Sector Peak Returns – Capital Goods Looking Like a Short

Print Friendly, PDF & Email

SEE LAST PAGE OF THIS REPORT Graham Copley / Anthony Salzillo

FOR IMPORTANT DISCLOSURES 203.901.1629 / 203.901.1627

gcopley@ /

September 21th 2018

Sector Peak Returns – Capital Goods Looking Like a Short

  • In a recent Weekly Findings, we highlighted the peak levels of skepticism we are seeing for many sectors, measuring a disconnect between the level of earnings and valuation and suggesting that either returns should fall, or valuations rise.
    • In this piece we take the analysis further and identify the most likely mechanisms to move high levels of skepticism back to normal. The highlight, in Exhibit 1, shows a very strong sell signal for Capital Goods when returns on capital hit a peak – regardless of the levels of valuation. We appear to be at or close to a peak today.
    • The other sector with the same dynamic today is Conglomerates, and ex-GE we are approaching a similar peak to Capital Goods, although not yet as pronounced.
  • For most other sectors the pattern is similar, but it is far less predictive of performance for Materials than for Industrials.
    • It works for Chemicals, but we are no-where near a return on capital peak at this time and relative outperformance looks far more likely for Chemicals than Capital Goods, notwithstanding the pressure of basic chemical margins today.
    • Metals is counter-correlated – you want to buy metals when returns are at a peak and sell at a low! We are close to a peak, but the trend is still improving.
    • Paper and packaging is the more obvious buy – very low valuation and good returns, but valuation so low that almost any SI correction should result in sector outperformance.
  • At the sector level we would want to be underweight Capital Goods and Conglomerates and overweight Metals and possibly Paper and Packaging. We would be more neutral on Electrical Equipment, E&C and Transports and see more select stock opportunities in Chemicals than in other more neutral sectors
    • More obvious stock underweights are PNR, DOV, IEX, MMM and probably LECO
    • Overweights would include: POL, HUN, WLK, PX, and most Paper and Packaging names

Exhibit 1

Source: Capital IQ and SSR Analysis


We first introduced our Skepticism Index in 2012, in an attempt to match valuation anomalies with earnings anomalies – the intent being to explain how much of a company’s deviation from normal value was explained by a deviation from normal earnings. Combining both our (well tested) normal value methodology and the new SI analysis resulted in a stock picking formula that has proven very effective over the last 6 years (see the very last Exhibit,13), generating better returns that the already effective “normal value” framework within cyclical sectors. When we take this analysis down to the company we can identify some interesting anomalies but generally we add a fundamental overview plus some of our additional methodologies, such as business optimism before making a company recommendation.

Both the normal value and SI methodologies rely on reversion to trend (sometimes this trend is also a mean, but we try not to force data to fit a pattern and let the data dictate whether there is a prominent trend or whether an average is more appropriate. For example, for Capital Goods, the return on capital component of the data in Exhibit 1 is driven by the trend in Exhibit 2, and when we discuss standard deviations of returns from normal we are basing off the trend and not an average. This, together with our discount from normal value methodology are more fully summarized in the deck linked here.

Exhibit 2

Source: Capital IQ and SSR Analysis

Sector Analysis

The Capital Goods example that we used in Exhibit 1 is the most compelling, as it suggests you should always underweight the group as return on capital hits a peak – whether the sector is expensive or not.

  • Expensive is when the green area is below the line and cheap is when it is above (relative to the S&P)
  • Any movement of the green line/area upwards over time is underperformance and any movement downward is out-performance.
  • For Capital Goods it is a pretty safe bet that an ROC peak drives underperformance.

Today, we have not necessarily peaked – Exhibit 2 – and one could argue that with the tax reforms the peak should be higher than prior peaks – which it already is in absolute terms – but not on a normalized basis and we saw statistically more significant deviations in 2006 and 1996, 1988 and 1980/1. There may be further to go, but probably not much.

Other sectors that look like Capital Goods:

  • Conglomerates – ex GE – Exhibit 3
  • E&C – Exhibit 4, but we do not have as much history and as many cycles – plus the sector is a disaster to try to model or for stock selection.
  • Electrical equipment has a less pronounced pattern than Capital Goods – there are some lags – and today we are a long way from a peak in returns – Exhibit 5

Exhibit 3

Source: Capital IQ and SSR Analysis

Exhibit 4

Source: Capital IQ and SSR Analysis

Exhibit 5

Source: Capital IQ and SSR Analysis

The Material sectors are less well behaved, with the Metals group having what appears to a counter indicator to Industrials – i.e. buy when returns peak – Exhibit 6. This suggests you should buy now and is consistent with our favored sector thesis, even if it is not working so far this year.

Paper and Packaging has the highest level of skepticism – with high returns and even higher discounts in valuation – Exhibit 7. The high ROC would normally be a trigger to sell, but the current mismatch with valuation is unusual, and we see overweight opportunity at the sector level and for several stocks. Chemicals is another sell at peak returns sector, but as shown in Exhibit 8 – we are not even close to return highs and the sector is very cheap. Late cycle concerns, genuine commodity profitability concerns in the US, as well as trade (Chemicals is the second largest component of US export revenue), have the sector beaten down before returns really got moving. There are stock opportunities here also.

Exhibit 6

Source: Capital IQ and SSR Analysis

Exhibit 7

Source: Capital IQ and SSR Analysis

Exhibit 8

Source: Capital IQ and SSR Analysis

Some Stock Ideas

At the sector level this analysis would suggest being underweight Capital Goods and Conglomerates – a trade that has worked to some extent this year. The analysis is more neutral on other sectors but suggests buying Metals and possibly Paper and Packaging although profitability would suggest otherwise.

At the stock level we have plotted the current valuation and return on capital components by company for Capital Goods – Exhibit 9, Chemicals – Exhibit 10 and Paper and Packaging – Exhibit 11. The format of these charts will be familiar to any regular reader of our sector monthly reports. The rule of thumb is that you should be more interested in stocks to the upper right and less interested in stocks to the lower left. We have not drawn the chart for Conglomerates because of the limited number of companies but the stand-out is MMM on the short side and we include its SI history in Exhibit 12. In Exhibits 9 through 11 we have highlighted clusters of stocks that look most interesting – on the short side for Capital Goods and on the long side for Chemicals and Paper and Packaging.

Our portfolio work, which we have been using for many years shows better results if you restrict yourself to shorting only stocks in the lower left quadrant – there are none for Capital Goods today – and buy those only in the upper right quadrant. The results of our methodology, which uses only the extremes in each category each month are shown in Exhibit 13.

Exhibit 9

Source: Capital IQ and SSR Analysis

Exhibit 10

Source: Capital IQ and SSR Analysis

Exhibit 11

Source: Capital IQ and SSR Analysis

Exhibit 12

Source: Capital IQ and SSR Analysis

Exhibit 13