Monday Morning Musings: The Still Untold Industrial Production...

 

Analyst: Tobias M. Levkovich

 

SALOMON SMITH BARNEY                                             Industry Note

 

Institutional Equity Strategy                                                  

Monday Morning Musings: The Still Untold Industrial Production Story          

                                                                              

May 3, 2002               SUMMARY                                              

                          * IP drives earnings pretty consistently            

Tobias M. Levkovich       * IP tends to reflect not only manufacturing activity

                            but also drives services.                           

                          * Inventory collapse cannot continue indefinitely   

                          * Semi demand is also affected by industrial        

                          activity.                                           

                          * Things look primed for a pickup                   

                                                                              

OPINION

 

(*) An immediate family member of Tobias Levkovich holds a long position in the

securities of Intel.

 

We have attempted over the past nine months to educate investors about a simple

fact -- industrial production (or factory sales) is one of the most reliable

indicators of earnings trends over the past 30-plus years (Figure 1).

Fascinatingly, despite all of its flaws, especially the fact that manufacturing

has become a smaller part of GDP, this driver has survived periods of a strong

and weak dollar, high and low interest rates, the 1970s energy sector frenzy,

the LBO craze of the late 1980s and tech-mania of the late 1990s. The rationale

for this relationship is also quite reasonable and has significant and credible

accounting basis; overhead cost absorption variances.  Accordingly, the rush to

clean up inventory last year meant meaningful under-absorption of often fixed

overhead cost even though variable labor costs benefited from productivity

gains.  Note that one of the first rules in managerial accounting is that a

firm should never produce goods or provide services when it cannot cover the

variable cost associated with that activity; consequently, it is overhead cost

that can be one of the greatest determinants of profit swings.  In addition,

during periods of inventory liquidation (i.e., production declines), pricing

trends often falter and exacerbate profit erosion trends.

 

Figure 1:  (Figures can be seen in PDF format)

 

Source:  DRI and Salomon Smith Barney

 

Indeed, another way to look at this issue is to assess profits versus capacity

utilization with the knowledge that capacity utilization is an alternative way

to think of overhead cost (the expense of maintaining existing plant capacity

or key employees) being absorbed by production (Figure 2).

 

Figure 2:  (Figures can be seen in PDF format)

 

Source:  DRI and Salomon Smith Barney

 

Interestingly, while industrial production did decline again year-over-year in

1Q02, the decline was not as severe as experienced in 4Q01, which itself was an

improvement from the sharp year-over-year drop in 3Q01.  While we do not yet

have the national income accounts data for 1Q02, it is illuminating to see that

in 4Q01, corporate margins (Figure 3) already began to bounce off lows seen in

3Q01 -- in other words, as the trends improve sequentially and the year over

year dips turn into gains, margins (and earnings) rebound.  If the drop was

caused by massive inventory liquidation (and not just the tech capital spending

slowdown), then a rebuilding of inventory should cause a substantive profits

snapback.  To a great degree, we strongly doubt that most investors even

believe that 4Q01 margins bounced higher, let alone 1Q02 corporate margins,

despite a 1Q02 survey by the National Association of Business Economics that

suggests they did and the 4Q01 NIPA data that one can see in Figure 3.

 

Figure 3:  (Figures can be seen in PDF format)

 

Source:  DRI and Salomon Smith Barney

 

One of the greatest questions we get on this industrial/manufacturing

production issue is; how can such a small part of the economy generate so much

impact on overall S&P 500 EPS?  While we might raise that same question for all

of the tech sector obsessed investors (since the manufacturing economy is about

4x the size of the tech industry's contribution to GDP and almost 10x its

contribution to industrial production), we think that one has to understand the

"multiplier" effect on services when it relates to production.  When more goods

are produced, they must be shipped (generating incremental transportation

services) and they must be financed in terms of working capital investment

(financial services), not to mention sold at the retail level.  In addition,

producing more goods often means adding workers (and temp staffing companies

are suggesting that this has begun to occur over the past two months), which,

in turn, should translate into more consumer spending.  Thus, the impact of

increased production is much more significant than many might think at first

blush.

