Vertical/Horizontal vs Exogenous/Endogenous

In MMT, a distinction is made between ‘vertical’ and ‘horizontal’ transactions. The former refer to transactions between the government and non-government; the latter to transactions within the non-government. I thought it might be helpful to clarify the significance of this distinction, and also what it does and does not imply about the modern monetary system and real economy.

The vertical/horizontal terminology is sometimes resisted by other economists working along Post Keynesian lines. I retain it here simply because it is the terminology currently used by the academic MMTers.

Vertical Transactions and the Level of NFA

In MMT, government money is depicted as vertical whereas bank money is considered horizontal. The reason for this does not concern the question of endogeneity or exogeneity of money but rather the fact that both the level of reserves and the level of net financial assets (NFA) – the sum of currency, reserves, and government bonds – can only change through vertical transactions.

Whereas horizontal transactions always involve both a non-government asset and a non-government liability, netting to zero, a vertical transaction involves a non-government asset or liability that is not offset within the non-government, but instead has its accounting offset in the government sector.

For example, if the government pays a pensioner, it credits the private bank account of the pensioner (asset of the pensioner, liability of the private bank) and credits the reserve account of the private bank held at the central bank (asset of the private bank, liability of the government).

In this example, the private bank’s position is unaffected by the transaction. Its liability to the pensioner is offset by its asset at the central bank. However, the pensioner’s asset has no offset within the non-government. It is matched by the government’s liability to the private bank. So, overall, the effect of this particular vertical transaction is to add net financial assets.

More generally, government spending increases net financial assets and tax payments destroy them. A budget deficit corresponds to an increase in net financial assets, a budget surplus corresponds to a decrease in them, and a balanced budget leaves them unchanged.

Although it is true to say that government spending causes an increase in net financial assets and tax payments cause a reduction in them, when discussing the budget position, the language adopted needs to be more circumspect. Without further information, we can only say that a budget deficit corresponds to an increase in net financial assets and a budget surplus corresponds to a decrease in them.

The reason for this more circumspect choice of words is that the budget outcome is endogenous. This is the case even though some components of fiscal policy are exogenous. Most obviously, government expenditure is exogenous. Its level is not dictated by current income. In addition, decisions over tax rates and the types of taxes and spending programs to introduce are exogenous.

Even so, the budget outcome itself is endogenous. This is because of the automatic stabilizers. When income rises, tax revenues automatically rise. This will reduce the budget deficit. When income falls, tax revenues fall. This increases the budget deficit.

Since the budget outcome changes endogenously with income, so does the level of net financial assets. In fact, changes in net financial assets mirror the budget outcome. A budget deficit corresponds to non-government net saving, whereas a budget surplus corresponds to non-government net dissaving.

The relevance of this is to highlight that the vertical/horizontal distinction in MMT does not correspond to exogenous/endogenous. Some aspects of vertical transactions are exogenous (e.g. government spending, setting of tax rates), but other aspects are endogenous (e.g. tax revenue, the budget outcome).

Because the budget outcome and level of net financial assets are endogenous, they can be influenced by both government and non-government spending decisions.

The government exerts both a direct and an indirect influence on the level of net financial assets through its spending. Government expenditure affects the level of net financial assets both directly (as the government credits private bank accounts) and indirectly through its impact on income, level of tax payments (debiting of private bank accounts), and the budget outcome.

The behavior of the non-government can only affect the level of net financial assets indirectly through its influence on economic activity. For example, an exogenous change in private investment, financed through private credit creation, will alter income, tax revenues, and the budget outcome.

Despite the indirect influence of non-government activity, including private credit creation, on the budget outcome, there is an important sense in which vertical transactions are key to the determination of the level of net financial assets. The key point is this: only the government can enable a change in net financial assets alongside stable income. If the non-government attempts to alter its net financial assets through its own actions, it can only do so in ways that cause income adjustments.

Suppose there is unemployment. This is evidence of a non-government desire for a higher level of net financial assets than is consistent with full employment given the government’s current fiscal policy settings. The non-government is powerless to realize this level of net saving alongside full employment through its own actions.

If the non-government attempts to increase its saving ratio, for instance, this will weaken demand and cause falling income unless the government counters the effect with greater deficit spending. Similarly, if capitalists cut back private investment, this will once again only increase non-government net saving in a way that causes negative income adjustments unless the effect is offset by deficit spending. So the consequence of non-government attempts to increase net saving are negative income adjustments.

Conversely, if the non-government steps up its autonomous expenditures, this will prevent the desired level of non-government net saving. The income growth will result in higher tax revenues, a reduction in the budget deficit, and lower net saving.

The two objectives – more net financial assets and full employment – cannot be met through non-government activity alone since getting closer to one goal moves the non-government further away from the other goal.

In contrast, the government, as issuer of the currency, is uniquely positioned to render non-government net saving behavior consistent with full employment. By increasing its deficit spending, the government can stimulate income growth alongside higher non-government net saving. This is because the effect of an exogenous increase in government spending or cut in tax rates more than offsets the endogenous impact of the higher income on tax revenue.

The above considerations illustrate that unemployment and unrealized non-government net saving occur whenever the government’s exogenous fiscal policy settings are incompatible with the behavior of the non-government. Even if the net saving behavior of the non-government is deemed inappropriate, resolution of the matter alongside full employment will require alterations in tax rates and possibly the introduction of new taxes. Either way, the incompatibility can only be resolved through an appropriate exogenous change in fiscal policy, and this change can only be effected through vertical transactions.

