Dow Falls 730 on Deteriorating Fundamentals, Evidence Rescue Efforts Not Taking Hold (Updated)

Boy, that was short lived. The massive EU and US rescue efforts to pump equity into banks, the TARP, the increase in the Term Auction Facility (from $150 billion to $900 billion), the Fed offering unlimited dollar swaps to foreign central banks, and those monetary authorities themselves engaging in liquidity operations, appears to have come to naught, or at least very little, as far as equity investors are concerned.

We had expected the market to revert to its old lows, because the economic bad news has only just started. Even with these unprecedented efforts, a successful fix to the banking system does not mean lending will resume on its former terms. Indeed, that would be a singularly bad idea, since overly generous lending created dud assets, which is at the core of this mess. But even if things were to normalize, lending standards will be tighter than before, which means deleveraging regardless, as debt is reduced to a sustainable level. But we never thought we’d see a reversal this fast.

However, we had pointed to this chart earlier from Paul Kedrosky (without previously reproducing it in this blog). Most readers and investors have seen the recent equity markets as representing reasonable value, forgetting that Alan Greenspan made his famous “irrational exuberance” remark in 1996, and that one can make a case that thanks to low interest rates (due to Greenspan failing to allow for the impact of cheap imports on prices in his interest rate policies) that equity prices were distorted for more than a decade. Thus what me may be getting is a nasty combo plate: a reversion of valuations to historic norms when fundmentals are taking a dive:

The Wall Street Journal gives an overview of the carnage (even that word is becoming commonplace):

Dire economic data knocked stocks sharply lower Wednesday as investors braced themselves for an ugly recession unlike the relatively brief, shallow downturns the U.S. has sometimes suffered over the last two decades…

“I don’t just think we’re going to test the lows. I think we’re going to violate them and break lower in a big way,” said Kent Engelke, managing director at the brokerage Capitol Securities Management, in Richmond, Va. Referring to the possible fallout in the broader economy from the credit crisis, he added: “We don’t yet know what that is, because this situation is so unprecedented. Every road sign has been obliterated.”

The Dow’s losses accelerated as the closing bell approached, leaving the blue-chip measure down 733.08 points for the day, off 7.9%, at 8577.91, hurt by losses in twenty-nine of its 30 components. The only exception was Coca-Cola, which climbed 1.1% after posting a strong profit report.

Further detail from Bloomberg:

The VIX, as the Chicago Board Options Exchange Volatility Index is known, jumped 26 percent to 69.25 for the biggest gain in three weeks….

Stocks in Europe and Asia fell for the first time in three days, helping push the MSCI World Index, a benchmark for 23 developed countries, to a 7.3 percent decline. Brazilian stock trading was briefly halted after the Bovespa index plunged 10 percent. The index closed down 13 percent after trading resumed.

Exxon Mobil, Chevron and ConocoPhillips, the three biggest U.S. oil companies, helped lead energy companies to the biggest retreat among 10 S&P 500 industries as crude fell below $75 a barrel for the first time in more than a year. The Organization of Petroleum Exporting Countries cut its 2009 demand forecast for a second month.

Some specific triggers for worry. Retail sales fell, which means our consumer driven economy is going into reverse (although the 1.2 percent decline in a month is far lower than analyst Gary Shilling forecast, who has called for a 4-5% fall). From Bloomberg:

The eroding U.S. economy drove retail sales into their longest slu in at least 16 years, even before this month’s market collapse signaled a deepening recession.

Consumer purchases fell 1.2 percent in September, extending the decline to three straight months, the first time that’s happened since comparable records began in 1992, Commerce Department figures showed today. In another sign of weakening demand, prices paid to U.S. producers fell last month on lower fuel costs.

Sales are slowing just as merchants prepare for the holiday selling season, on which they depend for the largest share of their revenue.

The Wall Street Journal’s MarketBeat blog noted that conditions in interbank lending markets have shown only marginal improvement:

Three-month LIBOR rates have started to decline — hitting 4.55% overnight — but the three-month Treasury bill was of late trading at 0.21%, putting the TED spread, a key indicator of market stress, at 3.34 percentage points, not much better than at the beginning of the week. Meanwhile, due to the need for safe credit, the repo markets have become strained — some participants reported not being able to find enough Treasurys in the repo market.

The Fed’s so called Beige Book report was less than cheery,as the Wall Street Journal tells us:

As problems in global financial markets intensified last month, economic activity weakened across all 12 Federal Reserve districts.

The gloomy report, prepared ahead of the Fed’s October policy-setting meeting and known as the “beige book,” shows that regions across the U.S. have taken on a more pessimistic view about the economic outlook. Most of the Fed’s 12 regional banks reported that manufacturing has slowed and consumer spending has decreased.

“Credit conditions were characterized as being tight across the 12 districts, with several reporting reduced credit availability for both financial and nonfinancial institutions,” the beige book said.

And some telling details from the Journal report citied initially:

A report on New York factory activity was grim, and core wholesale prices surged, suggesting corporate earnings could be pressured by still-high expenses and declining demand.

Morgan Stanley was taking a beating before the slide accelerated. From Clusterstock (hat tip reader Dwight) in the early PM:

Uh oh. The plan to rescue troubled banks doesn’t seem to be working. Despite government promises to back bank debt and inject capital, the credit default swaps are flashing “red alert” signs for Morgan Stanley. Stock is trading off more than 16 percent right now.

Print Friendly
Tweet about this on TwitterDigg thisShare on Reddit0Share on StumbleUpon0Share on Facebook0Share on LinkedIn0Share on Google+0Buffer this pageEmail this to someone


  1. CTMM

    “The only exception was Coca-Cola, which climbed 1.1% after posting a strong profit report.”

    Well, thank god the market for sugar water is holding up.

  2. Anonymous

    Anyone who plots an exponential growth trend line on a linear scale is obviously Palin material. The vertical plot has to be log. Since when has anything grown linearly? Growth has a doubling time.