 

As can be seen in Figure 4, Financials, which account for more than a quarter

of S&P 500 EPS tend to have a meaningful relationship to industrial activity,

especially the banks, given that commercial & industrial loan activity seems to

lag production slightly.  Hence, those who claim that ending activity remains

weak should recognize that this is pretty much par for the course.  While net

interest margin growth probably sustained earnings for the financial sector

last year (and will most likely not be able to continue to do so going

forward), the bigger driver will be credit costs and some renewed loan growth,

which again has pretty decent consistency over many years.  Thus, this is not

one of those often speculative "it's different this time" arguments.  We

consider it to be very much the same!

 

Figure 4:  (Figures can be seen in PDF format)

 

Source:  DRI and Salomon Smith Barney

 

In the Consumer Discretionary sector, earnings fell about $7 billion last year

(or roughly 15% year-over-year), with the auto industry accounting for most of

the decline and auto production rebounding which should translate into renewed

earnings vigor.  As Figure 5 demonstrates, auto earnings are very much driven

by production and the auto sector is generally a respectable lead indicator for

overall industrial production.

 

Figure 5:  (Figures can be seen in PDF format)

 

Source:  DRI and Salomon Smith Barney

 

Note that the Materials sector also got crushed by production cutbacks and

thereby much lower commodity prices (associated with ample supply facing much

lower demand).  Thus, this sector experienced a near 55% decline in profits,

which we think should begin to turn in 2002 and more impressively in 2003.

 

To be fair, the Telecom Services sector and the Information Technology sectors

earnings also got crushed.  To a certain degree, the Telecom arena is a bit

different in terms of the pricing impact from disruptive new technologies and

the excess of capacity that came on in the past few years that still may take

some time to be absorbed.  Within IT, the environment is affected by industrial

production in three ways: 1) Tech does account for about 7% of industrial

production and that collapsed as the industry attempted to cut excess

inventories also; 2) Many other industrial markets use semiconductors -- for

instance, the auto sector uses more semis in dollar value per vehicle than

steel today; and, 3) When industrial activity slides and earnings dip, capital

spending gets deferred.  Since industrial production dropped for 18 months

straight, it should not shock anyone that 2001 was a tough capital investment

year for IT and other capex-related products.

 

As can be seen in Figure 5, IT sector earnings collapsed $45 billion in 2001,

with the key culprits being telecom equipment and semiconductors (Figure 6).

While the telecom equipment market is likely to stay soft possibly through

2003, semiconductors are projected to experience strong EPS recovery in 2002

and 2003.   Keep in mind that industrial/automotive and consumer electronics

markets account for more than 30% of semiconductor industry shipments vs. 46%

for computer hardware; thus a pickup in consumer and industrial markets have

much more impact on semiconductor recovery than simply capital spending might

have for areas like telecom equipment, PC upgrade cycles or CRM software.

 

Figure 5:  (Figures can be seen in PDF format)

 

Source:  I/B/E/S and Salomon Smith Barney

 

Figure 6:  (Figures can be seen in PDF format)

 

Source:  I/B/E/S and Salomon Smith Barney

 

Nonetheless, in what we consider to be somewhat outrageous claims in some

quarters, the Chicago PMI data and the ISM data for April were interpreted by

some as proof of an imminent economic slowdown (which imperils the earnings

outlook), despite the fact that the new orders data registered satisfactory

expansion and inventories continued to contract, setting the stage for needed

production (and thereby, earnings) increases.  In very simple terms, the U.S.

economy cannot end up like Old Mrs. Hubbard with nothing left in the cupboard.

Thus, as industrial production climbs, so will profits, thereby generating

(internally) the cash flow for capital spending recovery in late 2002 and into

2003.

 

We do not consider this approach to be much different than what we have seen

happen before.  The big difference was that consumers never faltered, but the

consumer had to deplete excess inventory along with the production cuts.  As a

result, we are now shifting into the production pick-up portion of recovery and

by late 2002, we probably will enter the cyclical (not secular) capital

investment pickup period.

