Gavyn Davies has a column up over at the Financial Times, which I think is a masterpiece at integrating a number of threads from different economic schools of thought into a composite picture of the global economy that makes sense. His view is that slow wage growth, household debt and equity prices have been interrelated over the past thirty-five years due to economic policy responses and that it has been beneficial for shares. I want to riff off of his analysis below, adding some thoughts of my own.
I suggest you read the entire Davies article here but let me start out with the three most salient paragraphs:
Essentially, the argument is that lower real wages have increased inequality in the western economies, and this has depressed aggregate demand by redistributing real income and wealth away from the relatively poor towards the rich. Since the poor have a higher propensity to consume than the rich, this redistribution reduces consumer demand.
The decline in demand is then addressed by policy makers, either by fiscal expansion (reducing taxes and increasing subsidies on the poor) or by reducing interest rates set by the central banks. Since the fiscal response results in bigger budget deficits and higher public debt/GDP ratios, more and more of the burden of policy adjustment eventually falls on the monetary authorities. It is likely that this feedback loop will tend to increase both asset prices and inequality from one cycle to the next.
A pernicious additional consequence of this loop is that private sector debt/GDP ratios are also likely to rise through time. Falling real interest rates increase the incentive to borrow, while rising asset prices, especially in the housing market, increase credit worthiness and therefore the ability to borrow.
Debt ratios rise until they cause a crash, which of course is what occurred in 2008. This causes even greater and more permanent declines in real interest rates, which adds another twist to the cycle.
I think the linkages in this argument make a lot of sense. So let me flesh them out from my perspective. In the past thirty-five years, wage growth has stagnated across a broad swathe of advanced economies, including the United States, the United Kingdom and Germany where I have lived and have personal experience, and Japan, which was the first to experience secular stagnation. The fact that the wage stagnation is so widespread points to global phenomena as ultimate sources. A number of potential culprits come to mind, chief among them being global wage arbitrage. The ability to find workers abroad to do the same work as local labour has increased markedly for a number of reasons. There is the fall of the Iron Curtain and the integration of China into the global economy, which have added a huge supply of well-educated or cheap labour sources. There is also the Internet’s ability to replace local workers with offsite labour because work product data can be sent around the globe for pennies. And there is the falling power of unionized labour, which gives corporations more leverage over workers. All of these factors are going to depress wages where they are high and increase them in relative terms, where they are low until wages converge to a point where the arbitrage is too costly to consider given the risks associated with it like shipping costs, supply chains, language barriers, and different regulatory regimes.
Once we are in a situation in which advanced economy wages are depressed, you have a redistribution of income away from advanced economy wage earners toward emerging market wage earners and the advanced economy wealthy. In essence, the lower costs associated with global wage arbitrage are borne by advanced economy wage earners and then are recouped by a split between lower advanced economy prices, higher corporate profits and higher emerging market employment and wages. The lower prices do not fully compensate advanced economy wage earners for the stagnation or decline in wages, and so the economy stalls unless private debt can rise.
Let me give you an example away from the US and the UK that has resonance for me. Some 30-35 years ago, the “Made in China” wave started to pop up in manufactured goods sold in the United States. And as this wave advanced, whenever I went shopping, I got into the habit of flipping over items at stores just to see where the items were made. Now, because I lived in or visited Germany from the 1990s, I also did the same in Germany. And what I found was threefold. First, product labelling in Germany was less stringent than it was in the U.S.. Often, the sort of item that would have carried a label in the U.S. showing where it was made had no such label in Germany. So I couldn’t tell where everything was made. Second, when I did see a label, in the 1990s, the labels read “Made in (West) Germany” more often than not. Not only were these labels in English as if the product were always available for export even though it was being sold domestically, but there were many fewer products made outside of Germany in that time period than in the United States. Third, the shift away from Germany in labelling began in the late 1990s to early 2000s, and there I saw a shift toward Hungary, Czech Republic, Romania and other places in Eastern Europe for household wares like Braun shavers or blenders. And I also saw a shift toward China for toys like Schleich rubber miniature animals (which my daughter collected).
