Les dépenses d’investissement vont-elles repartir ?
- The leadership change in the German centre-left may strengthen some expectations in the market of a significant fiscal push in Berlin next year. We are more cautious. More generally, as the political situation in Germany becomes more uncertain, chances of a swift progress on the numerous issues pending on the institutional set-up of the Euro area are declining in our opinion.
- Diminishing uncertainty on trade may not be enough to re-start capex in the US in the medium run. We are concerned by the deterioration in corporate profitability, against a background of low productivity growth and lower policy capacity to continue to prop it up.
- Low real interest rates for long can lead to capital and labour misallocation towards low-productivity sector such as construction, impairing aggregate trend growth. We look at those mechanisms using the “Spanish housing boom” before 2007 as a natural experiment and apply them to the current German situation.
Don’t count on too much policy action in Europe
The European commentariat is likely to focus today on the victory of the left-wing candidates in the German Social Democratic Party (SPD) leadership race announced on Saturday night. We have seen in the sell-side literature (in particular from equity analysts) expectations of a decisive fiscal push in Germany rising in the last few weeks. Given Norbert Walter-Borjans and Saskia Esken’s support for a more ambitious fiscal policy, our best guess is that those expectations will strengthen further. For our part, while our baseline is that Berlin will dispose of the “black zero” in 2020 and will let automatic stabilisers play, we have remained very prudent on the likelihood and magnitude of a proper discretionary stimulus. The change at the top of the junior party in the Berlin coalition does not necessarily change this.
True, after his failure to take control of his party, Finance Minister Scholz is finding himself in a much weaker position, and with him the extra-cautious approach to fiscal policy. The new SPD leadership will probably push for more fiscal leeway as the review of the coalition agreement is looming. Beyond the tension within the ruling coalition, the policy debate in Germany has shifted a gear lately and the Social-Democrats could now join forces with the Greens in calling for a change of stance. But equally, centre-right CDU-CSU parties have recently made their opposition to fiscal activism even more explicit than before. Finding common ground is not going to be easy, especially since Walter-Borjans and Esken are reserved, to say the least, on the very notion of continuing the coalition. Disagreeing on public finances could actually be a handy pretext to end the coalition.
If early elections were to be the solution to break a gridlock – a big “if” at this stage - the number of potential combinations, on the basis of current polls, is high without clear indications on the outcome for fiscal policy. If conversely after a coalition implosion Germany were to be led by a centre-right minority government, a fairly orthodox fiscal policy would probably be implemented.
Beyond the fiscal question, the change of SPD leadership makes it even more difficult for Berlin to take a leading role in the European debate, with more time and energy devoted to domestic issues. The only recent initiative from Berlin, namely by Olaf Scholz, a new conditional push towards a common guarantee of bank deposits, was anyway a non-starter in our view (essentially a way to deflect pressure from fiscal policy with too many red lines to make it acceptable to other member states). Ursula Von der Leyen and Christine Lagarde are starting their new jobs with lots of ambition, but apart from the French government it is unlikely who in the EU’s member states will be willing and able to deliver on these ambitions.
Investment needs to be profitable!
We have made the point several times in Macrocast that the main transmission of the trade war to the US economy has been through heightened uncertainty taking corporate investment down. Symmetrically, assuming uncertainty recedes as a “phase one deal” is nearing between the US and China, could we expect a re-acceleration in capex? A knee-jerk, short-term improvement is plausible, but we also think there are endogenous forces which will cap any rebound.
Our focus here is the financial position of the US corporate sector. Leveraging continues at a fairly fast clip. This is usually seen as a concern in terms of financial stability. Our own view is that as long as monetary policy is very accommodative and keep debt servicing costs low, the risks remain in check from that angle – at least for the immediate future. “Zombification” is not a good thing in the long run, but by definition it comes with a reduced likelihood of a massive rise in bankruptcies. What concerns us more from the point of view of fostering economic growth is the juxtaposition of this high and rising level of debt with a decline in corporate profitability (Exhibit 1), here measured in its simplest way (value added minus labour costs).
Exhibit 1 – Watch corporate Americs’s numbers
Investment decisions are made taking into account the cost and availability of capital (whether it is funded on internal resources, debt or equity) and demand but also expected profits. Several years of actual decline in profitability reduce both the free cash flows available for investment and depress such expectations. More fundamentally we think that it is very unlikely we can have at the same time a rebound in profitability and the continuation of decent consumer spending. Very simplistically, it is either investment or consumption, one cannot get both without a positive shock to productivity and/or additional policy support.
We use the US integrated macroeconomic accounts to calculate various measures of profits in the non-financial corporate sector. The deep blue line at the bottom in Exhibit 2 is what national accounts call “corporate disposable income”. It is what remains of value added once labour costs, interests, dividends, tax and the consumption of fixed capital have been taken away. These are the resources firms can use to spend on investment. This ratio has halved since a recent peak just after the Great Recession in 2011. True, one could still take a relatively benign view since it is still – just – above the 2007 level and has stabilised over the last two years. However, without the support of the massive tax cuts offered by Donald Trump, profits would have continued to fall over the last three years (orange line in the middle).
