Within companies millions of employees are isolated from many of the beneficial mechanisms of the market, like the ability to form direct monetary relations with colleagues, to reward one another financially, to enjoy a higher degree of economic agency — in order to learn, develop and earn more.

It is in this asymmetry where the real ‘gap’ is hidden: companies that are internally devoid of the beneficial mechanisms of their market environment; employees that don’t have a chance to fully experience their economic potential as producers of value but only as final consumers.

You don’t have to be an independent entrepreneur in order to prosper, you can also become rich as an employee of a company. It is a mission of The Transaction Company to restore this faith in our workplaces.

I drew a pyramid to illustrate this important hierarchy.

Fig . 1. Business innovation is stacked into three different levels. Each level of innovation is constrained by the capabilities of the one below it.

To sum up, product innovation cannot be better than the underlying processes permit, and similarly, the capability to create innovative processes is limited by the organisation of the underlying institution.

At the end of his article, John points out “the single most wealth creating innovation over the past several centuries”. According to him, this is the development of the limited liability joint stock company. Is The Transaction Company going to become the next big institutional innovation?

Fig. 1. Traditional enterprises tend to compartmentalise their economic intelligence

Businesses have traditionally been organised in a way that compartmentalises economic decision making within the enterprise, constricting it to a group of people called the ‘managers’. This group has the task to be the ‘economic brain’ of the organisation, taking care of the economic survival and growth, while the rest are to be devoid of all economic thinking.

This division, however, requires a good communications channel between the ‘brain’ and the ‘body’. As corporations grow and responsiveness begins to matter more in our fast paced age, this channel becomes overloaded and noisy. The solution? Add more brain cells? Make the channel fatter? No, I would rather say “decentralise the economic reasoning of the enterprise!” But is this possible?

This long-standing status quo has created the unfortunate perception that employees generally have low economic intelligence and cannot be entrusted to deal with money matters. But at the same time, these same people, in their private lives, go shopping every day, seek out the best holiday packages, and sign contracts for car leases or house mortgages. Nobody was born with these particular skills, people have learned them.

It doesn’t take much to obtain the skills to work in The Transaction Company. We just have to bring sensible and simple economic feedback mechanisms into the workplace. This, in fact, is precisely what the concept is about, thus creating a self-learning environment where people can quickly orient themselves and progress much on their own. The higher degree of both independence and interaction, the simplicity of monetary transactions, the overall transparency of the organisation, all these factors contribute to a great place, not just to learn, but to have fun too.

This would gradually lead to the collective economic intelligence of the enterprise becoming both more evenly spread across it, as well as having a higher overall level compared to traditional businesses. Transaction companies will still have their ‘economic brain’, but it will be relieved to pursue more pleasurable and higher-level tasks.

Fig. 2. Distribution of economic intelligence across the enterprise

The greater independence, responsibility and social skills that The Transaction Company teaches would also benefit people beyond their jobs, in their personal and social life. Do not be surprised if we eventually witness some profound shifts in our society too!

* * *

PDF version of this article (641 KBytes)