    It’s amazing Bloomberg would insult us with such a misleading graph. Well, actually not. These are money guys and they make most of their “talk their book” points using the hockey stick graphs that invariably emerge from using a linear scale for a growth process.

    The “reversion to mean” using this graph scale would indicate that we might be at 8000 soon. But consider that from 82 to 94 the Dow went from 500 to 4000 – an eightfold increase. For that to continue from 94 to 06 would imply a current Dow of 32000, not 8000. That would be the real trend line.

    Why is it that you display such graphical nonsense on this blog without calling it garbage? I generally respect your take and it’s hard to believe your BS filter is that out of whack.

  3. Anonymous

    Anyone who plots an exponential growth trend line on a linear scale is obviously Palin material. The vertical plot has to be log. Since when has anything grown linearly? Growth has a doubling time

    Purely in terms of chartology this statement makes sense, but I have to ask – why is exponential growth simply taken as an assumption, and anything which suggests a non-exponential growth curve is automatically suspect?

    Isn’t exponential growth what the bacteria expects to experience right up to the moment just before it finds the edge of the petri dish?

  4. Dale

    why is exponential growth simply taken as an assumption?

    Well, because "growth" of the kind seen in economics is an exponential process by definition, and it only makes sense to graph it using logarithmic scales.

    Another flaw I see in the methodology is using the Dow Jones, a price-weighted index, rather than the S&P, a market-cap weighted index. The Dow Jones is useful for its shock or entertainment value only, particularly when used by reporters doing comparisons with the 1929 crash. No serious person would use it to analyze the market.

  5. Pascal

    Can we stop spending thousands of billions of taxpayer dollars (and next-generation real disposable incomes) on trying to fix an unfixable mess (while millions if not billions “fall through the cracks” and end up in private pockets of well-connected Wall Street people!) and concentrate on padding people and businesses against the consequences of the necessary but painful global deleveraging that is taking place?

    We need to accept the very basic truth here: the West has been living beyond its means for 2 decades now (real spending growth is above aggregate productivity growth) and this shows up in aggregate private market credit as a % of GDP which has reached historical proportions now and has been on a tear for 2 decades (see graph). Although part of this upward trend may be attributed to a “structural break” due to globalization and financial innovation, I doubt very strongly this is possible at that point.

    This “living beyond our means” has been made possible by borrowing (in fiscal and environmental terms) from future generations and from Asia. We are now literally throwing free money at the very people who should get as little as possible, thus structurally integrating moral hazard to the long term mess as well, but that is so “old school”, right?

    Of course “punishment” is not a responsible economic policy – I agree and I totally understand the issue of “clogged” credit markets. But pragmatism IS good economic policy. And using thousands of billions of taxpayer dollars in what seems like ad-hoc “tests” and “approximate” policies with very unsure results that benefit some at the expense of many seems quite un-pragmatic and irresponsible to me.

    The reality is that there are 2 types of recessions: one is due to the Central Bank slowing the economy to decrease inflation (Romer and Romer) and keep the economy “close to potential.” The other type of recession is due to massive misallocation of resources, dislocations in interest rates, and poor risk assessment. The recession just starting now is a “type 2”, and standard policy will not work. We (especially the USA and Europe) will spend astronomical amounts of taxpayer dollars and Central Bank “printing of free money” while discovering this and running into a liquidity trap. We will then wake up in a few years with very serious fiscal issues which will limit the intervention ability of the public sector when it will be really required and useful. I rest my case. I have been playing this record too much now – it is broken.

    Pascal Bédard, Montreal

  6. Anonymous

    Why are the Federal Reserve and Treasury Department so concerned with the financial system when it is the real economy that is most important. Using $2 trillion dollars of taxpayer money to bailout the crooks on Wall Street is a sham, they will just keep the money and live high on the hog during the severe recession they created. To add insult to injury, Paulson got the worst terms possible on the first $125 billion dollars given to the banks, my 12 year old niece could have done better. Where is the Congressional oversight?

  7. wintermute

    Even before you debate the merits of linear or exponential projections – you need to consider the underlying data.

    The Dow Jones 30 Index is a genetically selected subset from a pool of competing companies which happen to have the greatest “fitness value” at any time. Strong companies are selected in and weak ones discarded (AIG/Kraft recently). The process is a form of genetic algorithm with the whole Russell 5000 as a gene pool.

    So in a properly functioning capitalist economy the Dow should outperform other gross measures of GDP. It should have some exponential behaviour in a linear background.

    The water is further muddied with the effect of inflation on stock prices and arbitrary selection of timescales such as 1982-94.

  8. Max

    Well, because “growth” of the kind seen in economics is an exponential process by definition,

    No, it isn’t. It’s only “defined” this way in your head only, so we are eagerly awaiting some reasoning as to why the economic growth must be exponential.

    Already a refutation – developing economies grow much faster than the developed ones.

  9. Sion

    I’m no expert but the graph looks right to me.

    Could it be that the effects of growth are exponential not the measure of it?

    The effects of growth certainly are exponential that much is true. To me it is unclear that price as a measure of value should be exponential also.

    Another question is does size measure value? Many men would argue not…lol….some women too.

  10. Pascal

    Wait anonymous 5:40… the financial intermediation HAS intrinsic and actually quite high value-added in terms of economic role. We need to be careful of what we say here. There IS a real problem of financial intermediation simply not working anymore, and this can grind the whole economy to a halt even MORE. So we need to at least try to do something about this, but as a wrote earlier (see 5:38 posting), there is an issue of the wisdom in using those funds, which you rightfully mention.

  11. Pascal

    by the way the graph I talk about in my original post is NOT the linear graph shown here, it is the graph of personal savings rate and also the graph of private sector credit relative to GDP, which is sky-high and hit historical (1920-2007) peaks in the last few years!

