Equity Insight
Have cyclicals run too far?
􀀗 Some cyclical trend-earnings PE relatives look full...
􀀗 ...mostly in Europe
􀀗 However, macro considerations may matter more

Have cyclical stocks run too far? Many investors think they have, and we’ve met a few
(including some long-only money) that have positioned their portfolios for a correction. As we
write, global stock indices are now down around 4-5% from their 13 May high, and cyclicals
(and financials) have begun to underperform defensives (see chart below and inside).
In this week’s Insight we update our estimates of trend-earnings-based sector PEs (PTEs). In
the US, the bloc’s PTE is still below its historical average, and more so than the wider
market’s. In Europe, however, the bloc’s relative PTE is actually above its historical average,
and no lower than in October 2007. On this view, we can see why some investors – and
analysts – may feel that the cyclical bounce has gone far enough.
However, we feel instinctively that if any valuations matter at the moment, they are absolute
ones, not relatives. The market is digging itself out of such a deep hole that the key drivers of
performance in H2 will likely be stabilising macro data and rising risk appetite, in which case
cyclical momentum can resume and carry a little further – probably led by the industrial bloc,
which is where activity fell most sharply and the beta bounce even now may be incomplete.
The rally has been associated with genuine improvements on two fronts – in the banking crisis,
and in the economic data. We think the setback may be modest, so we continue to grit our teeth
and retain our modestly pro-cyclical (and pro-financial) positioning within the market.

Have cyclicals run too far?
Between 9 March, when the S&P troughed, and 8
May, our cyclical ‘supersector’ outperformed its
defensive peer by 20% in the US, by 22% in EU,
and by 20% at the global level (these comparisons,
like the chart, are based on weekly data). Over the
same period, the financial supersectors outperformed
their defensive peers in the US, EU and world
contexts by 57%, 58% and 44% respectively.
As of the close on Friday, the cyclical blocs had
given back 5%, 5% and 4% of this outperformance
in the US, EU and globally respectively – that is,
between a quarter (the US and EU) and a fifth
(global) of the rally. Financials had given back 9% in
the US, 7% in Europe and 6% globally, or between a
sixth (the US) and an eighth (Europe).
The scale of the outperformance in cyclicals (and
financials) needs to be viewed in the context of the
wider rally in markets: with the main indices up so
sharply, there has been room even for the
underperforming defensive bloc to make significant
absolute gains.
We are open-minded about this pattern continuing.
We think many defensive stocks offer attractive
long-term investment opportunities, but their lack of
macro appeal in the context of a stabilising global
economy and rising risk appetite means that other
sectors may do better again in H2.
The tables below update our estimates of ‘trend
earnings’ PEs (or PTEs) – that is, PE ratios
calculated on the basis of today’s stock price but
trend levels of earnings. We also show how the
selected sectors’ positions differ from those seen
closer to the market’s high (in October 2007, when
the crisis was still in its infancy).
Investors familiar with the data will know that there
are any number of ways of skinning this statistical
cat. In this approach, we take inflation-adjusted
Datastream earnings for each sector since 1972,
mechanically calculate an exponential time trend,
and use it to estimate where earnings would be today
if they were on this trend. We use Datastream
earnings, despite their imperfections, simply because
a long run of sector data is easiest to get there.
Not all sector earnings trends are positive. Indeed, in
some cases the notion of ‘trend’ earnings itself is a
misnomer, and the smooth line fitted to the data
masks a discontinuous, noisy reality. We have not
tried to identify these instances carefully here.
In others, even where there may have been some sort
of meaningful positive progression in earnings, that
trend might now change – the financial sector being
an obvious case in point, which is why for the time
being we exclude explicit estimates for this sector
from the table (logically, this also impairs the overall
market’s ‘trend earnings’ too, something to bear in
mind when looking at the aggregates).