Glass Half Full Exercise

[moore_capital54]

Let us pretend for a moment that we are only interested in positive macro and micro economic data. We all know about all the negative data points being constantly pounded into our brains. I follow no less than ten different writers, all of whom have turned bearish. Let's try and use this thread to post the data points which tangibly demonstrate an improvement over the last 12-24 months. You can use company specific data and/or macro data. I understand the flaws in this exercise, but believe the results will help us seek market truth especially over a period of time. Moreover there is nothing of this sort (an aggregation of only positive data) anywhere on the internet (while there are plenty of negative-only aggregators), and so the contrarian in me feels COBF is the right place for it. Be objective and use only those data points that will stand scrutinization.

[twacowfca]

Good idea.  It's like the 'think' system that worked so well in the musical, The Music Man, for learning how to play a band instrument. :)

 

Seriously, here's my input:

 

All the liquidity in the system ( the sources are too numerous to cite ) will eventually result in large nominal gains. 

 

If your statement that all sources are negative is representative of general opinion, that in itself is a powerful positive indicator.

[Parsad]

I don't think we are that far off from Merkel relenting...it is coming...just exactly how much more pain and angst does she want to endure from her peers?  And once they start down that road, the whole deleveraging process will take 3-4 years...but the primary wall would have finally been breached.  We'll see what happens.  Cheers!

[Uccmal]

Bank downgrades causing a rally in financials:

 

Bank Investors Dismiss Moody’s Cuts As Years Too Late

 

 

http://www.bloomberg.com/news/2012-06-21/credit-suisse-cut-3-levels-as-moody-s-downgrades-biggest-banks.html

 

[watsa_is_a_randian_hero]

-Europe surprises with coordinated effort / Eurobond

-China could surprise to upside and avoid "bust".  I feel like everyone loves the idea of China "busting".  It provides assurance that America is still #1 while conveniently invalidating their planned economy.  And while there is evidence of a decline, their population is still very rural/uneducated.  Technology today makes it very easy and speeds up the process for a larger and larger % to become urbanized and contribute towards China's industrialization.  With so much negative news baked in, what if China surprises to upside?

-US continues its trend of employment growth (albeit slow).  Every additional employee earning $40k and paying taxes rather than consuming forms of welfare (99 weeks unemployment, etc) is a huge benefit.  So, even reducing unemployment by 1% / year (which we are trending) should be huge for the market. 

-Energy commodities remain relatively cheap

-Possibility of more market-friendly president in 4.5 months.

-Possibility of removal of all of ObamaCare (while I think insurers may fall in short-run as this hurts near-term enrollment, I believe it will improve long-term profitability)

-While the market is expensive on Shiller's PE basis, the ratio of current adjusted earnings relative to the min of the last 10 years is at an all-time high (indicating Shiller's historical 'E' is most likely underestimating future 'E')

-Supercheap interest rates for corporations to lower their costs of capital.

-HARP should provide refinancing servicing/origination fees to banks.

 

[racemize]

I'll add the positive news in housing being discussed elsewhere on the board.

[bmichaud]

1. The American consumer has deleveraged phenomenally well as a result of government deficits (unlike the wallowing European consumer) and only has a few years to go to get back to historical norms - let's hope we do not get a "market friendly" president that believes immediately balancing the budget is the best prescription....

 

2. Commodities are plummeting, which should provide a tailwind for those companies that raised prices in response to input cost gains and have the ability to keep those price hikes.

 

3. US intermodal rail data continues to indicate a muddle-through economy.

 

4. The US stock market has held up unbelievably well in the face of European implosion and softening US headline data....

 

5. Merkel can talk tough all she wants, but she let LTRO in through the backdoor.....not for one second should someone believe she will let the Euro project break up and/or the German economy to continue its decline....too much is at stake, particularly the 40% of German exports that find their way to European peers.

 

6. EVERYONE is talking about "Lehman 2.0"

[Uccmal]

RE: lehman 2 talk.  Funny how no one was worrying much about this ahead of Lehman 1.

 

I recall the word contained being used by the Treasury, fed, and financial pundits ahead of the fall 2008 crash.

 

 

 

[watsa_is_a_randian_hero]

1. The American consumer has deleveraged phenomenally well as a result of government deficits (unlike the wallowing European consumer) and only has a few years to go to get back to historical norms - let's hope we do not get a "market friendly" president that believes immediately balancing the budget is the best prescription....