  12. Pascal

    max, I think you need to think this over a bit: 1% relative to 100is 1. 1% relative to 1000 is 10… so the LEVEL will go like this: 1, 10, etc… in other words the LEVEL will indeed show an exponential growth, just like the aggregate price level will display on a linear scale. The convergence of standards of living between rich and poor countries has little to do with this!

  13. Anonymous

    If we believe that money is an illusion, just a numeraire with no greater meaning, and if we assume further that real growth and the inflation of money are constant, we will get exponential growth in nominal terms.

  14. Anonymous

    Everyone’s missing the point on an exponential scale – it is merely a way of showing data over time in such a way that the slope of the line coressponds to the growth rate. A straight line on an exponential scale has a constant growth rate. Same data on a linear scale gets ever steeper, making it look like growth is accelerating, when in fact it is not.

  15. bg

    decades long trend lines need to be logarithmic, not linear. I prefer to use forward looking profits (30%-50% lower than historical) against historical PE ranges ( 10 in bear to 20 in bull market. By this measure we still have 20-40% room to fall.

  16. Anonymous

    would someone be so kind as to show the alternative graphs side by side (i.e. someone make a logarithmic one) I don’t know have the data

  17. Don

    Why Didn't The Market Drop Further?
    And in the ironic department:

    "Community banking executives around the country responded with anger yesterday to the Bush administration's strategy of investing $250 billion in financial firms, saying they don't need the money, resent the intrusion and feel it's unfair to rescue companies from their own mistakes….

    And in offices around the country, bankers simmered.

    Peter Fitzgerald, chairman of Chain Bridge Bank in McLean, said he was "much chagrined that we will be punished for behaving prudently by now having to face reckless competitors who all of a sudden are subsidized by the federal government."

    Holy God. And This:
    "The government decided not to impose an explicit requirement that banks use their taxpayer dollars to increase lending. But regulators said they will watch banks closely. They also noted that banks have less reason to hoard money now that they can borrow more easily. Most important, however, they said, banks want to make money."

    Of course not. Why have a plan that makes sense.

    Also yesterday, the Federal Deposit Insurance Corp. said it will create, essentially, two new insurance programs.

    "The basic insurance program still guarantees all bank deposits up to $250,000. A new supplemental program guarantees all deposits above $250,000 in accounts that don't pay interest. The program basically covers accounts used by small businesses.
    Some European governments had already guaranteed deposits, creating a competitive advantage for banks in those countries. "

    Now we know why they did this. Finally:

    "Bair acknowledged that the new guarantees shelter banks from the immediate consequences of misbehavior because depositors and investors have no incentive to remove their money from an institution if they know that the government stands behind it.

    But Bair said the government's first priority was to stabilize the industry.

    "The risks of moral hazard were simply outweighed by the need to act and act dramatically and act quickly," Bair said."

    I don't know whether to laugh or cry.

    So, let's see:

    1) Small banks don't want TARP and resent it.

    2) We're not forcing TARP recipients to loan money out, thereby obviating it as a credit stimulus program.

    3) We need to guarantee deposits because those socialist Europeans did it.

    4) There are no immediate disincentives to reckless behavior.

    5) We given up on any notion of implicit or explicit guarantees.

    Your government's dollars at work.

    TARP is an awful mess.

    Don the libertarian Democrat

  18. Anonymous

    Yes it was a short lived market upper. Credit deflation has been raging for awhile and the effects on assets are starting to get some attention. Was “contained” when it struck the low end of the housing spectrum but suddenly like a raging forest fire out of control it moved into the better part of town and then set off in all directions.
    Besides the asset destruction taking place what will be discussed for years will be the speed at which it moved and how powerless the authorities were to contain the damage. Leverage will never be looked at in the same way again!

  19. Anonymous

    The graph should be log scale. End of story. You don’t have to be a bull or a bear, you just have to have a basic understanding of mathematics. The fellow who prepared the graph is an idiot.

  20. tompain

    Growth is exponential because of the compounding effect. A society has a certain amount of productive capital. That allows it to produce stuff. If some of what it produces is not consumed but invested, then the productive capital base grows. If productivity doesn’t change, and the same amount of production is invested in each period, the result will be a constant growth rate in percentage terms. That’s exponential.

    There’s no rule that says productivity can’t change, or that the proportion of production that is invested for future consumption can’t change, but it would be quite strange for these things to change in just such a way as to create a linear rather than exponential growth chart over time.

    We can measure the size of the capital base and the level of productivity. If you are looking at the period from 1982, then yes the capital base has grown exponentially and it is foolish to draw and extrapolate a straight line between 1982 and now. In the future (which is what the stock market tries to look at), it is certainly possible that the capital base will grow much more slowly than it has since 1982 (either because of higher consumption rate or a decline in capital productivity), but the likelihood that the stock market’s estimate of future growth would lead the Dow to fall right on a linear trend line drawn from 82 to now is pretty small, and would be coincidental rather than causal.

    Also, growth is a function of labor, and labor is a function of population, which usually is assumed and seen to grow exponentially (ex migration and the like).

  21. JP

    Plotting the data in different ways tells you different things.

    Plotting it logarithmically: If the data are linear on a log graph, then the underlying process has a constant multiplicative growth. Compounding interest falls into this category.

    Plotting it linearly: If the data are linear, then the underlying process has constant additive growth. Example: A printing press runs at a constant output, so it’s contribution to M3 is linear.

    Now, note the LINEAR part of the graph over the years 82-94. There was CONSTANT LINEAR growth over that period. Those years (Reagan/Bush) did not result in compounded multiplicative growth.

    You can plot that portion of the curve logarithmically, and all you will see is a curve that bends over logarithmically. It won’t show you the underlying process was ADDITIVE growth.

  22. Anonymous

    The only thing that is obvious is:
    1) The DOW data is linear from 1982-1995
    2) A sudden increase in the slope occurs starting Jan. 1995.

    The same result is seen in the S&P plot. I don't need mathematics to tell me that something changed starting in 1995 that caused a sharp increase in the rate of price growth.