 

Thus, as we go though production pick-up, auto companies, component

manufacturers, industrial products makers, steel and aluminum producers,

specialty chemicals, semiconductors and financial companies benefit, as do

their earnings.  Back in 1998 and 1999, investors were told to "follow the

cash," in other words, to look at where the cash-rich dot-com and next

generation telecom startups would spend their newfound money.  Thus, demand for

computers, telecom equipment, fiber, etc. soared as did stock prices.  Now, the

initial cash (that, in some cases, must be raised through financial

intermediaries) will be spent to replenish vastly depleted inventories, so the

manufacturers benefit and eventually so will the capital spending

beneficiaries.  Thus, despite their current lack of popularity, we still want

to own names like Ford and GM as well as Borg Warner and Magna International.

Plus, American Express, Bank of America, JP Morgan, Bank One and Lehman

Brothers make sense to us, as do Ingersoll Rand, Deere, SPX, Canadian National,

Nucor, Alcoa and Air Products.  Furthermore, select semiconductor and retail

names continue to hold appeal including Intel, Analog Devices, and Target, not

to mention special situations like Apple Computer. Travel sensitive names like

Continental and Hilton still are attractive to us as well.

 

 

To be fair, investors chose to look for the bad on Friday by focusing on the 6%

unemployment rate (due to a surge in new job seekers, not fewer actual jobs)

and the fact that the non-manufacturing ISM numbers did not show a jump (but

still showed expansion).  The overtime data for the 132 million people who have

jobs (and its associated higher pay) means that consumption can be sustained

and that companies are still cautious on hiring (not a big surprise at economic

inflection points).  One should not and cannot expect that every data point

will be positive every day and, in fact, it seems like more attention is being

paid to the slant or bias on those data points anyway rather than the actual

numbers themselves.  If one can look out beyond the next day or two, the

industrial production factor seems to have a much bigger and more lasting

impact.

 

Indeed, one of the lead indicators for production, freight tonnage

(transportation services), has broken into positive territory and seems as if

it is trying to edge higher (Figure 7), which we think could be positive and

indicative of a coming production pop.  In fact, the only time we saw this not

be a decent indicator for production coming off the bottom was back in the

early 1980s due to inflation/monetary policy issues that simply do not seem to

be recurring.

 

Figure 7:  (Figures can be seen in PDF format)

 

Source:  DRI and Salomon Smith Barney

 

Companies Mentioned

 

(*) An immediate family member of Tobias Levkovich holds a long position in the

securities of Intel.

 

Air Products & Chemicals Inc. (APD-$50.69; 1M)

 

Alcoa Inc.# (AA-$34.43; 1M)

 

American Express Company (AXP-$42.22; 1M)

 

Analog Devices, Inc. (ADI-$36.16; 1H)

 

Apple Computer (AAPL-$23.69; 1H)

 

Bank One# (ONE-$41.50; 1M)

 

Bank of America# (BAC-$73.55; 1M)

 

BorgWarner (BWA-$63.70; 1M)

 

Canadian National Railway Company (CNI-$49.80; 1M)

 

Continental Airlines Inc.# (CAL-$25.00; 1H)

 

Deere & Company# (DE-$44.20; 1H)

 

Ford# (F-$16.10; 1M)

 

General Motors# (GM-$66.24; 2M)

 

Hilton Hotels Corp.# (HLT-$16.09; 1H)

 

Ingersoll-Rand Co.# (IR-$49.96; 1H)

 

Intel Corporation (INTC-$27.87; 1M) (*)

 

JP Morgan Chase# (JPM-$36.25; 1M)

 

Lehman Brothers# (LEH-$62.12; 1H)

 

Magna# (MGA-$77.38; 1M)

 

Nucor Corporation (NUE-$59.69; 1M)

 

SPX Corporation (SPW-$132.00; 1M)

 

Target Corporation# (TGT-$44.12; 1M)