 

2. Commodities are plummeting, which should provide a tailwind for those companies that raised prices in response to input cost gains and have the ability to keep those price hikes.

 

2 things...

 

I don't think either candidate would try to immediately balance the budget.  However, I do believe a gradual shift in policy towards a balanced budget (including entitlements) would improve our nation's health.  Further, my comment was regarding "market friendly" was more about regulations/obamacare and less about budget deficits.  I do believe that the additional regulations put in place, and the uncertainty over implimentation is holding back hiring and business investment, at the margin.

 

Re: commodities.  I had the same comment, I am with you here.  However, I would like to now add the caviet that a freefall in commodity prices would be damaging.  A large portion of the economy is tied to the resource space - refinement, exploration, transportation, capital equipment, etc.  Any fall in prices large enough to cause resource companies to cancel/limit future exploration projects would be bad for the market.  I think we are in the sweet spot right now - prices not to high, not to low. 

[dowfin1]

David Rosenberg has changed sentiment because voters may be acting like adults.

 

Mr. Rosenberg says his newfound optimism was crystallized by the votes last week in Wisconsin and California. He sees those representing a sea change in the way state and local governments are addressing their fiscal problems, and expects that attitude will sweep across statehouses and eventually filter up to the national level.

 

He said that those political events were an eye-opener, but the thoughts have been coalescing for some time. A few weeks ago, he published a note listing 10 “silver-linings” amid the dark clouds.

 

All of this leads him to believe the next 15 years or so won’t be like the past 12 or so. Anybody who tries to project the future based on the immediate past is making a mistake of “gargantuan” proportions, he says. “I think we may look back at the events of last week as a real inflection point,” he says.

 

http://blogs.wsj.com/marketbeat/2012/06/14/a-noted-market-bear-gets-a-little-bullish/

[moore_capital54]

Lets stick to actual data points ala Bmichaud's post, otherwise this will turn into a discussion... Good post Bmichaud.. keep them coming. Randian Hero had some good data points that were mixed with some projections, lets not have any projections only clear and concise data that provides demonstrable improvement.

 

My partner and I went through this same exercise in 2001-2003, only in 2003 did things appear to have bottomed.. while the best buys occurred in 2001.

[onyx1]

American Staffing Index, a measure of demand for temporary employment, is now tracking at pre-crisis levels:

 

http://www.americanstaffing.net/statistics/graph_52_weeks.cfm

 

Architecture Billings Index, a leading indicator of construction spendind, is off it's recent highs, but considerably above crisis levels:

 

http://www.google.com/imgres?q=architecture+billings+index&hl=en&sa=X&qscrl=1&nord=1&rlz=1T4GGLL_enUS398US399&biw=1399&bih=725&tbm=isch&prmd=imvnsu&tbnid=opSYR1KHTDAlZM:&imgrefurl=http://mjperry.blogspot.com/2012/04/architecture-billings-index-pos-for-5th.html&docid=T6lSj9o36zwtLM&imgurl=http://2.bp.blogspot.com/-UiFHGPT9t-U/T47Ntdl-JNI/AAAAAAAARP8/L6aunVYmRfs/s1600/abi.jpg&w=1040&h=860&ei=yPXkT5G_PObV0QHN_uW-CQ&zoom=1&iact=hc&vpx=430&vpy=157&dur=1442&hovh=204&hovw=247&tx=86&ty=120&sig=118437345317959465096&page=1&tbnh=164&tbnw=200&start=0&ndsp=15&ved=1t:429,r:1,s:0,i:77

[Uccmal]

At the Extreme macro level we are roughly at the time we should be entering a secular bull market.

 

In 1999-2000 Buffett was warning of a long period of below average performance.  He was spot on. 

 

Secular bear : 99/2000 - 2011/2013% (12-13 yrs)

 

Secular Bull: 1982-1999 (17 yrs)

 

Bear: 68/69 - 1982 (13 yrs)

 

Bull: 50-68/69

 

Bear: 29-49

 

The defined dates are negotiable; the general concept is sound.

 

[alertmeipp]

Clear data point:

 

I am down like 40% from my previous peak.