  23. joe

    I want to thank y’all for grabbing by the horns a dilemma that has manifested itself on these pages thanks to a “graph”.

    Exponential growth.
    Exponential economic growth.
    At least implied by the capitalist system.

    Exponential economic growth funded, fed and sustained, if you will, by a vehicle that defies exponential growth.

    You guessed it.
    That same old vile.

    Debt, at its creation, in today’s misfiring fractional-reserve banking system, requires in its deliverance of economic activity, that it be paid back around three times over.

    Think about that.
    Three times in the future, you must pay for WHATEVER economic activity has just happened.
    All of it.
    Three times over.

    And that exponential growth thing.

    It’s not the hair of the dog that bit ya.
    More like the deadly poison in the air that you breathe.
    The definition of unsustainability.

    Economically speaking.

  24. Juan

    The chart depicts “growth” or merely a rise in price of a particular subset of financial assets? Guess I would have to say the latter since 1) we (and most of world’s econ) have been in a tendential slowing since the early 1970s, which might also be called a long wave phase with contractionary tonality while 2) state and private attempts to overcome this have helped generated progressively greater financial activity until, during the 1990s, the system became finance dependent, which is to say the ‘normal’ relation between real and financial was turned upside down.

    Beyond a point, which I believe has been reached, asset prices must fall to, or re-equilibriate with, not only underlying (and devaluing) collateral but a diminished flow of global surplus value. What was turned upside down is righted.

  25. Anonymous

    Well many opinions do make a market.

    I really don’t care about the entrails of how you plot a chart.

    Fact is the economy is going south and the market will continue to deteriorate until all the leveraged funds are busted out.

    Fat Tony

  26. DMD

    Debt grows exponentially, not labor.

    People work linearly for the most part. Technology provides step function-type increases in personal productivity, which resumes as linear growth.

    Based on my personal observation of 35 years in the work force (academic, consulting, residential construction as carpenter and US Marine), and a PhD dissertation in computational mechanics (civil eng, Berkeley), my first pass at a model for productive output would look something like this:

    O = P1 + P2 +- P1*P2.

    Yes, there is a non-linear effect, but there is no way to say whether P1*P1 interacts in a positive way or a negative way.

    The upshot is that labor scales arithmetically.

    This makes sense.

    There are only 24 hours in a day after all.

    Note: I don’t consider creating debt as “productive,” and I do consider the phrase “financial industry” an oxymoron.

    So, no, I am not an economist, nor an investment advisor, but this little epiphany and some perusal of DJIA history got me out last October with nice gains.

  27. DMD

    Also, if you do this plot using a log scale, it comes out to around 10,500, IIRC.

    It would be very interesting to plot economic growth normalized by some index of the debt creation that has fueled the last 15 years. My hunch is that would be arithmetically linear.

  28. john bougearel


    You’d think this bunch had never seen a chart b4! What gives!

    Paul draws a valid line on a chart with his ruler and you call it dishonest, untrustworthy! You’d think the damn thing was a bankster or something.

    Have you all lost it? :-)

  29. Juan

    anon 8;38,

    OK, too short, probably garbled, but

    I come at this from a particular 19th century perspective which sees labor as the sole source of value, that value creation takes place within the processes of production and that within capitalist production the wage paid does not pay workers for the total amount of value they create but only what is required for the (non-static) reproduction of that class — surplus value then is the seemingly paid but actually unpaid portion of labor time which is ‘objectified’ in product and forms the basis for profit. To become profit in fact, a portion or all of this surplus must be realized and that can only take place through the process of exchange, the real market (which helps create an illusion that the market itself creates profit even though it can only circulate and distribute, not create).

    Productivity enhancement and capital deepening, as a displacing of living labor, tend to reduce the mass of surplus value relative to capital, so pressure rate of profit, though this ‘pressure’ can be offset via greater intensity and duration of labor among those remaining employed.
    A further offset can take place through the running down, the non-reproduction, of means of production and nature.
    Credit as a necessarily temporary substitute for surplus also enters in and over the last decades came to play an extreme role.

    Notice that all of these have limits.

    In short, ‘diminished flow of surplus value’ = diminished mass of surplus value relative to total claims against that mass, claims which became ‘valued’ as though there were no limits, as though money from money could expand ad infinitum with no care of the underlying real economy.

  30. ciccocicco

    What is this non sense discussion on log – non log?

    The graph is obviously a stupid one.

    The article makes the point that the market is out of sync with respect to the performance of the market 82-94 on which the trend line is based.

    Yes it is out of sync, but for the opposite reasons. During the period 1982 – 1994 the Dow multiplied by 7, during 1994 – 2008 (even 2 years more) it only multiplies by 3.

    So if the thesis of the article are right Dow should be at 30.000 now.

    I know its difficoult to understand for all you Paulson bashers, so I will give you another example.

    If I extend the graph to 500 years the dow at the trendline will be at 150.000.
    Basically if I have to believe this trendline, the average market gain should be 1500% in 500 years or in other words 0,55% per annum. And this based on trendline taken from one of highest growth periods in stock market history.

    And to think the market is tanking now because everibody is cared of people like you a la Roubini and its 3 trillion black hole.

  31. observer of the insane

    In an effort to calm markets FED Head Hank Paulson announced He and the President would not make any more speeches… (A possible headline in the near future).

  32. ciccocicco

    Guys, I have a secret for you to look smart at the next party. Go to print a Dow graph from 1950 to 2008 with non log transformation.

    Voila, here you have a beautiful trendline that says the Dow should be at 1500, still minus 80% to go.

    Whenever you feel too good about life, have a look at this chart and get a good dose of depression, no need for for a couple of hours.

