When Albert Edwards predicted in late 2016 that a surge in wage inflation was imminent, we were confused by this prediction from the world's preeminent deflationist: after all, not only had not a single economic indicator validated a tighter labor market despite unemployment just above 4%, but as we have have repeatedly demonstrated what little wage inflation existed, was attributable to managerial-level, supervisory positions while the bulk of job creation remained with minimum-wage jobs, which have continued to see virtually no wage growth. Even Morgan Stanley, a far greater bull than Edwards, one month ago admitted that "wage growth is leveling off, may be slowing."
Which is why we have to give Edwards credit: some 6 months after his initial call, he had the courage to do what is never easy and admit he was wrong, and that contrary to his expectations wages are not going up after all.
Talking about wrong, I have to put my hands up. I have been expecting US wage inflation to roar ahead over the past three months to well above 3%, yet every data release has surprised on the downside. Wage inflation, as measured by average hourly earnings, has actually levelled off at close to 2½% while wage inflation for ‘the workers’ is actually slowing (see chart below)! Strictly speaking, "the workers" are defined (by the BLS) as "those who are not primarily employed to direct, supervise, or plan the work of others. Hey, that's me!
So with the concession aside, Edwards is left with even more question, starting with "What on earth is going on with US average hourly earnings?"
Three consecutive Employment Reports have seen this key measure of wage inflation surprise by its weakness. I feel especially foolish as I had written that wages were set to accelerate sharply, forcing the Fed to tighten aggressively and thereby driving both bond yields and the dollar higher. Doh! While many commentators last year, including the Fed, expressed surprise that US wage inflation had been so quiescent despite a tight labour market, I thought there was a simple explanation. I believe that nominal wages had not accelerated more rapidly through 2016 primarily because headline CPI inflation had been so subdued, staying in a 0-1% range for most of the last couple of years. Hence nominal wages did not need to accelerate rapidly for workers to be much better off as 2-2½% nominal wage inflation translated into strong real wage rises of around 1½-2% - the most rapid for years (see circled area in chart below).
As headline CPI inflation surged this past six months, rapid real wage growth turned into real wage stagnation (see chart above). I believed that a tight labour market would prompt an aggressive reaction from "the workers" to maintain the previous 1½-2% rate of real wage inflation they had enjoyed and got used to through 2015 and 1H 2016. Hence I expected nominal wage inflation would roar upwards in 1Q this year. How wrong I was!
There is even more confusion in the data, because Edwards points out another disconnect: while the BLS' measure of hourly earnings has gone nowhere, and real earnings have in fact tumbled, the employment cost index has spiked, "with wage and Salaries jumping from a 0.5% rise in 4Q to rise by 0.8% in 1Q 2017 ? the fastest quarterly rise since 2007. On a yoy basis, this measure of wage inflation still showed a 45 degree upward trajectory into 1Q 2017 (see left-hand chart below). Adding benefits to wages and salaries, total compensation also rose by 2½%."
Then there is the issue of declining productivity, because when calculating productivity and unit labour cost growth, the BLS estimates non-farm businesses saw their workers compensation jump from the 3% average rate seen in 2016 to just shy of 4% yoy in 1Q 2017?. This has implications on corporate profits:
"Together with sluggish 1% productivity growth, this means that unit labour costs are rising by almost 3% yoy, well in advance of the rate by which corporates are able to raise their output prices (see right-hand chart above). The bottom line is that US corporate margins are suffering a savage squeeze and have been for some time. What then do I make of the heady 1Q company reporting round? Not much."
Perhaps in retrospect, between the divergent AHE and ECI data, Edwards was not entirely wrong, as he suggests:
The truth is that the closely watched average hourly earnings measure of wage inflation has not accelerated in response to a surge in headline CPI in the way I had expected. So strictly speaking I have been wrong and as such I must throw myself upon your bountiful mercy. But let me say in my defence that other measures of wage inflation have shown exactly the acceleration I had expected. The fight-back by labour to secure their rightful share of the economic pie is ongoing, but it seems likely that the savage downward trend in the share of labour compensation that had been in place since the 2001 recession seems to have at last been broken (see chart below). The laws of economics have not been abolished after all ? at least not in the US.
Yet while the jury may still be out on US wages between two contradictory data sets from the BLS, when one looks outside the US, things are clear: despite years of QE, there is no wage growth. For evidence, look no further than Japan. Edwards again:
Japan is becoming an economic enigma. Last week saw some truly astonishingly weak wage inflation data ? so weak that it sent the yen sharply lower on expectations that the Bank of Japan might need to step up their already ridiculously outsized QE programme to even higher levels. Wages for March fell by 0.4% yoy, well below both the expected 0.5% gain and February's 0.4% rise. Even the far less erratic underlying wages (excluding overtime and bonus payments) weakened sharply and declined yoy in March. In real terms, total cash earnings were miserable too, falling by 0.8% from a flat reading in February (see charts below). Certainly on this measure Abenomics has been a total and utter failure.
The idea was simple, QE (or QQE as the Japanese call it) would as an indirect consequence send the yen sharply lower (as it did in 2013/14), which would push up headline CPI inflation (also buoyed by the 2014 VAT hike) and drive wages higher in what was a tight labour market.
And when I say tight, I mean properly tight. This is not the US, where most commentators agree there is likely to be more slack than the low headline unemployment numbers suggest due to the sharp decline in the participation rate since the last recession. By contrast, the Japanese labour market is unambiguously as tight as it ever has been in history (see left-hand chart below). Yet wage inflation remains moribund.
Without any real cost-push wage pressures, and with the initial inflationary impulse on headline and core CPI of the declining yen of 2013-14 receding into a distant memory, core CPI inflation (ex food and energy) has begun to fall once again (for this see right-hand chart above, and note that headline and CPI ex-food are rising moderately only because the yoy impact of the oil price has gone from negative last year to positive this year). So after all the trillions of dollars of QE and huffing and puffing, Abenomics has failed to deliver its much touted exit from the deflationary mire.
And before readers respond with "there is always more QE", the problem is that for both the ECB and BOJ, the answer is increasingly, "there isn't" as both central banks are just months away from running out of eligible bonds to buy, beyond which point the entire bond market may simply lock up, or the central banks will have to even more actively start buying equities, with both outcomes effectively a nationalization of capital markets. And the last time we checked with the USSR, that strategy did not work out too well...