[bmichaud]

Valuation-wise we are not even close to where secular bulls begin. 10y median Schiller eps is about $70 - to Watsa's point, the headline schiller eps number is dragged down by 2008/2009, so I use median - so at current levels, the market is 19 times earnings. Secular bulls begin at under 10 times. Even if this secular bear doesn't end for another three years, the market would still have to fall under 850 for this bear to end.

 

That said, other times in history when the market has fallen below ten times were in the 1940s/1950s and 1970/1980s - in the former, we were on the gold standard and there was not nearly the amount of fiscal and monetary stimulus as we have now, and in the latter, interest rates were SKY high. Plus tax rates are currently at record lows.

 

Given the extreme, on-the-precipice-of-depression nature of the environment that took us down to March 2009 lows, I have a hard time seeing the market get down to 850 without an extreme European collapse and/or us balancing the budget here in the US.

[link01]

i find it difficult to view the investment landscape from a glass half full perspective right now.

 

not necessarily because of all the problems with debts, deficits, deleveragings, the worsening plight of the poor, the disenfranchised middle class, the broken promise of upward mobility & rising tides that lift all boats big & small economically,  the tragic degradation of political leadership every where....and too many more to mention.

 

not because of all those problems, but because i believe market valuations today are too high compared to historical trend earnings, just as bmichaud points out above.

 

another poster from another board puts it very well, i think:

 

I think I've asked this question before, but is there a readily

(freely accessible) location I can access this data to construct my own models and graphs?

 

If I may interject---

There are much better valuation metrics that are almost as easy to calculate.

So why bother looking it up?

 

I suspect this is not Mr Buffett's favourite valuation metric, only his

favourite from among those so simple they only use two numbers.

That's assuming that he cares much about such things at all.

 

Some of the things that change a lot over time that are not covered in the market-cap-to-GDP ratio:

- The fraction of US economic activity which is accomplished by US-listed public firms

- The fraction of US-listed corporate profits coming from economic activity within the US

- The fraction of national economic activity ending up sustainably as corporate profits

- The ratio of GDP to cyclically adjusted GDP

Some of these are absolutely enormous factors!

Fifty or a hundred years ago it might have been a useful approximation

that the profits of US listed firms came from US economic activity

and vice versa, but it just isn't even close any more.

Around 46% of aggregate S&P 500 profits come from outside the US--up 16%

in just the last decade--as well of around half of production and sales.

 

Ultimately the value of any given collection of equities comes from real

future net profits of that same specific set of companies and nowhere else.

Since aggregate net profits grow at only a finite rate over time, any

reasonable measure of current cyclically adjusted trend earnings will give

a pretty good guess of what the trajectory of future net earnings will be.

The estimate will be flawed, but it's the best one possible unless

you think you can predict really big macro trends accurately.

The ratio of current price to current on-trend real earnings will work

(does work) a heck of a lot better than market cap to GDP.

The advantage is that you can use the aggregate earnings of some set

of companies (the S&P 500 or whatever) and compare that to the price

of the exact same set of companies no matter where they or their

economic activity are based. Market cap to GDP doesn't manage that.

 

The easiest good-enough very simple model is Mr Shiller's CAPE:

(a) Download the history of earnings and prices from here http://www.econ.yale.edu/~shiller/data.htm (first link second paragraph)

(b) Convert all the prices and earnings into inflation-adjusted values by dividing each by the then-current CPI figure in the file.

© Calculate the 10 year trailing average of real earnings for each month

(d) Look at the ratio of current on-trend real earnings to current price.

Anybody who can use Excel should find this pretty easy.

These are the average compound annual total return 3 and 10 years forward

from each valuation decile, after counting dividends and inflation:

3 years  10 years -0.67%    0.40% most expensive decile based on lowest initial trend earnings yield using real E10  4.36%    2.01%  5.21%    4.20%  4.18%    4.80%  4.53%    5.70%  6.24%    5.13%  9.58%    5.86%  7.59%    7.37% 12.59%    8.57% 12.30%  10.12% least expensive decile based on highest initial trend earnings yield using real E10

[the 9 cutoffs to pick the current decile would be <=4.27% 4.89% 5.47% 5.94% 6.51% 7.27% 8.38% 9.09% >=10.91%]

 

We've been running around the 18th percentile lately (second row in the table).

Expect below-average returns from the broad US market in the next few years.

You can do this for the S&P 500 or all US firms and you reach the same conclusion.

