Stock-Flow consistent model
Stock-Flow Consistent (SFC) models are a family of macroeconomic models based on a rigorous accounting framework, which guarantees a correct and comprehensive integration of all the flows and the stocks of an economy. These models were first developed in the mid-20th century but have recently become popular, particularly within the post-Keynesian school of thought.
The accounting framework behind stock flow consistent macroeconomic modelling can be traced back to Morris Copeland's development of flow of funds analysis back in 1949. Copeland wanted to understand where the money to finance increases in Gross National Product came from, and what happened to unspent money if GNP declined. He developed a set of tables to show the relationship between flows of income and expenditure and changes to the stocks of outstanding debt and financial assets held in the US economy.
James Tobin and his collaborators used features of stock flow consistent modelling including the social accounting matrix and discrete time to develop a macroeconomic model that integrated financial and non-financial variables. He outlined the following distinguishing features of his approach in his Nobel lecture 
- Modelling changes between discrete short-run time periods rather than a long run equilibrium
- Tracking changes in stocks of assets held by different groups
- Multiple assets with different rates of return,
- Modelling of monetary policy operations
- Subjecting the demand functions to "adding up constraints"
The current SFC models mainly emerged from the separate economic tradition of the Post Keynesians, Wynne Godley being the most famous contributor in this regard. Godley argued in favour of wider adoption of stock flow consistent methods, expressing the view that they would improve the transparency and logical coherence of most macro models
Structure of the Models
SFC models usually consist of two main components: an accounting part and a set of equations describing the laws of motion of the system. The consistency of the accounting is ensured by the use of three matrices: i) the aggregate balance sheets, with all the initial stocks, ii) the transaction flow, recording all the transactions taking places in the economy (e.g. consumption, interests payments); iii) the stock revaluation matrix, showing the changes in the stocks resulting from the transactions (the transaction flow and the stock revaluation matrix are often merged in the full integration matrix). The matrices are built respecting intuitive principles. Someone’s asset is someone else’s liability and someone’s inflow is someone else’s outflows. Furthermore, each sector and the economy as a whole must respect their budget constraint. No fund can come from (or end up) nowhere.
The second component of SFC models, the behavioural equations, include the main theoretical assumption of the model. Most of the papers in the existing literature are based on post-Keynesian theory. However, the behavioural equations are not restricted to a single school of thought.[nb 1]
Advantages and Disadvantages
The comprehensive accounting framework has several advantages. Tracking all the monetary flows taking place in an economy and the way they accumulate, allows for a consistent integration of the real and the financial side of the economy (for a detailed discussion see Godley and Lavoie, 2007). Furthermore, as balance sheets are updated in any period, SFC models can be used to identify unsustainable processes, for example a prolonged deficit of a sector will result in an unsustainable stock of debt. These models were used by Wynne Godley in forecasting, showing promising results. Moreover from a modelling perspective, the consistent accounting framework prevents the modellers from leaving "black holes" i.e., unexplained parts of the model.
The SFC are usually aggregated macroeconomic models and as such they incur all the disadvantages typical of this approach. For example, they explain little of what happens at microeconomic level. Furthermore, they often rely heavily on parameter values used by the modellers. A possible way to overcome these drawbacks is to combine these models with an agent-based model (ABM). A recent 'benchmark' model was developed by Alessandro Caiani and colleagues.
Example of SFC model
Flow of funds between sectors in a closed economy
|Households||Firms||Government||Rest of the World||∑|
|Changes in Money||-ΔHh||+ΔHs||0|
The above table shows the flow of funds between different sectors for a closed economy with no explicit financial sector. The minus (-) sign in the table represents that the sector has paid out while the plus (+) sign indicates the receipts of that sector, e.g, -C for the household sector shows that the household has paid for their consumption whereas the counter party of this transaction is the firm which receives +C. This implies that the firms have received the payments from the households. Similarly, all the respective flows in the economy are reported in the flow of funds. More advanced SFC models consist of a financial sector including banks and is further extended to an open economy by introducing the Rest of World sector. Introducing the financial sector enables in tracing the flow of loans between the sectors, which in turn helps in determining the level of debt every sector holds. These models become more complicated as new sectors and assets are added to the system.
The model structure
Once the accounting framework is fulfilled then the structure of the model, based on stylized facts, is defined. The set of equations in the model defines relationship between different variables, not determined by the accounting framework. The model structure basically helps in understanding how the flows are connected from a behavioral perspective or in simple words how the behavior of a sector affects the flow of funds in the system, e.g., the factors that affect the consumption (C) of the household is not clear from the flow of funds but can be explained by the model. The model structure with a set of equations for a simple closed economy is given by:
Y = C + G
T = θY
YD = Y - T
C = α1 Y + α2 Ht-1
ΔHs = G - T
ΔHh = YD - C
H = ΔH + Ht-1
Y (Income), C (Consumption), G (Government Expenditures), T (Taxes), YD (Disposable Income), ΔH (Changes in stock of money) and θ is the tax rate on the income of household sector. α1 is the household consumption out of disposable income. α2 is the household consumption out of previous wealth.
The SFC models are solved in different ways depending on the aspect of research but in general initial values are assigned to the stocks and then the model is calibrated or estimated.
- " “According to Gennaro Zezza the accounting consistency should be a requirement for all macro model. Models with post-Keynesian behavioural assumption should therefore be a sub class of macro model labelled stock-flow-consistent post-Keynesian models” 
- Caverzasi, E., & Godin, A. (2014). Post-Keynesian stock-flow-consistent modelling: a survey. Cambridge Journal of Economics
- Caverzasi, Eugenio; Godin, Antoine. "Stock Flow Consistent Modeling Through the Ages" (PDF). Levy Institute.
- C.H. Dos Santos. "Notes on the Stock Flow Consistent Approach to Macroeconomic Modeling" (PDF). Retrieved 14 December 2014.
- Dos Santos, Claudio H. (2002). “Cambridge and Yale on Stock-Flow Consistent Macroeconomic Modeling.” Department of Economics, New School University, New York
- Godley, W. (1999). "Seven Unsustainable Processes" (PDF). Special Report.
- Caiani, Alessandro; Godin, Antoine; Caverzasi, Eugenio; Gallegati, Mauro; Kinsella, Stephen; Stiglitz, Joseph E. (2016-06-06). "Agent Based-Stock Flow Consistent Macroeconomics: Towards a Benchmark Model". Rochester, NY. SSRN .
- Lavoie, Marc "Stock Flow Consistent Modeling", video and slides