Sometimes data revisions matter

Sometimes data revisions matter

05-08-2019 | インサイト
Sometimes data revisions really matter – this is one of those times. Recent revisions to US GDP data shed a different light on inflation data, and on the resilience of corporate profits.
  • Rikkert  Scholten
    Portfolio Manager

Speed read

  • Downward revisions in US corporate profit data for the past three years…
  • …are accompanied by upward revision in labor’s share of GDP
  • The new data implies cyclical mechanisms still work

The rewards are in the details

Data revisions are not everyone's cup of tea and some may question their relevance given our forward-looking nature. So, why bother?

In some cases it is worth the effort of looking into the details. That is clearly the case for the recent revisions of US GDP data for the 2014-2018 period. The interesting information is not on the front page, but in the national income revision and the details provided on page 47 of the BEA report (thanks Will Denyer from Gavekal for pointing that out). The BEA revised its calculation of corporate profits downwards for the three years 2016 to 2018, by 1.2%, 4.4% and 8.4% respectively. That amounts to a hefty USD 188bn for 2018.

Where did the money go? The adjustment is allocated to net interest expenses and labor compensation. The latter is particularly interesting.


Labor not doing so badly after all

Economists have spent countless hours in recent years debating why the share of US labor income has not risen in this current expansion. To be fair, the income share is still low and the downtrend remains intact, but after the revision, the cyclical dynamic of the data suddenly looks a lot more normal.

Instead of having remained stuck for the past few years, compensation now shows a decent move upwards (See Figure 1). The cyclical upswing from the low in 2015 is now 3.5 percentage points, which is in line with the rises experienced in 2006-2008 and 1996-2000.

Figure 1 | Labor compensation as % of US total value added

Source: Bloomberg

It may well be that the Phillips curve (linking inflation to unemployment) has been declared dead, but it seems that some of the cyclical mechanics of sharing the income pie are still intact.

What is the upshot, then? The implications certainly are not as dramatic as an upward revision in the inflation numbers, but this new data does suggest that the risks of an unexpected rise in inflation have increased, and that corporate earnings are not immune to cyclical trends.

We can also look at the profit-to-GDP levels after the revision, as indicated in the charts below. The below chart shows the data that was available at the time of our June Quarterly Outlook, and the following chart reflects the updated data.

Figure 2 | Unit labor costs were rising into the 2000 and 2008 elections

Source: Bloomberg

Figure 3 | US profits to GDP now a descending line at 9.3%

Source: Bloomberg

Notice the difference: instead of a flat line at around 11% of GDP since 2015, profits are now gradually declining as a proportion of GDP.

What a difference a little data revision can make…

Unpriced inflation risk

We have long positions in shorter-maturity linkers, on the view that the risk of stable to higher core inflation is not properly reflected in market prices. Within the asset class we have a preference for European linkers as their valuation has reached attractive levels.


当資料は情報提供を目的として、Robeco Institutional Asset Management B.V.が作成した英文資料、もしくはその英文資料をロベコ・ジャパン株式会社が翻訳したものです。資料中の個別の金融商品の売買の勧誘や推奨等を目的とするものではありません。記載された情報は十分信頼できるものであると考えておりますが、その正確性、完全性を保証するものではありません。意見や見通しはあくまで作成日における弊社の判断に基づくものであり、今後予告なしに変更されることがあります。運用状況、市場動向、意見等は、過去の一時点あるいは過去の一定期間についてのものであり、過去の実績は将来の運用成果を保証または示唆するものではありません。また、記載された投資方針・戦略等は全ての投資家の皆様に適合するとは限りません。当資料は法律、税務、会計面での助言の提供を意図するものではありません。




商号等: ロベコ・ジャパン株式会社  金融商品取引業者 関東財務局長(金商)第2780号

加入協会: 一般社団法人 日本投資顧問業協会