Someone Just Turned the Lights On

Dil Green, 18th February 2021. Image credits: Fleischman, T. et al.

In the autumn of 2020, several members of Mutual Credit Services launched a loose symposium with the not-too-catchy identifier ‘Circular Trade Analytics’.

What we wanted to do was to develop tools and techniques that enabled us to look at the relationships and dynamics in economies which were practicing various mechanisms we identified as ‘circular’ in character – in particular, how various flavours of mutual credit could underpin a circular (material) economy.

A small but wide-ranging group began to meet fortnightly and it became clear that there was significant work to be done – that despite the enormous worldwide output of economics-related papers, books, articles, and analytical tools, there seemed to be very little that was applicable to the kinds of networks we are setting out to build – and that what work does exist tends toward the socio-economic, rather than the quantitative.

We know that we need to be to able quantify benefits for members, to be able to understand when trading networks are healthy or not, what the correlates are of these conditions, and how to advise network managers as to how flow of trade may be encouraged. These and and many other questions will need good answers if we are to build out and support thriving networks that support community wealth and rebuild the commons.

It seemed as though, despite the work already achieved by attendees, such as the output from Grassroots Economics, there was little to build upon, and a great deal to do.

And then we were introduced to a new paper, by Tomaž Fleischman, Paolo Dini, and Giuseppe Littera; Liquidity-Saving through Obligation-Clearing and Mutual Credit: An Effective Monetary Innovation for SMEs in Times of Crisis, and it was as if someone had turned the lights on.

It turns out that, unbeknown to most, a significant body of technique and experience has been developing over decades in Slovenia for the analysis of trading networks at scale, and, in particular, the discovery of ‘cyclic structures’ or ‘loops’ of obligations within the economy (A owes B, B owes C, C owes D, D owes E, E owes A) – the very circularity that the Circular Trade Analytics group set out to understand. 

The paper describes an ‘invoice clearing’ mechanism that, through detecting these structures, helped the Slovenian economy to bounce back from its war of independence from Yugoslavia by allowing up to 8% of the country’s trade transactions to be settled without money changing hands. It considers the impact such a system might have if deployed in the existing Sardex business network. The authors thereby demonstrate the remarkable impact that two key ‘new economy’ approaches in combination – namely, multilateral obligation set-off and mutual credit – can have on business finances.

The paper analyses a large number of Sardex business-to-business mutual credit transactions (over 130 000, totalling >€30M’s worth of trade for the year 2019), looking at the impact of different approaches to the settlement of these transactions on the fiat balance-sheet positions of the participant firms. At any point, each business will have a trade balance of outstanding ‘accounts payable’ (to creditors) and ‘accounts receivable’ (from debtors).

The default position, of course, is that each invoice is settled individually with a bank transfer. This is the trade world as we know it, with all its attendant risks and impacts – late payments, defaults, invoice factoring, expensive overdraft finance, and the rest. This is known in the jargon as ‘Real Time Gross Settlement’ – and is taken in the paper as the ‘baseline’ for comparison.

The graph above shows the distribution of final balances for each individual business represented in the data set. This demonstrates a fairly wide range of outcomes, with businesses of small and large turnover represented among those with net deficits, balanced outcomes, and net surpluses, and – as expected – a preponderance of firms with the largest net balance positions (whether positive or negative) having relatively large turnover.

Multilateral obligation set-off

The first mechanism analysed is invoice clearing. Within a sufficiently large set of transactions, it is possible to discover cyclic patterns which connect loops of obligations together – meaning that all of these can be ‘set off’ – resolved without any bank-money transfers. Such a set-off does not change the eventual balance position of the firm, but does reduce the amount of bank currency it needs to conduct its trade, and thus improves cash-flow and availability of working capital.

Using a loop detection algorithm called TETRIS that was developed in Slovenia in the 1970s (a significant piece of work in itself, since this is a computationally expensive NP-hard problem, further complicated by the fact that the loops are not ‘disjoint’ – to use the term from the paper – but deeply interconnected), the authors show that, for a significant proportion of Sardex members, a notable proportion of their trade can be cleared in this way. For some, applying this mechanism could halve the amount of bank-money needed across the year, while the average would be about 25%. This is significantly greater than the corresponding figures for Slovenia, as Sardex is embedded within Sardinia's relatively closed island economy (resulting in a denser network of obligations with more loops).

Note that (for smaller, high-trust groups) this computationally tricky approach can be avoided by the simple expedient of all participants submitting invoice data to a shared ledger. This ledger simply adds up all credits and debits for each member to produce a cleared balance – the net sum owed to or by a firm in respect of the whole network. At the end of each clearing period, each firm either pays or receives a single bank-money transfer from the central clearing account to achieve settlement for all invoices. This simpler approach is used as the first stage of Mutual Credit Services’ Trade Credit Clubs proposal for business-to-business mutual credit.