 

It would not be smart to expect above-average broad US market returns going forward

until the trend earnings yield gets above its long run median or average.

In round numbers, a >25% drop in prices, a very long wait for trend

real earnings to catch up to current real prices, or a mix of the two.

 

Jim

[moore_capital54]

Jim the problem with your analysis, and the reason why I wanted to see data points, is that your analysis is backward looking. You have no tools that will allow you to project what earnings growth will be or how an increase in housing value may stimulate above average growth in GDP. Looking backwards will not work right now.. imho. We need to try and objectively look for data points (like the one provided by onyx) that demonstrate things are improving even though we don't feel it just yet.

[SharperDingaan]

Not visible yet, but we know it is occurring:

 

US. 0-6 months. Lot of drama, but the fiscal cliff ultimately gets punted as part of the 'global growth agenda'. Neither party will want to wear austerity cuts. Fed is already positioning to counter the anticipated drama.

• Press Reports: Operation Twist extension, QE3 on standby, etc.

 

Eurozone. 0-6 months. Framework to transfer fiscal & political powers to new global eurozone governance institutions. You want German help ? agree to it.

• Press Reports: Merkel ‘relenting’. Bail-out funds increased to 750B Euros. Louder calls for growth & sudden silence on ‘Eurobonds’.

• Cdn Visitations: PM ‘Runway’ comment, Finance Minister ‘governance comments’, BOC Governor presence in senior committees. 

 

Eurozone. 6-18 months. Coordinated eurozone attack to 'grow their way out', stabilize/reduce unemployment levels, & keep politicians in power (Germany, France, Italy, Spain, Greece)

(1) Sovereign debt restructuring. 40-60% of existing debt re-termed, converted into zero-coupon, & backed by gold, forex, & global SDR reserves. FX rates centrally 'managed' against global SDRs. Martial Plan 2 (MP2).

(2) Extensive eurozone infrastructure rebuilding as fiscal policy, paid for with new printing. Hard assets grow at the inflation rate, bail out funds re-tasked to buying in devaluing sovereign debt & mitigating the jump in the yield curve.

• Press Reports. MP2 concept recently floated in senior places. Sudden foreign interest in SDR’s

 

Eurozone. 3-5 yrs. Bretton Woods FX reset (BW2).

(3) Co-ordinated Basel IV restructuring & trust busting of the financial industry. A few heavily regulated systemically important oligarchs - & a lot of smaller & more easily controlled entities. Zombies allowed to collapse.

(4) New global securities act. Direct nation state investment in the traditional hedge fund areas. Securitization, trade clearing, share lending etc. standardized, commoditized, & state supervised. Bulk central bank long/short trading (for state pension plans) disciplining markets.

 

According to the 'chattering class' this is the 'Great Recession'. Does anyone really think the policy response can be anything else but real change - & extensive modernization of the 80 year global '1930 Investment Act' ?

 

[link01]

here's a little more color on the historical trend earnings model which jim espouses & has provided some interesting commentary on over the years on another board:

 

<<I'm well aware that random functions can give the appearance of

reversion to the mean. In this case I don't assume reversion, I

observe that the appropriate mathematical functions demonstrate

evidence of mean reversion with very high confidence.

There is a big difference between assuming a function will mean

revert and depending on the metrics which demonstrate mean reversion.

 

But I'm also aware a priori that the growth of earnings of US companies is not a random series.

It isn't a constant growth rate either, but it correlates strongly with

aggregate financial output, which can be thought of as population

growth combined with output per capita. Both of those trend pretty

smoothly, so the combined function itself trends smoothly.

Output varies with the business cycle, but neither function is random.

Do we think the US economy will never grow again? It seems unlikely.

Do we think it will grow at 10% a year forever? No, it seems unlikely.

Do we think trend real corporate earnings will grow at 5% in the next decade?

Possibly, but pretty unlikely, since it has never exceeded even 4% in the past.

So, some sort of central slope estimate in the 0-4%/year range seems a good starting guess.

 

Even ignoring all that, my starting point remains valid even if the

series were truly random, provided you merely accept merely that the

BIG factors change slowly after taking out the short-run business cycle.

Once you smooth it enough to take out the business cycle, the remaining

function is very stiff and seems likely to remain so.

