Yesterday was a big conference day in Washington, D.C., with the IMF and World Bank meeting and with the Brookings Institution holding an event featuring the Greek and German finance ministers. While I attended the Brookings event, the event I thought most interesting was the 24th Annual Hyman P. Minsky Conference at the National Press Club held by the Levy Economics Institute. And I want to make this the basis of today’s newsletter entry.
What I am doing today is building a mental model for forecasting the future of the U.S. economy based on a very few economic and political variables and the presentations by Daniel Alpert of Westwood Capital, Lakshman Achuthan of ECRI, Bob Barbera of Johns Hopkins and Levy’s President Dimitri Papadimitriou. I am going to use the Wynne Godley sectoral balances approach and the CBO figures from Papadimitriou’s presentation as a baseline. This is not an ‘economic model’ but a mental model of how to think about macro outcomes in the U.S.
I think the Godley model of using sectoral balances to tease out flows between sectors of an economy and between economies is a very good approach in understanding how an economy operates. I will use this approach to demonstrate constraints that lead me to believe that the U.S. economy is entering a period in which a Japanese outcome has a high likelihood of occurring.
First let me present the assumptions.
- Monetary offset: in this mental model, I am assuming that the Federal Reserve will do its utmost to pursue its dual mandate of price stability and full employment while maintaining the value of the currency and keeping a watchful eye on financial stability. Thus, to the degree fiscal policy is restrictive, the Fed will offset this restrictiveness via policy accommodation. And to the degree fiscal policy is too loose, monetary policy will offset this looseness via policy tightening.
- Fiscal balance endogeneity: a core assumption here is that the fiscal balance is mostly endogenous. And by this I mean that any linear regression of employment or nominal GDP growth outcomes and the fiscal balance will have a strong positive correlation in any country. The fiscal balance, then, is largely determined by a mix of prior government spending and taxation choices, the private sector’s desire to save net of investment and the trade balance with influence from the currency. The fiscal balance is endogenous in the sense that it will automatically move up when macro outcomes improve and it will automatically move down when they worsen.
- Fiscal determinism: this is a term I made up yesterday to denote the view that fiscal outcomes are targets as well as outcomes. A real world example of precise fiscal determinism is the stability and growth pact in the eurozone which targets fiscal deficits no higher than 3% with government debt shrinking toward a goal of 60% or under that level. Fiscal determinism takes an endogenous fiscal outcome and creates reflexivity by trying to make this outcome an exogenous target variable. And this will have unintended consequences that can be measured via the sectoral balances approach.
- Monetary policy ‘derivative-ness’: a final assumption I am making here is that monetary policy acts primarily through credit, asset and portfolio balance channels. Thus, monetary policy only has a ‘derivative’ impact on the real economy. Thus, the policy mix from a specific level of monetary offset can be relatively more or less stimulative to asset and credit markets than the real economy. (See the ECB’s mental model of how this works. This is not my own model but it shows that the ECB recognizes that monetary policy has no direct impact on the real economy.)
Now, having laid out these assumptions, I want to get to the Papadimitriou/CBO stock-flow modeling figures. The CBO assumes budget deficits in the range of 2.5 to 2.8% through 2018 from an actual 2.8% figure.
(Source: CBO)
If the CBO estimates were true, then we would need to see the fiscal balance divided between the current account and the private sector’s saving net of investment. This is how the flows looked from 1990 to 2012 minus the state and local government contribution.
Given the present US current account deficit as a percentage of GDP is about the same as the federal deficit, this would imply a private savings level net of investment of about zero.
(Source: Trading Economics)
Thus, for the CBO projections to hold, you would either need to see a surge in capital investment to reduce private savings or a surge in household debt. The first outcome is contrary to what we have seen so far in this recovery and thus assumes an increase in capital investment which has to be predicated on wage growth that translates into robust consumer demand and GDP growth which mandates more capital investment. The second outcome is one Dan Alpert discussed in noting that nominal household debt levels now are only 4% lower than they were at the peak before the financial crisis and that household debt a s a percentage of GDP is well above the levels from the last two economic upturns.
For the first scenario to hold then, you would need a level of growth which would increase government tax receipts much more than the CBO forecasts and would be predicated on wage growth taking off. For the second scenario to hold, we would need a full scale return to the asset-based economy in which households re-levered, introducing a renewed level of financial fragility.
Now, if the U.S. economy remains on an upward path, I believe the fiscal balance will be more positive due to fiscal determinism reducing outlays. But then there might be enough monetary policy offset to give the Fed policy space to deal with a downturn following the upturn’s credit and asset price growth.
However, I see the recovery policy mix as more heavily geared toward asset prices than the real economy because asset prices are already elevated today. If the CBO forecasts were to pan out, we would be in an asset bubble scenario.