Exhibit 2 – Without the Fed's stimulus and Trump's tax cuts, profits are back to a recent trough
Exhibit 3 – “Goodwin cycles” appearing after years of stability in the wage/profit shares of US output
Finally, on top of taking out tax we also corrected profits from the impact of the decline in interest rates (pale blue line at the top). In this case, the deterioration in profitability is continuing at a fairly quick pace and we have reached the lowest level observed over the 15 last years. This is the “spontaneous” profit ratio, i.e. before economic policy intervenes on taxation and debt servicing costs. The combined dynamics of labour costs and productivity are its main driver.
We want here to revisit a theoretical model designed by Richard Goodwin in 1967. He took his inspiration from the predator/prey models developed in biology, in which the population of antagonistic but complementary animals moves in cycles. According to Goodwin, employers and employees compete for the same scarce resource: the firm’s value added. Productivity is determined exogenously by the state of technology. As demand improves, the unemployment rate falls and wages accelerate (as per the Philips curve), to the point at which they exceed productivity growth. There is no other solution than allowing profits to fall, which in turn depletes investment and gradually triggers a slowdown in economic activity. The downturn then reduces the employees’ bargaining power, allowing wage growth to move again below productivity growth, restoring profits and investment. The cycle can be slow, given the institutional rigidities in wage negotiations and information asymmetries: employers know in real-time the level of productivity. Employees don’t.
The “only” problem of the Goodwin model and its exotic late-sixties Marxian flavour is that its predictions were continuously disproved by the actual pattern of wages and profits in the United States. Indeed, until quite recently the profit share did not move in cycles but was merely displaying what looked like random noise around a very stable average (see Exhibit 3). This may be explained by the fact that productivity growth is often pro-cyclical: periods of fast economic activity coincide with higher wages but also a more intense use of the workforce which helps protect the share of profits. Productivity is not 100% exogenous.
However the pattern has changed since the early 2000s. Commentators usually focus on the fairly regular decline observed between 2001 and 2011 to levels unseen since data became available just after the end of the second world war, suggesting a structural shift in the labour/capital allocation of income. This may actually be true, but the rebound since 2012 is for us equally interesting. We think a “Goodwin narrative” could fit the pattern relatively well. The recession of 2001 occurred after years of “productivity euphoria”, a belief that new information technology would trigger an explosion in productivity. The dot come bubble burst challenged these expectations and wage moderation settled in, actually exceeding what was required by the actual slowdown in productivity. By 2007 the share of wages had started to catch up again, following the improvement in the labour market, but this was stopped in its tracks by the Great Recession of 2008-2009.
Productivity growth has not rebounded firmly since then, but wage moderation has gradually abated in line with the decline in unemployment, re-growing the share of wages back. As we suggested above, this was hidden for some time by the tax cuts and the drop in interest rates, but policy space is now missing to provide the same quantum of support.
Many commentators as well as the Fed have taken comfort in the resilience of consumer spending in the US. However, this resilience is essentially the product of a rise in labour costs which, given the current pace of productivity, is detrimental to corporate profits. In terms of growth dynamics, this is normally consistent with a continuation of the current mediocrity in capex. This may also squeeze the capacity of firms to continue paying decent dividends and/or engage in equity buy-back programmes. The other adjustment valve, obviously, is another round of wage moderation and or a decline in job creation.
Looking into the cost of low interest rates via the labour reallocation channel
The ECB’s negative interest rate policy is under severe criticism in Germany. One prominent argument is that it is fuelling a bubble in house prices. The traditional response of central banks to this is that such adverse side-effects are better addressed by macro-prudential measures, for instance via taxation or modulating the capital requirements of banks, than by altering the monetary policy stance. Things would be different though if on top of the financial stability risks, a housing bubble would have a detrimental impact on potential GDP growth.
The usual channel for a unfavourable effect of low interest rates on trend GDP growth is “zombification”, i.e. the fact that too many unproductive firms are allowed to survive thanks to low debt servicing costs, thus depressing aggregate productivity growth. This is notoriously difficult to quantify. Here we investigate the downside of negative rates from another angle: the re-allocation of labour to the most interest rate-sensitive sectors. Construction is one of those. A lasting drop in real interest rates should push up activity and prices in real estate, attracting an increasing share of labour and capital. An issue from a macroeconomic point of view is that an inflating construction sector is detrimental to potential growth: capital crystallised in dwellings does not provide much basis for future production, and the level of labour productivity in construction is low.
We take here the example of Spain between the beginning of monetary union in 1999 and the bubble burst of 2008. The convergence towards the core countries went faster for interest rates than for inflation (between 1998 and 2003 inflation averaged 1.8% in the Euro area but 3% in Spain). The decline in real interest rates unsurprisingly triggered a massive housing boom, during which the share of construction in total employment rose by four percentage point. While on the surface Spain was outperforming most of its European competitors in aggregate GDP growth, in reality it was building up a very significant “productivity deficit” which would force a massive downside revision in potential GDP growth once the bubble burst.
