US secular growth: donkey or racehorse?

I have been hacking on for several years about the downward pressures on US long-term growth prospects. What amazed me two years ago was to see the authorities, including the Fed, estimating a nearly 3% trend for the US, which seemed to me then and now as impossible. Negligible growth in population and man hours offered to the workforce is the most important brake to growth, with a net drop of a full 1% from the pre-2000 trend. 

Less capital investment and growing income inequality do not help. But the most underappreciated important factor, in my opinion, is the drag on growth from the loss of sustained cheap energy, as oil has moved from a $16/barrel 100-year trend pre-1972 to today’s approximate $75/barrel trend price. The GFC, in contrast, was a temporary factor and one that I believe (on my own, apparently) was given an exaggerated importance. Given these negative factors, I estimated a few years ago that the US trend line growth for GDP was likely to be no higher than 1.5% a year, and perhaps only 1% for Europe and Japan. Well, wheels turn and estimates are re-estimated: official estimates for the longer-term growth trend of the US have been falling slowly but surely over the past few years to a range from about 2% to 2.5%. In the two or three years since 2012, I had expected to see them in the range of 1.5% to 2%. 

Well, into this quiet world of creeping adjustment, an IMF paper released in early April this year acted as an unexpected jolt of excitement as, unusually, estimates tumbled all the way down to 1.5%. Wonders never cease. Now, the question is how much will this affect the Fed’s beliefs? Presumably enough to matter. This would be timely because, as you may remember, I have been anxious about the Fed’s whipping our actual 1.5% donkey in the mistaken belief that it was a 3% racehorse. The danger was, as I said, that they would keep on whipping it until either the donkey turned into a racehorse or dropped dead. Death from overstimulation. 

Not only has the IMF paper been a necessary gust of reality that might just convince Ms. Yellen that she is indeed dealing only with a humble donkey, but it has also raised some further interesting questions. It made its main point the reduced rate of growth and the ageing of the workforce. How, by the way, does this point, straight from the US Bureau of Census, take over five years to make it into semi-official GDP growth estimates? It then references lower capital investment ratios in a traditional way; also obvious enough. But what does it leave out? Resource limitations. I like to joke that the only thing that unites Austrians and Keynesians is their complete disregard for the limitations imposed by Spaceship Earth. In their view, a dramatic increase in price trend from the old $16/barrel to the new $75/barrel had no effect. Mainstream economics continues to represent our economic system as made up of capital, labour and a perpetual motion machine. It apparently does not need resources, finite or otherwise. Mainstream economics is generous in its assumptions. Just as it assumes market efficiency and perpetually rational economic players, feeling no compulsion to reconcile the data of an inconvenient real world, so it also assumes away any long-term resource problems. “It’s just a question of price.” Yes, but one day just a price that a workable economy simply can’t afford. 

I am still just about certain about three things: first, our secular growth rate in the US is indeed about 1.5% (at least as stated in traditional GDP accounting, wherein expensive barrels of oil increase GDP; perhaps closer to 1% in real life); second, economists move their estimates slowly and carefully in order to stay near the pack and minimise career risk (despite the recent IMF heroics); and third, that we do not like to give or receive bad news and, when in doubt, we tend to be optimistic. So, the US market is still not bubbling yet, but I think it will.

The key point here is that in our strange, manipulated world, as long as the Fed is on the side of a strong market, there is considerable hope for the bulls. In the Greenspan/ Bernanke/ Yellen era, the Fed historically did not stop its asset price pushing until full-fledged bubbles had occurred, as they did in US growth stocks in 2000 and in US housing in 2006.

Both of these were in fact stunning three-sigma events, by far the biggest equity bubble and housing bubble in US history. Yellen, like both of her predecessors, has bragged about the Fed’s role in pushing up asset prices in order to get a wealth effect.

Thus far, she seems to also share the Fed’s view on feeling no responsibility to interfere with any asset bubble that may form. For me, recognising the power of the Fed to move assets (although desperately limited power to boost the economy), it seems logical to assume that absent a major international economic accident, the current Fed is bound and determined to continue stimulating asset prices until we once again have a full-fledged bubble. And we are not there yet. 

We at GMO still believe that bubble territory for the S&P 500 is about 2,250 on our traditional assumption that a two-sigma event, based on historical price data only, is a good definition of a bubble. (As we like to describe it, arbitrary but reasonable, for it fits the historical patterns nicely.) 

For the record, probably the best two measures of market value — Shiller P/E and Tobin’s Q — have moved up over the past six months to 1.5 and 1.8 standard deviations (sigma), respectively. So, just as with the price-only series, they are also well on the way to bubble-dom but, clearly, not there yet.

If we used these value series instead of just price, it would add 5-10% to the bubble threshold, further improving my case that the current market still has a way to go before reaching bubble territory. Historically, we have often used the price series as both less judgemental than using measures of value, and as a much fairer comparison with other bubbles (for example, commodities, currencies and housing). 

Bear in mind that in the Greenspan era (from 1987 and still running), the market, measured by Shiller P/E, has averaged about 60% higher than it did prior to when the Fed’s habit (as in, addiction) of pushing the market up in order to get a wealth effect became standard operating procedure. It would be tempting to adjust the “normal” P/E downward for this, and treat it as a temporary anomalous phase, but we do not — we give each year equal weight. 

Perhaps most importantly, the current economic cycle still seems only middle-aged, despite its measured long duration. For example, there is still plenty of available labour hiding in the participation rate for sure (male labour anyway). It turns out that even the female labour force probably still holds substantial potential, for after a quick surge in the US back in the 60s through the 90s, in recent years the female participation rate has pulled back, while in other developed economies it has continued to increase. We are now well behind the average of the other 22 developed countries. So, we are in no danger at all of running out of labour. 

There is also room for an increase in capital spending, which has been quiet for years. The current capex-to-GDP ratio is still depressed below any level from 1950 through 2007. A steady rise in this capex ratio, particularly if coupled with an uptick in real hourly wages, is what is needed to inject more life into this slow-moving economic cycle. 

The lower oil price should also have more of a stimulus effect for the US as time passes. I argued in the last two quarters, however, that globally (in contrast to the US) there would be surprisingly little benefit from lower-priced oil, and that in GDP terms, as opposed to reality, it might well even be negative. 

The US housing market in terms of houses built is still way below the old average, and house prices are only around long-term fair value; there is room for improvement in both in the next two years. It would, in my opinion, be odd to have a Fed-driven cycle end before the economy is working more or less flat out as it was in 1929, 2000, and 2007, to take the three other biggest equity bubbles of the last 100 years. I still believe that before this cycle ends, the quantity of US deals, including co-investments, should rise to a record given the unprecedented low rates and the current extreme reluctance to make new investments in plant and equipment (how old-fashioned that sounds these days) rather than into stock buy-backs, which may be good for corporate officers and stockholders but is bad for GDP growth and employment and, hence, wages. 

We could easily, of course, have a normal, modest bear market, down 10-20%, given all of the global troubles that we have. If we do, then the odds of this super-cycle bull market lasting until the election would go from pretty good to even better. So, “2,250, here we come” is still my view of the most likely track, but foreign markets are of course to be preferred if you believe our numbers. Stay tuned.

Edited extract from GMO’s quarterly letter