  33. alan von altendorf

    Seems silly to talk about it now, but 25 years ago in correspondence with Milton Friedman, I graphed the arithmetic increase of all US debt (fed + state + local + corp + household) / total liquidity (L), and Friedman pooh-poohed it, saying that the rate of increase (log scale) was all that mattered. Well, I ain’t the one wearing pooh-pooh any more. Rate of increase has been an article of faith for a generation of bankers and montary policy wonks, expecting infinite growth and slightly higher money supply from ocean to ocean forever. Rubbish. Arithmetic sums matter, too. Pray tell what the mark-to-market value of Social Security and Medicare Trust Funds will be next year or in 2015? What percent of GDP will fed + state + local govt expenditure be in 2015?

    No one looking over the horizon any more?

  34. Anonymous

    Anyone who draws a two-decade old chart plotting returns on a linear scale does not know anything about finance.

    There is no dearth of junk and therefore it can crop up in good blogs, however, it would be useful to remove this kind of junk immediately. Ask the author to redo the chart using a log scale for returns.

  35. ciccocicco

    Dear alan von altendorf,

    why can’t you understand?

    Suppose your salary in 2008 is 1000, you get am increase to 1050, you are happy.

    Now you are in 2108 your salary is 6000, and you still get a raise of 50 to 6050, is this raise equivalent? Are you happy?

    Not applying a log scale for dummies. TM


    please take out this chart.

  36. Anonymous

    Sy Krass said…

    You can ue statistics to prove anything. A log graph is useful when comparing eras (1915-2008) when a linear graph will not even measure the crash of 1929-32; it will appear as almost a straight line. A log graph is equally useless looking at just 2002-presnt, it does not acurately reflect destruction of prices in REAL dolar terms which are linear. Graphs are merely representations of asset values. The reason the graph went parabolic in 1995 was because outside money began pouring in to form a bubble.

    As far as the true value of the chart – We’e on the other side of the mountain- we’ll either form a giant head and shoulder top, or go crashing through to the low of 1987, about 1,700. If this is a typical K-wave, the DOW will lose 70-90% of its value.

  37. Dave Raithel

    I can’t embed pictures here. But two links, one is log, the other not log.


    It’s been a good while since I’ve had to use either calc or stats, but everyone made their points clear enough to follow. I am not surprised, but it is not at all comforting, that even the simplest of data provokes such disagreement on how to display it to show what the data may tell us about the ground reality. Small wonder that discussions which invoke “6 sigma” and “delta hedges” (six standard deviations from mean? rates of change plus or minus?) and then drag us into the fat tails of the bell curve which ain’t really normal gives me no encouragement that the mathematization of moral philosophy is going to solve these problems.

  38. alan von altendorf

    Total US debt 2008 is 300% GDP
    govt spending rising rapidly

    In 1943 debt was only 250% GDP
    max war spending was 52% GDP

    There is an absolute ceiling (putatively 52% GDP) of how many bailouts, beggars, bureaucrats, bombs, bullets, xmas toys and taxes that private sector productive firms can carry.

    Log scale irrelevant.

  39. cent21

    Clever group, here I was going to post a blind comment about a linear graph plotting a 20-fold expontial gain.

    Regarding the TED spread, at how many times in the past 50 years has a 4.55% Libor rate represented a high interest rate when the trailing inflation figures suggest a real return close to zero?

    Could it be that the problem with the TED spread is primarily in the treasury rate being artificially low, due in part to a government-induced panic?

    And speaking of exponentials, has anyone anywhere modified their assumption of long term equity returns around 12% if you wait long enough? And just how badly are pension funds doing if the market ever settles down enough that it can take a closer look at that?

    Yes to the idea of Paulson, Bush, and Bernanke putting some systems on auto-pilot and signing themselves in to an undisclosed location for a few weeks. I do hope some of all the dollars that have been pushed out to the edges of the earth have been put into the hands of people who aren’t speculating against the powers that-be-who-dont-like-the-status-quo with the funds.

  40. IF

    cent21: my personal expectation for long term equity growth is inflation + GDP growth or about 3..4 percent plus 3..4 percent for a total of 6..8 percent. This assumes no share dilution for company fat cats etc. (which in the case of the financial sector seems naive).

  41. DMD


    Data doesn’t care about finance degrees, experience whatever.

    ciciocici: yeah… it’s going to be really interesting to see where economic activity finally proves sustainable, once 10s of millions of people realize that all that overtime they worked for decades just vanished, and retirement is a pipe dream.

    You finance types are always real quick on the trigger about past performance not guaranteeing future results. Then you create models based on grossly misplaced assumptions of, say, past consumer behavior: they always increased spending in the past, so they will always increase spending in the future. Cockamamie. Violates common sense.

    The trend change in the early 80s was likely our transition from a cash culture to a debt culture. Each trend change since 88-89 probably represents the repeal of some archaic, depression era regulation inhibiting the economy. I dunno. You tell me. I’m just a dumb engineer with a dumb phd. I look at data, and try to understand the story the data tells me. I don’t dictate my story to the data.

    von altendorf: I’m with you. The data support labor as scaling linearly, arithmetically. And we have a long number of years to owrk off our current liabilities as a nation, one paycheck at a time, week after week, year after year. Arithmetically.

  42. alex

    pardon by naivete and i’m sorry if i missed the answer somewhere in the posts…but can someone explain to me the value of drawing trend lines in an index that constantly changes? the components of the DJIA are different now than they were in 1982, so does’t that make comparisons of growth rates irrelevant? am i supposed to believe that every company, in every sector, grows at the same rate?

  43. David Habakkuk

    Rather than looking at trend lines — where I cannot see any reason to believe that the past predicts the present — does not it make better sense to look at relationships which have been stable over the long term, but with very large variations in the medium term?

    One such relationship is that between equity values and the cyclically adjusted P/E ratio (CAPE), where the Yale economist Robert Shiller has produced figures going back to 1900. The other is the measure 'q' — the replacement value of assets — originally devised by another Yale economist, James Tobin, where figures going back to 1900 have been computed by the British economist Stephen Wright.

    Charts relating both measures to equity values from 1900 to the present are available on the website of the company run by Andrew Smithers, who has long championed 'q' as a measure of valuation.


    The good news is that recent falls have bought the S&P 500 close to what these measures suggest is long-term value.