In other words, even if earnings trend at a very low rate from here,

or trend at an historically very high rate, they will still only

change on trend at a finite speed, and will still have a central

function around which the business cycle oscillations occur.

Unless you think there will never be another recession, we will see

earnings lower than today's earnings in real terms soon enough.

Unless you think the central trend of an irregular oscillating function

passes through peaks, today's level is pretty certainly above the trend.

What's the trend earnings level in Dec 2011 dollars for the S&P 500?

Build an estimator from the data and let me know what you think*. $55? $60? $70? $80? 85!?

I'm sure you'll come up with a number somewhere around there.

They all lead to the same conclusion, just to varying degrees.

If one merely accepts that the concept of cyclically smoothed trend

earnings represent a meaningful concept, one can't argue with the conclusion

that today's market level represents an historically typical valuation level

if and only if on-trend real earnings were ~$100 right now, which they ain't.

Conversely if you think aggregate earnings are not trending, nor cyclical,

nor even subject to an upper limit on growth rate, then fuggedaboudit.

Just go with Cramer's calls ; )

 

Of course, if it's random, then this following can be ignored.

Here are the historical average 3-year-forward real total returns since

1871 based on the ratio of initial price-to-trend-earnings ratio.

-2.51%/year compounded after dividends and inflation

2.85%

4.89%

4.84%

6.48%

7.07%

7.09%

7.55%

12.91%

14.60%

 

Here are the forward real total returns by decile after 10 years:

-0.21%

1.40%

4.45%

5.14%

5.32%

5.04%

5.24%

6.83%

9.20%

11.30%

 

Right now I estimate we're at about the 10th percentile, but certainly bottom quartile.

Certainly earnings could do anything in the next ten years, but

when you hear hoofbeats it's usually better to expect horses than zebras.

Either (a) we're at an unusually high valuation level compared to history

and history will be a useful guide to the fugure meaning returns in

the next 5-10+ years will be lacklustre, or (b) it's really different this time.

 

Jim

 

* I can post the monthly real earnings series from three different data sources if you like.

Trailing "earnings in time interval" known only after the fact from

Mr Shiller, trailing "last four quarters as known at the time" as

reported by Dow Jones News, and national accounts data from the Fed.

They all give very similar conclusions.

They start in 1871, 1942, and 1947 respectively.

The first two are fairly up to date, the Fed data set ends about 8 months ago>>

[bmichaud]

I dug up this old post I made back in late 2010 comparing current tax rates with those in 1974:

 

http://www.cornerofberkshireandfairfax.ca/forum/general-discussion/poll-percentage-of-cash-in-your-portfolio/msg30239/#msg30239)

 

Bottom line, a 9X PE ratio back in 1974 is equivalent to a 14.78X PE today after adjusting for the difference in tax rates. I just have a tough time seeing today's market getting back to single-digit PE ratios barring an utter collapse in economic activity such as what transpired post-Lehman.

 

Here is the post:

 

Another thing to consider when looking at PE ratios in a historical context is the level of tax rates. Consider the following (when I say pre-tax, I mean pre dividend & capital gains taxes):

 

In 1974 - the top marginal tax rate was 60%, dividends were taxed at the marginal rate, and capital gains were taxed at 36.5%. Based on the PE chart in a prior post, the PE ratio was roughly 9X in 1974 for a pre-dividend/capital gains tax yield of 11.11%. Assuming the market payout ratio was 50%, the after-tax yield drops to 5.75%, and the after-tax PE is 17.39X.

 

Currently - the top marginal tax rate is 35%, and dividends/capital gains are taxed at 15%. Again, assuming a 50% payout ratio, the after-tax yield is 3.88% and the after-tax PE is 25.74X.

 

So - the current market pre-tax PE of 21.88X divided by the 1974 pre-tax PE of 9X is 243%; however, the current after-tax PE of 25.74X divided by the 1974 after-tax PE of 17.39X is 148% - a huge discrepancy due to the difference in tax rates.

 

The 1974-equivalent PE ratio based on today's tax rates can be calculated as follows: 1974 pre-tax earnings yield was 11.11% (100/PE of 9) and the after-tax earnings yield was 5.75% - let's call 5.75/11.11 the after-tax yield margin, which was 52% in 1974; currently the pre-tax earnings yield is 4.57% (100/PE of 21.88) and the after-tax earnings yield is 3.88% for an after-tax yield margin of 85%.