Moreover, monetary policy offset implies higher interest rates and a stronger dollar and thus weaker net exports and a larger current account deficit. Think of this as an outcome similar to the late 1990s or 2007.
Lakshman Achuthan talked about the Fed’s raising rates as if it were not a negative outcome. And the rest of the panel was incredulous because they did not believe the US economy was in a position to deal with hikes. But, if the CBO scenario actually came true, raising rates would be the appropriate outcome.
More likely, however, is that the CBO forecasts are not realistic because the sectoral balances make them so. Let me present an alternative scenario then.
In this scenario, the baseline economic outlook is weaker. Q1 was not an aberration and so GDP growth stalls as output remains below forecasts, inventories are not accumulated and capital investment growth is weak. That would mean monetary offset keeping interest rates relatively lower, the dollar is weaker than anticipated and the current account shifts in a more positive direction.
But fiscal determinism means that deficits stay low, just as the CBO has assumed. And as a result, recession comes earlier than anticipated, causing a shift toward increased private savings net of investment in both the household and corporate sector.
If you look at top line revenue growth at US companies right now, EPS and share price growth are completely out of line with this due to the four horsemen of the cyclical recovery (multiple expansion, stock buybacks, operating leverage, and cost cutting). That goes away and reverses in recession and means that capital investment drops. In the household sector, the still high debt levels will mean deleveraging and increased private savings.
The recession would mean a larger deficit, and depending on who is in power and how difficult the downturn is, fiscal determinism could be invoked again, adding reflexivity to the retrenchment already underway.
This is an outcome which is similar to 1997-98 Japan. And would be deflationary. Yesterday, Dan Alpert mentioned that the US CPI ex-housing is already in negative territory. So, a recession would mean that deflation would take hold and we would move into a Japanese situation.
The pernicious part of a Japanese outcome is the destruction of balance sheets at financial institutions due to a flat yield curve implied by permanent zero rates. Net interest margins would plummet and we would see credit growth stall as a result. Financial institutions would be more concerned with rollovers and work outs for existing borrowers in order not to impair their capital base. Existing borrowers could hang on because of low rates, creating a sort of zombie economy with low or negative GDP growth, weak credit growth and zombie corporates and banks.
None of this takes into account potential disruptions from abroad, of course. But this is just a simple baseline. My mental model for the U.S. economy gives a very narrow path to a Goldilocks outcome. The Scylla we are trying to avoid is Japanese-style secular stagnation of low nominal GDP growth, borderline or entrenched deflation and zombie banks and private debtors. The Charybdis we are trying to avoid is an asset-based bubble economy with the attendant financial fragility, stark deleveraging and financial crisis that come with it.
The narrow path between these two outcomes comes largely from a less monetary policy-heavy mix but we not be able to get there from this starting point. A given policy mix can be relatively more or less stimulative to asset and credit markets than the real economy depending upon how much monetary or fiscal policy is invoked to reach full employment. This particular recovery has been extremely heavy on monetary policy. And this is why signs of asset market excess are everywhere even though signs that the real economy is charging ahead have been limited to date.
The Goldilocks scenario is one in which the fed does begin to raise interest rates, restraining future asset price growth while wage growth continues upward to buttress the real economy and bring asset prices back into line. The problem I have believing this outcome is possible stems from the oversupply of labor relative to capital in a global economy. My question is how you get robust US wage growth in a world in which an increasing percentage of US wage laborers are competing with wage laborers from other countries where wage rates are much less. The US benefits from high levels of accumulated capital wealth that translate into productivity. But US wage laborers see an ever-decreasing portion of that productivity differential due to the weak bargaining position of US wage laborers relative to employers. And this true throughout advanced economies.
I should also note that the relative lack of bargaining power for labor and fiscal determinism necessarily combine to create a monetary-heavy policy mix. Weak wage growth implies weak nominal GDP growth unless household debt rises. And a monetary-heavy policy mix allows this to occur.
In sum, the reason the asset-based model even existed is that it allowed consumption and nominal GDP growth to continue unabated in the face of weak US and developed economy wage growth as wage benefits passed to emerging markets workers. Germany has adopted the only other viable model given this outcome, and that is to bolster nominal GDP via a current account surplus, something the US cannot do given its currency’s status as the world’s reserve currency and the Triffin dilemma mandates current account deficits.
With current account deficits entrenched or even increasing during cyclical upswings, due to a strong dollar, fiscal determinism that mandates lower deficits as a target implies private sector dissaving and creates a bias toward boom-bust and financial fragility. We seem to be at the end of the line on this score, with weak nominal GDP growth relatively entrenched. And weaker nominal GDP growth implies greater susceptibility to recession due to the lower baseline growth levels being much closer to stall speed.