    According to the Smithers website: 'As at 9th October, 2008, with the S&P 500 at 909.9, the US stock market was almost exactly at fair value according to CAPE and 7% over-priced according to q (excluding the statistical discontinuities from net worth.)'

    The bad news is that historically periods where the value substantially overshoots these measures have been succeeded by periods where it undershoots. Measured against CAPE, these undershoots can be prolonged.

    One of the morals would seem to be that adaptive expectations play us false. When recent historical performance suggests that equities are both profitable and safe, they are likely to be neither. When recent historical experience suggests that they are likely to be neither profitable nor safe, there is some chance that they may be profitable, and the risks involved are at least less.

    As to the question of how the burdens of all these bail-outs are to be born, with the Fed's balance sheet ballooning out of control I do not see how one can any longer rule out the possibility — in the medium term — of inflation or indeed of default. What the medium term is however I no longer know. Nor do I know what is an appropriate investment strategy, where both deflation and inflation are possible outcomes.

  44. Richard Kline

    Re: the value of the chart posted . . . Look. As Sy Krass puts it concisely above, graphs are just representations; it’s interpretations that matter. Different representations isolate different inputs to the underlying data. Seldom is a single input determinative. It’s up to the observer to put a relevant correlation to the indicated relationship: _that_ is where we have value added, or more often subtracted, from the pattern represented in the graph, if any. Log that data if you want, boys, you will still have a serious deflection point even if the slope of the relationship changes. Explaining the deflection is the key. Anyone assuming the nerverending story of ‘exponential growth’ most likely saw the deflection as a meaningless anomaly, and has just lost the smile on their face as well as most of their hair looking at the returns on their positions. That deflection was HIGHLY meaningful, and exactly why bubbles are _not_ hard to see in real time, even in unadjusted data.

    Log scales are clearly superior for interpretation of long-term time series _IF_ the particular log transformation fits the underlying trend dynamic _AND_ if that trend dynamic is comparatively constant. Neither of those conditions holds for economic data over the long-term, there’s the rub. I’m sorry to tell you folks that soaked up all that log-rot in your business courses whilst getting your union cards that the methods you were taught don’t well-map the data you seek to understand on a daily basis, but there it is.

    Logistic transformations are far more accurate correspondents of long-term economic series, at the least because market saturation and other diminishing returns functions flatten individual and sector growth slopes inherently over mid-term time frames. I’m with DMD on this regarding stepwise growth sequentiation in time series. Separately, productivity surges and declines do not map well to assumed near-constant growth functions. Then too, one has to map in and out inflationary/deflationary pressures in the underlying data collected, a form of data prep usually under-performed. There are underlying arithmetic functions which tend to ‘attract’ otherwise ‘constant’ growth functions away from presumed trends—but gradually with sudden readjustments. Like, ahhh, when mortgage price growth exponentials are attracted back to linear wage growth trends; like, well, now. Just like those exponential deflections in equity price trends are going to get smashed back face first into underlying and much lesser _corrected_ earnings trends, which if not arithmetic are much smaller exponentials not least because they are linked to productivity which in turn has a sticky tie with linear factors in labor. I’m with A. von A. and DMD that labor is an arithmetic. If certain trends are log accurate on their face, they operate in contexts which are not log-determined. This is something that trained economists as a group just do not seem to ‘get,’ which is why I don’t spend too much time with their perspective when considering long trend issues.

    Beyond that, I would add that growth rates are most certainly _NOT_ constant over multi-decade sequences: they are phase sequenced higher and lower relative to long-term trends. If anyone commenting on this issue does not understand that, it’s because they believed everything they were taught including being taught not to think for or to educate themselves. Log transformation of trend data is a powerful and very useful operation, which, however, introduces as many gross distortions into the data representation as it removes. Use with care and in context. The belief that log transformations remove noise and refine trend is the kind of methodology that got us rididulous risk assessments based on trend-deflected numbers, and still yields nonsensical ‘profit projections’ when a look out your window makes plain that profits are well lower and heading down like a plane with only half of the wings it went airborne with still attached. Look: and you will see.

    Anyone who could come up with a formulation approriately weighting growth exponentials, growth logistics, source arithmetics, and phase differentials has just won themselves a Bank of Sweden Prize. Any takers?

    So Juan: There is no _global_ long wave tonality. Having read most all of the literature on the issue, I would say that the evidence is there to prove that, though it is sufficiently fragmentary that no one there has overtly asserted that. Well, having other evidence _I_ assert that rather more categorically, but I’m afraid you’ll have to take my word for it. But consider: If we are in a putative long wave contraction, how do you account for expansionary vectors in E and S Asia? That’s just one problem, but there are many others. I say that even though we are now going to get a synchronized global contraction. This synchronized tank diving is _induced_ though, by financial mismanagement and related production malinvestment, rather than a baseline trend in all industrialized economies. Over the mid-term, three to five years, separate long functions will separate different economic regions as they reorient to their intrinsic phase structures. The worse the crash, the longer the disentanglement from the wreckage. The US does _not_ figure to be a winner on relative growth post-smack up. Unless we get a major structural revision of our structural economy, which I think quite unlikely but which can’t be ruled out.

    And Sy Krass, even if we had a 70-90% deflation of equity prices in real terms, something somewhat more than I expect, that would still be in _real_ terms so we’d see much higher numbers for inflation-impacted price series such as the DOW. It is not in the picture that we will get real declines in output of that order, though, thankfully.

    There is an easy way to test if a trendline suggesting DOW 7000 is ‘a stupid number’: look up the trend in 2010 and see if it plots out. 2012 would be better, but you’ll get a fair read on unadjusted price vectors in three years. We will probably undershoot that number in the interim in my rather inexact view. I suspect that it will be a pretty good number, perhaps a tad on the high side. . . . Of course if we get a major or hyper inflation, we _will_ have to log the data to see the trend.

