Wednesday, February 11, 2009

OECD paints a dreary picture

The Organisation for Economic Co-Operation and Development (OECD) released its December composite leading indicator (CLI) report. The OECD has 30 member nations, and economic activity has fallen, and continues to decline, to its lowest level since 1975.

The chart illustrates the OECD CLI since 1961; the CLI is measured to indicate turning points in economic activity. The CLI level 100 represents the long-term trend in economic activity: a CLI above 100 indicates above-trend economy activity, while a CLI below 100 indicates below-trend economic activity. The OECD as a whole is producing well below trend, 92.9.

The CLI pattern is constructed to lead economic activity, so wait to see the CLI rise to indicate that economic stabilization may be underway. That certainly did not occur in December, as the OECD CLI fell 1.1%, its fifth consecutive decline of 1% or more.

At the country level, Russia is showing the most slack in economic activity, while Brazil is showing the least.

This chart illustrates December country-level CLIs across most member OECD countries and key non-member countries (you can see the member countries here). The world economy is suffering jointly - all countries measured in the CLI are below trend (100) and underutilizing resources (i.e., there are unemployed workers and capital not being used for the time being).

However, the last time that the total OECD CLI fell below 100 coincided with the 2001 U.S. recession (see first chart). In October 2001, total OECD economic activity fell below trend to 97. At that time, country-level data indicated that several countries were producing above or close to trend.

As measured by the CLIs, the 2008/2009 cycle is much, much worse than the last cycle.

This chart illustrates that in October 2001, when total OECD economic activity bottomed at 97, Australia and New Zealand producing very close to trend. It reiterates the severity of the global recession (we can call it a recession if the IMF forecasts 0.5% global growth in 2009).

Rebecca Wilder


  1. If the OECD CLI were true, then any man could take them and either "short" those economies whose economic output should fall according to the OECD and could take call option contracts on those economies whose economic output should rise.

    If the CLI were leading truly, these numbers would have predictive power about any future.

    Yet, the OECD "economists" collect data from already happened events. According to the OECD, their economists use "indices  of  industrial  production (IIP)" and intend the IIP as a surrogate for GDP.

    Of course, such indices come from after-the-fact, already produced output.

    Yet, the sum of output is what we want a "leading indicator" to predict before being able to tally up such output.

    In short, the CLI is a useless measure for prediction about any future since it cannot predict the future measure of any output.

    The CLI reflects typical self-deception from the world of Academia Economics.

    True Prediction, An Example

    True prediction rests on relationships, say for example, the ratio of diesel used in farming to the ratio of cooking gas used in restaurants.

    If restaurants have become busier, faster, than farmers, this means more persons have been eating out more owing to higher disposable personal income.

    By looking at the growth rate of such a ratio, a predictor can tell if such a relationship is increasing at an increasing rate (expanding), increasing at a decreasing rate (growth slowing), decreasing at an increasing rate (collapsing), decreasing at a decreasing rate (collapse slowing).

    I give the farm-to-restaurant example above to illustrate a point and am I not making a claim that such is a true leading indicator, although it could be.

    The OECD CLI is not a leading indicator in spite of the claims by their academia trained "economists".

  2. A seemingly convincing argument. However, just for fun, lets try to poke some holes in your relationship example for predictive purposes. Firstly, do most people in fact change their eating habits, and eat out more if their disposable personal income increases? If Harry loses his job, perhaps he is going to drown his sorrows over greasy french fries. Perhaps we should only measure the higher end restaurants? Further, if farmers are using less diesel, might not they be in trouble too?
    I am not familiar enough with CLI to trust it as a leading indicator, but there seems to be at least a direction of trend being presented by the charts. Let's wait and see. When the economy starts booming, we can discount those downward trends we seem to be getting from various sources. But beware, (dare I predict), the "sucker's rally", or countertrend fake.

  3. Since you're game for fun Charles, for fun, let's drive a supertanker through your empty logic.

    I wrote, "True prediction rests on relationships..."

    And then I added, "say for example ...".

    Most importantly, I end with, "I give the farm-to-restaurant example above to illustrate a point and am I not making a claim that such is a true leading indicator."

    All of the world can see that I never claimed my example is, in fact, a leading indicator.

    However, I taught all the key criteria, 1) relationship, 2) measured ratio 3) prediction of trend

    On the OECD CLI

    Charles says, "...there seems to be at least a direction of trend being presented by the charts".

    The graph shows after-the-fact data collected from events already past.

    The OECD "economists" use indices of industrial production as surrogate for GDP.

    No trend exists because no prediction exists. This is a plot of the past, not any future.

    All anyone can say, in truth, is the depth and the rate of change of the fall in the past.

    We don't know that the next plot point shall be higher thus trending the graph up.

    Why don't we know this? Because we cannot predict what the future shall be using data from the past, the OECD summary of the indices of industrial production.

    Are we deficient because of this? No. Because the self-deluded OECD "economists" and their leading indicators are not leading anything.

    OECD "economists" are much like their Academia counterparts who trained them.

    Their beliefs are false because they misunderstand the logic of statistics (the logic of recording past events).

    They're taking data from the past. They're lagging the future events as such events come into the right now, the present.

    On Foretelling Any Future

    It's impossible, always, everywhere and forever to use statistics to predict any future in an open system.

    Men cannot know all of the outcomes that would sum to one.

    Thus, men cannot calculate a percentage chance of outcome from the whole.

    Only in closed systems of either finite outcome (blackjack) or infinite outcome (roulette) can men predict the payoff.

    Only relationships expressed as ratios might be useful for prediction, for example the rate of change between diesel used in food production to natural gas used in restaurant cooking.

    Yet, even here, we're guessing merely.

  4. ok. So we agree there is an indicated trend, not a predicted trend, albeit not "leading"? Or must all trends be predictive?

  5. If someone or others are going to claim they have a thing called a "leading indicator", something that is ahead of what has happened, it better be useful for prediction.

    Otherwise, what he or they have is a lagging indicator and only can use such a thing to say what has happened.

    He or they can say some words like "This was the trend as it happened in the past.".

    In short, he or they are economic historians, useless for anyone who is trying to decide what to do next with their time, hence their money in any future.

    A sane man must conclude that if such economic historians were trying to pass themselves off as oracles, predictors of any future, that such economic historians either suffer from self-delusion or must be charlatans.

    Many have awakened and discovered that they were caught up in trends -- a house price appreciation bubble, a stock exchange paper asset price appreciation bubble.

    If only they could have predicted the trend, they might be financially solvent, divested of their 401k middle class lottery tickets (common stocks without dividends) and fully invested in gold since 2000.

  6. I was inspired by your comments to try to educate myself on the term. I found this link useful:
    I take your point about predictive short comings, but I am not sure prediction is claimed. Perhaps you are right, and the term "leading" is in fact "misleading". In any case, I found the explanation for "lagging" indicator interesting, in that it can be used to confirm a pattern is underway or in the works. The best predictive model, I suggest, is wave analysis. Are you familiar with Elliott Wave studies? A related site that I found very interesting is Socionomics located on the web at:
    Check out their streaming, (and free), documentary entitled History's Hidden Engine.

  7. Thanks Charles. I'll investigate the site.

    I've read about Prechter and the five waves.