 

If we divide the 1974 after-tax earnings yield of 5.75% by today's after-tax yield margin of 85%, we arrive at a 1974-equivalent pre-tax earnings yield of 6.76%. By dividing 100 by 6.76, we arrive at a 1974-equivalent PE ratio of 14.78X.

 

It would take a 32% drop in the market to reach this "1974-equivalent" PE ratio and 54% drop to reach the "Golden Level" of 10X.

 

In order to evaluate the fair value of the market, one must look at interest rate levels. The BAA yield averaged 9.58% in 1974 for an after-tax yield of 3.83% (based on a 60% marginal tax rate), and the current BAA yield sits at 6% pre-tax or 3.90% after-tax (based on a 35% marginal tax rate).

 

For the sake of argument, let's assume the after-tax BAA yield is an appropriate rate to use to capitalize after-tax market earnings. In 1974 the after-tax earnings yield was 5.75% versus the after-tax BAA yield of 3.83% - assuming the market should yield the after-tax BAA yield, then one could say the 1974 market was undervalued. Today's market yields 3.88% after-tax versus an after-tax BAA yield of 3.90% - hence, one could say today's market is roughly fairly valued.

 

I say all this merely to point out that tax rates make a HUGE difference when comparing PE ratios across time periods and that investors waiting for 10X PE ratios may be waiting for a long while. I hope I am wrong though because a 10X PE market would be a dream.

 

(Excel sheet with calculations attached)

[ERICOPOLY]

If you look at any age category under 55 (census data), home ownership rates are lower at the end of 2011 than at any point in history going back 50 years (as far as the data goes).  So,  if historical trends per age category hold, then home ownership rates will CLIMB from here.

[twacowfca]

If you look at any age category under 55 (census data), home ownership rates are lower at the end of 2011 than at any point in history going back 50 years (as far as the data goes).  So,  if historical trends per age category hold, then home ownership rates will CLIMB from here.

 

 

 

 

Will this trend change soon? 

 

It seems that political policy has the unintended consequence of suppressing potential employment in young age groups, exacerbated by Baby Boomers hanging on to jobs in older age groups.  Subsidized  loans for 'higher' education take young people out of the potential workforce, but do not always increase practical employment skills.  Age descrimination laws make it difficult for young people to get jobs or move up in hierarchies.  Occupational licensure requires often unnecessary extended education.  Etc.

[onyx1]

 

 

"... according to the Census report, one million more U.S. households in the first quarter than a year earlier, marking the biggest increase since 2006."

 

http://online.wsj.com/article/SB10001424052702304868004577376342935387520.html#articleTabs%3Darticle

[ERICOPOLY]

 

Will this trend change soon? 

 

 

Gary Shilling argues that home ownership rates are mean reverting.  I do too, only I look further into the data than he does -- I recognize that the new blended (all age groups) home ownership rate will be higher than in the past because of the shifting age demographic (more top heavy because of the boomers).

[moore_capital54]

Both April and May US Car Sales are tangibly higher than last year:

 

http://www.reuters.com/article/2012/06/01/usa-autosales-idUSL4E8H16YI20120601

 

AUTOMAKER              YTD  YTD 2011  PCT CHNG

1  General Motors    1,066,963  1,046,275        2.0

2  Ford Motor          935,864    878,600        6.5

3  Toyota Motor        868,301    701,851      23.7

4  Chrysler            689,257    519,538      32.7

5  Honda Motor Co      576,174    523,550      10.1

6  Nissan              485,484    433,032      12.1

7  Hyundai            292,856    263,588      11.1

8  Kia                237,381    200,060      18.7

9  Volkswagen          170,555    125,681      35.7

11  Subaru              136,602    112,255      21.7

12  BMW                130,843    116,656      12.2

10  Mazda Motor Co      123,886    103,072      20.2

13  Mercedes/Smart      117,231    95,458      22.8

14  Audi                52,494    45,858      14.5

15  Mitsubishi          27,462    35,816      -23.3

16  Land Rover          17,389    13,254      31.2

17  Porsche              13,448    12,996        3.5

18  Suzuki              10,695    11,124      -3.9

19  Jaguar                5,476      5,021        9.1

                                                     

    TOTAL            5,958,361  5,243,685      13.6

    SOURCE: Autodata Corp and Reuters calculations