My takeaway here was that the global wage arbitrage may have come late to Germany but it came nonetheless. And so the forces bearing down on US and UK workers also had an impact in Germany as well. We know this from the difficulties in the Rhein-Ruhr region but it was interesting to see it play out in products for sale in stores.
The story I would tell here about depressed wage growth is slightly different than the one Gavyn tells because I see high interest rates as a major reason that a disproportionate amount of the policy response fell onto central banks. Whether inflation came down because of global wage arbitrage or because of other macro reasons, we did have a secular decrease in inflation from the early 1980s that translated into a secular decrease in interest rates. In the US at a minimum, this meant a de facto monetary policy asymmetry as rates came down quickly in a recession but never went up as far in recovery due in part to the drop in inflation. The result was lower debt service costs at any given level of debt burden, and as such, an increase in private debt burdens without an increase in debt service costs. Here, I am not ascribing causality to why debts increase but it is key to remember that wages stagnated at this point. In the U.S. real hourly wages were actually declining through the mid-1990s.
Clearly, stagnant incomes can be overcome by higher debt loads if interest rates and debt service costs are declining on a secular basis. Every time, interest rates went up cyclically during this period, so too did debt service costs. But a recession soon ensued and interest rates dropped to another secular low afterwards, easing the burden, and allowing households to increase indebtedness without debt stress.
Going back to Gavyn’s analysis, he makes a good point that the lower interest rates imply lower discount rates, and, therefore, higher equity valuations so that slow wage growth and higher household debt are inextricably linked to higher equity prices. Now, in Gavyn’s analysis, the wage share of GDP is key to understanding the secular bull market, even today. But I believe we are already in a secular bear market that began in 1998 or 2000 because of the zero lower bound. Basically, when discount rates hit what was then their nadir with the 1% Fed Funds rate in 2002, we were about as low as we were going to get in terms of discount rates. And so the wage share of GDP has become less relevant because every bust from here forward will be an event during which monetary policy is limited by the zero lower bound but private debt levels are still elevated and wage growth is stagnant. That’s a recipe for deleveraging of the kind we saw in 2007 and beyond. I believe the next downturn will also be one of pronounced deleveraging due to the zero lower bound. And that means lower GDP, lower profits, and lower equity prices. But of course, it probably also means lower price-earnings multiples, setting us up for a large decline in equity values.
So, my analysis says that the 35 year up trend in the profits share in GDP is not reversing any time soon for the same reasons Gavyn gives. However, I believe equity prices will decline or stagnate nonetheless because the zero lower bound limits monetary policy and fiscal policy is politically constrained. With wage growth slow and household debt elevated, it means cycle trough debt stress can only be relieved via deleveraging, resulting in a large cut in equity prices, so large that we should consider this one large secular bear market.
The country to look at to understand this thesis is Japan. They have gone through all of these stages already. Though it was corporate balance sheets in Japan that caused a deleveraging, the depressed GDP growth and the outsized decline in equity prices during recessions was so severe that equity markets declined outright from nearly 39,000 to almost 7,000 over two decades. If you look at Japan, the financial surplus of the corporate sector is large. In fact, it is so large as to be unwanted, as Martin Wolf pointed out last week in the FT. So we have the unsatisfactory co-existence of large corporate surpluses, stagnant wages and GDP plus a long period of declining equity prices. Japan may be breaking away from that right now, as equities are at a 7-year high. but the carnage since 1989 has been immense and there is no reason to believe the confluence of zero rates and no wage growth won’t create the same dynamic in North America and Western Europe. Right now, this may not be a concern. But if we are just in a cyclical bull market within a larger secular bear market, understanding these interconnections will be crucial for investors as the economy turns down.