Mutual credit

The paper goes on to analyse the impact for firms of adding a further mechanism – mutual credit.

In a business-to-business context, this can be understood as a pooled trade credit arrangement. Participants mutually agree to allow a certain amount of value exchange to be settled using an accounting unit internal to the network. This reduces the amount of bank-money needed still further, as non-cleared balances at the end of each clearing period need not be paid in full, but simply brought back within acceptable limits. The network, through the aggregate commitments of all of its members, is effectively providing its own money supply.

The paper assumed that each firm has access to ‘rolling trade credit’ amounting to 2% of annual turnover (the rule applied in the Sardex network), and shows that the impact of this is approximately equal to the impact of the multilateral obligation set-off mechanism – a reduction of the order of 25% of bank-money needed to clear obligations at the end of the trading period.

The Trade Credit Club model includes this mechanism, too, but rather than imposing a blanket credit limit on participating firms, offers a process whereby each firm proposes both the maximum credit balance it is prepared to hold and the maximum credit it wishes to access within the network. Once all members have made proposals, an optimisation routine can set workable balance limits. This process can be repeated at intervals, so that liquidity provided by the network can respond to changing conditions.

Combining both mechanisms

Finally, the authors show that it is not only possible to combine both mechanisms, but that the impact is additive, so that (for the average Sardex member) an approximately 50% reduction in bank-money requirements can be achieved.

In discussing other large transaction data sets with Tomaž, it seems that provision of internal liquidity within the network at the level of around 20% of the total turnover can allow for essentially all trade within the network to be settled without need for bank-money.

Historical context

Alongside the quantitative findings outlined above – which are dramatic enough in themselves, the paper also provides some fascinating history with a potted history of the development of the Slovenian system.

In the section with the heading ‘History and Context’, a rather dry and minimal account is given of the routine usage of centralised obligation clearing by the Slovenian government from 1991 onwards, and the contribution of this mechanism to the weathering of economic turbulence in the country. This graphic shows the counter-cyclical take-up of this mechanism during the credit crunch that followed the 2008 crisis.

What is not described in the paper, but which came up in conversation with Tomaž is the attempts by the EU to convince the Slovenian authorities to stop using this mechanism when they were negotiating to join the Eurozone, despite the fact that it had made a major contribution to saving their economy on several occasions. Although the Slovenians refused, they did have to accept restrictions on the operation and scope of the mechanism. The result of this is that the clearing mechanism is not utilised to the full. It should also be noted that there is no mutual credit supply into the Slovenian system.

… and that’s not all!

The above outline covers only my interpretation of the most immediate findings from the paper, but it repays close reading, so I do encourage you to spend some time with the full text.

Some other nuggets include:

• Some rather lovely graphical network mapping images (which are themselves just a taster for watching Tomaž demonstrate them ‘live’);

• Useful discussion of the character and effects of loops in trade networks – in particular around the risk that such loops pose in credit squeeze situations – since they can cause a ‘freeze’ that extends throughout an economy – where no-one can pay anyone else;

• The software used for loop detection is also described in some detail.

And finally, an eye opener…

The section on ‘liquidity saving mechanisms’ was an eye opener. It makes it clear that various mechanisms to manage liquidity in money systems are well-understood in the banking sector. Clearing and Settlement Systems from Around the World: A Qualitative Analysis, referred to in the paper, documents the extensive use of mechanisms used in bank payments systems around the world. Again, some reading between the lines and looking at referenced material makes it clear that the advantages of these are not the subject of public discussion or made available outside the banking sector or multinational corporates.

In addition to the clearing approaches used, it is clear that many of these systems include close analogues to mutual credit mechanisms (albeit of restricted scope) when they incorporate mechanisms described as “… liquidity reservations, transaction prioritisation and timing, and active queue management…”.

It’s pretty obvious why these practices are kept ‘under wraps’ – the business model of commercial banks depends upon making access to money costly. Offering liquidity saving mechanisms to networks of businesses which allow them to diminish their need for bank finance is not something they are likely to promote. 

There are a number of references to follow up here, but the most important learning so far from this section for me has been that the mechanisms underlying mutual credit and obligation clearing are not in themselves either ‘novel’ or ‘radical’. The innovation we provide is not in the mechanics, but in their implementation – in bringing them to a different context. Bringing them out from the financial management back rooms of banks and corporates, and into the service of the value-producing economy. This theme is explored further in a companion piece to this post – ‘Complementary’ economics won’t do it.