  45. Richard Kline

    So David Habukkuk, agreed, to most all of that. I think q may be a bit high, and certainly overstates where S & P will be in a severe recessionary environment; think 700 with an undershoot, which clearly you already have. It's in the right ballpark on what a trend number should be, though. Let's just say that quadruple digits on the 500 here are the stuff crashes are made of. And one doesn't need a slide rule to know that, just a pair of eyes and the appropriate neural connectivity between them.

  46. Anonymous

    Folks, I think you need to better learn how markets function.

    Here is a blog I have been following, and it has nailed all major bottoms and tops (including many daily tops and bottoms). The latest are the tops of this week, and the bottom of last Friday.

    Found it on google. Google search URL is:

    Direct links is:

  47. Anonymous

    “Then too, one has to map in and out inflationary/deflationary pressures in the underlying data collected, a form of data prep usually under-performed.” — Ricardo Kline

    PRECISELY! It is nonsense to compare stock prices in 1982 dollars to stock prices in 2008 dollars, which have a third of the purchasing power.

    Even log scales assume that inflation and growth compound at a steady rate, when they don’t.

    In the real world, arithmetic charts (often with the lower part of the y-axis truncated to enhance the effect) historically have been favored by sell-side analysts to pimp stocks, not to mention the braindead WSJ. In other words, they are tainted by association.

    Heuristic: arithmetic chart = tendentious propaganda.

    — Juan Falcone

  48. Anonymous

    oh gawd… how long is this thread going on??

    OK, a couple of people already said the obvious – it’s fine to plot a graph of data however you like, so long as you say why you’re doing it in that particular way, what your assumptions are and how you reach your concusions.

    I think this graph was first mentioned by Yves in a quite specific context – i.e. when would be an appropriate time to get back into the market, or where approx. do we think this bear market is going to beginning to bottom out. To answer that, he used a trend line – nothing new.

    So here are two concepts – exponential growth and trend lines – that are important here. But neither of these is a thing in itself, both are emergent properties of processes. (1) economic growth is often exponential because over years approximately the same proportion of product is invested in expanding production each year, so the economy grows by approximately the same proportion each year, this approximate proportion being an underlying trend. (2) On the stock market, which is several times removed from the real economy, these trend lines show up only in a very confused kind of way. But investors BELIEVE that the market is trending in a particular direction and what’s more, they SUSPECT that lots of other investors also BELIEVE that. Out of these beliefs and confidence/despair in the underlying real economy arise the trend lines that we can sometimes easily recognise.

    Sometimes the stock market grows arithmetically – no wonder, since it is often powered from behind by liquidity (arithmetically added) rather than drawn from in front by economic growth (a multiple).

    Got that off my chest – if everyone else can contribute to this endless thread, so can I!

  49. Dave Raithel

    Good stuff.

    Mr. Kline: I tried both Wiki and Google and can't find a definition of "logistic transformations", which you distinguished from log transformations. Care to clarify?

    The Juan who invoked "surplus value" reminded me that in past two days, Krugmann and Stigler have both had reason to pop up on the tube, and both have been asked their take on why the stock markets have been doing what they have. Their answers were nearly identical: Well, the stock market just does what it does, for whatever…

    That's an economist's way of saying, I guess, that the stock markets are epiphenominae. Like the Juan who invoked "surplus value", and the other posters who referenced both linear (labor productivity) and non-linear (population growth) inputs to economic models (what we want is predictable concomittant variation), the real economy is elsewhere. The stock market is an indication of what some people THINK about the real economy; what it indicates is complicated by the business of persuading people what to think about it. (And so I don't dispute that eventually, this has to have an effect on the real economy.)

    One last point, in possible defense of what I took the Kedrosky chart to suggest: I'm not sure if it was HIM, or another observer, BUT someone suggested that the price run-up of stocks (the S&P charts from 1945 have the same patterns as the DOW charts I linked above, though with one extreme deflection) could be correlated to the death of defined pension plans and the forced march into 401k's. There was a simple demand-pull on prices. But how much is THAT change a function of the real economy growth (which has as input, population) and how much is that political – a policy change accomplished by an interest group? As additional "evidence" for this interpretation, other (again, cannot recall the source) bloggers have pointed to P/E ratios falling more toward historical "norms" … but hey, that just invites another argument over what curve best fits the data allowing predictable concomittant variation ….

  50. Yves Smith

    FWIW, French mathematician Benoit Mandelbrot, who made extensive study of price movements in actively traded markets, found that fractals (as in fractal geometry) described how prices change.

    So we are all on the wrong path when we try to use lines….

  51. Dave Raithel

    Anon of 9:19 –

    Didn’t mean to be redundant; I was apparently writing while you were posting …

  52. Dave Raithel

    Yves Smith:

    Oh, so this was a setup?

    What I know about fractal geometry I got out of that book “Chaos Theory” sometime in the late 80’s. Makes for wonderful graphic art. But this does raise the question: If the ability to predict the future requires such complicated analysis, and since so few of us are naturally gifted for that kind of analysis, could most any “stock advisor” be engaged in anything other than “innocent fraud”?

    The real economy calls, time for me (CST) to go to work ….

  53. DMD

    dave raithel: the problem is that while people rage about whether the market follows arithmetic or linear growth (or fractal), *nobody* can predict the stock market anywhere near as well as the weather can be predicted. It’s mental wankery.

    This inability to predict the market seems as common in PhD economists as truck drivers.

  54. Juan

    ‘So’ Richard: it might benefit if we knew what each other meant by ‘long wave’ and further, in re. S and E Asia, tried to take uneven and combined development into account.

    Spatially particular patterns are within the larger wave, i.e. a long wave of contractionary tone does _not_ mean no recessions, no recoveries; is _not_ simply a price wave as much of Kondratieff’s work relied on, but more correctly long waves in the rate of profit (which does not mean earnings, does not mean opportunity cost but does have to do with accumulation/overaccumulation of production capital, profit/competition driven change within/of its technologies, displacement of living labor relative to mass of capital, greater relative mass of unproductive labor and unproductive consumption, etc., and does _not_ have a national border anymore than capital does). IOW, I tend to agree with the late Ernest Mandel’s critique and approach to long waves as was brought out in his early 1970s book ‘Late Capitalism’ and refined in his Cambridge lectures.