In Exhibit 4 we build an “alternative history” scenario in which we keep the share of construction in total employment constant at their 1995 level, so that job creation after that date goes into the more productive sectors. GDP is then the result of the actual productivity levels of each sector multiplied by the alternative figures for employment per sector At the peak of the housing boom, between 2003 and 2007, GDP growth would have been 0.3% higher annually without the labour re-allocation. This is the effect of both the initial difference in productivity levels between construction and the rest of the economy and the fact that productivity fell faster in construction than in the rest of the economy at the time of the boom.
Of course this is limited approach, in particular because it does not take into account the spill-over effects (the housing boom fuelled growth in the rest of the economy, for instance by raising the value of collateral), but we think it provides an interesting order of magnitude of the “reallocation cost” to potential GDP growth. Another lesson from the Spanish experience is that those costs to “underlying growth” are difficult to come into focus in real time, since the economy is usually enjoying a strong acceleration in actual GDP during the housing boom. Observers have to realise the “quality” of economic growth is deteriorating, but his may be a second-order issue for policy-makers. This is exactly what happened to the Spanish authorities of the time.
Exhibit 4 – Building a counter-factual: Spain without the pre-2008 housing boom
Exhibit 5 – Exploring similar scenarios for Germany
We replicate this for Germany, imposing a gradual shift in the share of labour employed in construction of 4 percentage point over the coming decade. To break the effect down we provide in scenario 1 a static analysis in which productivity levels across sectors are “frozen” at their 2018 level and GDP growth is only hampered by the fact that a rising proportion of workers are employed in the lower-productivity construction sector. The difference in GDP over 10 years is almost invisible. In scenario 2, on top of the shift in the initial level we replicate over the decade the decline in productivity seen in construction at the time of the housing boom in Spain. In this case over 10 years GDP growth would fall by 0.2% annually. This is not insignificant for an economy where potential GDP growth is currently estimated at 1.2% per annum.
Note however that by using Spain as a “natural experiment” we picked a very extreme case which we don’t think would be easily replicated in Germany. Indeed, on top of a rise in domestic-driven construction, as the drop in real interest rates was fuelling a rise in the housing purchasing power of residents, the Spanish housing boom was also fuelled by massive inflows of foreign capital trying to take advantage of the country’s tourism appeal. This, together with different lending habits, a deeply-ingrained preference for renting over owning and lower tourism potential, makes it very unlikely that Germany would go on a similar path.
We think in the medium to long run negative rates are more a headwind than a tailwind, but they do not pose an immediate risk to growth dynamics in the Euro area. Their side-effects, via “zombification” or re-allocation of production factors to the most leveraging-sensitive sectors of the economy, take time to materialise. The key issue thus for us is to create a credible path for exiting this state of affairs within the next few years. We continue to think that an exogenous shock, in the form of higher public spending in the countries where fiscal space exists, is the most efficient approach.
Mon: ISM mfg index, final mfg PMI; Wed: ADP employment change, ISM non-mfg index; Thu: trade balance, Fri: non-farm payrolls, unemployment, prel. Michigan consumer sentiment
Mon: Final Eurozone mfg PMI, ECB President Lagarde testifies to European Parliament. Wed: final Eurozone composite PMI; Thu: final Eurozone GDP; Fri: German industrial production
Mon: final mfg PMI; Wed: final composite PMI, final services PMI; Fri: BBC Johnson-Corbyn debate
Mon: Caixin mfg PMI; Wed: Caixin non-mfg PMI
Mon: final mfg PMI, Tue: final services PMI
This communication is intended for professional adviser use only and should not be relied upon by retail clients. Circulation must be restricted accordingly.
Issued by AXA Investment Managers UK Limited which is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales No: 01431068 Registered Office is 7 Newgate Street, London, EC1A 7NX. A member of the Investment Management Association. Telephone calls may be recorded or monitored for quality.
Information relating to investments may have been based on research and analysis undertaken or procured by AXA Investment Managers UK Limited for its own purposes and may have been made available to other members of the AXA Investment Managers Group who in turn may have acted upon it. This material should not be regarded as an offer, solicitation, invitation or recommendation to subscribe for any AXA investment service or product and is provided to you for information purposes only. The views expressed do not constitute investment advice and do not necessarily represent the views of any company within the AXA Investment Managers Group and may be subject to change without notice. No representation or warranty (including liability towards third parties), express or implied, is made as to the accuracy, reliability or completeness of the information contained herein.
Past performance is not a guide to future performance. The value of investments and the income from them can fluctuate and investors may not get back the amount originally invested. Changes in exchange rates will affect the value of investments made overseas. Investments in newer markets and smaller companies offer the possibility of higher returns but may also involve a higher degree of risk.
© AXA Investment Managers 2019. All rights reserved