Money is Endogenous

In light of the above discussion, ‘vertical money’/’horizontal money’ – or, equivalently, ‘government money’/’bank money’ – should not be interpreted as ‘exogenous money’/’endogenous money’. Government money and bank money are both endogenous.

The endogeneity of bank money follows from the analysis of the circuitists and horizontalists. On the basis of credit-worthiness, subject to capital constraints, private banks issue loans and create deposits (bank money). In doing so, they are not constrained by reserves. If caught short when it comes to settlement or meeting reserve requirements, they can always borrow in the interbank market or from the central bank, which stands ready to accommodate all such demands.

It is therefore impossible for the government, through its agent the central bank, to control the supply of bank money. All it can do is set the price of government money – in practice, usually the short-term interest rate, though it could set other rates as well if it wished – and accommodate whatever demand for liquidity eventuates. Since credit-worthiness, as assessed by the private banks, will be strongly influenced by the state of the economy and the prospects for income growth, the supply of bank money is endogenous.

The endogeneity of government money – in particular, high-powered money (HPM), which comprises currency and reserves – follows from the endogeneity of the budget outcome. Government deficits or surpluses result in an overall increase in reserves. If the central bank is targeting a positive interest rate, it issues or purchases bonds to drain or add reserves. The extent to which the central bank has to do this in response to fiscal actions is dictated by the budget outcome. If there is a zero interest-rate target, the build up in reserves is still influenced by the budget outcome.

Clearly, in view of the endogeneity of the budget outcome, both the level of net financial assets and the level of reserves (and currency) are endogenous. The level of net financial assets is always a reflection of budget outcomes; the level of reserves reflects both budget outcomes and other central bank balance sheet operations.


9 thoughts on “Vertical/Horizontal vs Exogenous/Endogenous

  1. “On the basis of credit-worthiness, subject to capital constraints,”

    Interesting that those capital constraints are again price, not quantity.

    Bank A can loan to Bank B to buy equity in bank C. Run that round in a circle enough times until its not worth doing it any more.

    All roads lead back to the central bank’s rate.

  2. Neil, that is too theoretical. In practice if you use equity as collateral then it is very close to having no collateral at all. Even if we forget the credit risk, there is still a market risk which requires capital.

  3. The reason the terminologies “horizontal” and “vertical” are not very appropriate is that Basil Moore used the same phrase in his fantastic 1988 book, “Horizontalists and Verticalists”, where he used the phrases for a simplified description of the Post-Keynesian versus Neoclassical money supply curve.

    “In MMT, government money is depicted as vertical whereas bank money is considered horizontal. The reason for this does not concern the question of endogeneity or exogeneity of money but rather the fact that both the level of reserves and the level of net financial assets (NFA) – the sum of currency, reserves, and government bonds – can only change through vertical transactions.”

    Well, that is not true really.

    Take a closed economy. Imagine the central bank drastically reduces or increases short term interest rates, thereby causing the whole yield curve to move. This induces revaluation gains/losses of the government bond holdings of the private sector and hence changes it’s NFA.

    The central bank is not doing any vertical transactions here than required to offset funds in and out of the banking sector – the movement is purely due to announcement and communication.

    The more accurate statement can be made using NAFA – the flow.

    Also, take an open economy. Now what changes NFA depends on which sectors one is consolidating. IMO, it is counterproductive to merge the domestic private and foreign sectors, and in that case where one does not, current account surpluses also increase the domestic private sector’s NFA.

    [Some commentators have cribbed that my first point (made in another blog) does not work out because government bonds are valued at par etc. I think it is a bit silly from SNA2008 viewpoint to do so. Imagine two bonds today – one issued in 2001 and maturing in 2031 and paying a coupon of 5% and another issued in 2011 and maturing in 2031 paying a coupon of 3%. It is ridiculous to value both using the same price. More importantly national accountants do not do that]

  4. Also, the usage of the phrase “capital constrained” is a bit misleading. Only during a credit crunch can this be an issue and not always so.

    Away from a credit crunch, banks have less issues in raising capital by equity issuance (to non-banks and not to each other!) and retaining enough profits. The constraint is again demand and credit-worthy borrowers.

    If capital adequacy is raised by regulators (in a non-hurtful way) or banks themselves, lending rate will rise but this is just a temporary phenomenon.

    In fact, there is some mainstream literature wondering why banks have excess capital (Diamond, Rajan et. al.). The authors don’t seem to realize that the answer lies in the Kaleckian-Keynesian principle of effective demand!

  5. “This induces revaluation gains/losses of the government bond holdings of the private sector and hence changes it’s NFA”

    No it doesn’t. The redemption yield is exactly the same as it was before.

    Market value is not value.

  6. “Also, the usage of the phrase “capital constrained” is a bit misleading. Only during a credit crunch can this be an issue and not always so.”

    And only under the current type of credit licence.

    Alter the credit licence to a fixed amount – say $Xbn and you’ll find them very capital constrained.

    The current settings of parameters a the system isn’t the entirety of the system. It’s just one instance of it.

  7. Why not simplify the terminology and refer to it as private credit expansion/public credit expansion. Public credit expansion occurs when the issuer expands the currency supply by running deficits. When government expenditures exceed revenues then by law the currency issuer must issue debt. Private credit expansion typically involves the issuer lowering interest rates. The private banking sector responds to the lower rates by extending more credit.

    Can’t a simplified definition be that public credit expansion increase government liabilities and private credit expansion does not?

Comments are closed.