    What we’ve been seeing on the surface since the early 1970s is the downside of the postwar wave; or should it be imagined that phenomena such as long-run slowing global growth and the multiple national/international attempts, including ‘financialization’, to revive a faltering system have been unreal and did not come to constitute a reactionary one-sided and internationally incoherent situation of ‘permanent crisis management’. I believe you are assigning causality to to the monetary and financial [“financial mismanagement and related production malinvestment”] without considering that these have been effects.

    Today’s (and yesterday’s) crises did not simply fall from the sky but grew out of a for profit, capital accumulating, system which attempts to reduce so much as possible its own real basis, productive living labor. Steeped in subjectivist ideologies, its personifications remain ignorant of its own contradictions.

    The _real problem_ with long wave theory/data is not that there have been no such waves but the tendency to see these as guarantee of system permancence.

  55. Juan

    Dave R,

    “Invoking” surplus value is to distinguish between productive and unproductive labor, is to state that – no matter how necessary – processes of circulation and distribution cannot directly create one iota of new value, and that, last instance, finance is dependent. Or that _contradiction_ between surplus producing, financial appropriating and price of claims develops, and that looking at only one side or the other misses oncoming limits, leading to unhappy surprises.
    Super short, price falls to value and value is not simply capitalized anticipations.

  56. Richard Kline

    So Dave Raithel, sidle over to Wiki and look under ‘logistic regression;’ read all the realted linked subjects. That’s a good enough place to start. Let me be clear, though: it would be as inappropriate to model typical exponential functions with logistic curves as vice versa. The isssue is that significant change functions which show up in tire series data, especially with regard to economic ‘growth’ or ‘decline,’ are logistic processes, which are best modeled by logistic curves—but which are seldom so modeled in standard economic analyses. The real issue is to make an assessment of the underlying process and to model it appropriately rather than to the tools of reflexive orthodoxy.

    But re: markets, if not epiphenomenae these are perhaps more, as Anon of 9:19 suggests, in fact tertiary data series of much more basic trends. These teriary series are sufficiently fragmented by multiple but disparate inputs that, as Yves says, they are often best mapped by fractal models, but are by definition then ‘not predictable.’ However, investor/speculator working fantasies of where background trends are going as shown in market data _do_ influence capital allocation and decisions and conditions, so there is a feedback, negative and positive, into real trend inputs from surreal trend aggregates. One could say that the ‘data shadows’ on the wall ‘attract’ behavior toward their limits because they are visible but what they represent is much less so—until we bang into hard real limits or step in the chamber pot. Oh well, it’s all a full employment programs for economics doctoral candidates.

    So Juan, I do not attribute cause of long wave functions to anything in economic behavior: they are not ‘economic’ waves, but more broadly social ones. Economic activity is, loosely, organized by their movements as a driving function, but exactly because economic behaviors are have their own constraints and even more because they are subject to intentional intervention by economic actors economic activity has tangential vectors and is noisy compared to the driving background wave functions. And those functions are not a mono-wave but discrete waves of common form in discrete populations. A ‘post-war wave?’ Well, one could map a downswing in the US for much of the last generation, but an upswing in Europe, for example.

    There is a great deal of background to those assertions, Juan, and I’m going to stop there, but I think you get the drift. I’ve read and have respect for Mandel’s work. He, like others interested in this stuff, did his damdest to think it true and test his models against empirical evidence. But fundamental methodological errors doomed the effort. Long waves are not economic, and hence they are not ’caused’ bay _any_ economic changes of any kind. Since you like Mandel, I recommend reading Jacob van Duijn as well. His workd was similar but even a better effort to get there. His best known text took the mono-wave concept and did an analysis—and he couldn’t find it, like most others to crunch numbers. He did a later paper, however, where he disaggregated economic data by national areas and ran them separately, and got separate waves. It’s been awhile since I read that, and I think I’m slightly misquoting his summation toward my own reading of it, but this was one source I had in mind in my comment above in this thread. I think that paper was in Kleinknecht, Mandel, and Wallerstein, eds., 1992. New Findings in Long-wave Research_, but might have been in Tibor Vasko, ed., 1987. _The Long Wave Debate_. I’m not sure without checking my notes which I don’t have with me.

    Personally, I’m far more interested in things like generational and age-cadre social gestalts which are periodized by social cyclical functions than I am in the economic byproducts of such functions. Economics bores me dreadfully, truth be told. But the fat part of the literature on the phenomenon has been written by economists so I’ve had to pole my punt down their foetid bourne to get where I’m going.

  57. Juan


    Yes, long waves are ‘broadly social’ but the social must be able to REproduce itself, which – necessarily having to do with production – at least some might see as “economic” activity, an activity performed within/informed by (historically) particular sets of social relations that are conditioned by and condition the cultural and political, and as we see with every recession/slump, come into contradiction with their own success. Economic activity is social activity.

    ‘Discrete populations’, yes, but unless we ignore the last centuries, also increasingly combined, which should not be mistaken for ‘convergence’. Too much emphasis on the discrete runs risk of overstating the individual while ignoring that which makes individual individual, the social.

    Given the post-war boom, including the ‘miracles’, followed by structural crisis during the 1970s and further problems created by a became global neoliberalism, the ‘post-war wave’ has not been in hiding and, looking at rate of fixed capital accumulation, unemployment, nonfinancial profit after interest and dividend deduction, a long upswing in Europe fails to appear.

    I do accept that your assertions rest on ‘a great deal of background’ and also, I think, a substantially different perspective than mine.

    Van Duijn’s is a theory of innovation, investment, overinvestment isn’t it, but not clear to me whether he sees innovation as the driving factor?

Comments are closed.