Seeking Alpha

Jeff Miller

View as an RSS Feed
View Jeff Miller's Comments BY TICKER:
Latest  |  Highest rated
  • Best Recession Forecaster: Robert F. Dieli [View article]
    Young -- I'll have more on the indicator this week, I hope.

    There is a logic behind the organization of the series, which I think will become apparent.

    Thanks for your patience:)

    Jeff
    Jan 15 10:10 AM | 1 Like Like |Link to Comment
  • Weighing The Week Ahead: Have Big Companies Lost Their Earnings Mojo? [View article]
    pigdog67 -- Just to be clear, I said that we closed our IWM position last week with a 6% gain since Dec 19th. Those subscribing to our weekly ETF ratings know that oil (via USO) has been near the top of the list but in the penalty box. IWM still has a positive rating, but other sectors are better.

    Gas prices in the Chicago suburbs have indeed moved higher and my son forgot to fill up the tank before heading back to school!

    Jeff
    Jan 15 09:59 AM | Likes Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    CW --- You are changing the subject, which is recession forecasting.

    I haven't made any comments about investment performance, which can happen in many ways.

    I ask only that everyone evaluate the Dieli model on equal terms with other forecasts. I have invited nominations for many months, with stated criteria. If there is something to show, then please do so. The question is whether any of these people have a long-term real-time business cycle forecasting record that is better?

    Jeff
    Jan 14 06:48 PM | 2 Likes Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    Fabien Hug -- Yes, as one approach. To be explained....

    Thanks,

    Jeff
    Jan 14 01:14 PM | 1 Like Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    CW -- I agree with you about GDI and the potential importance in any NBER decision about a cycle peak.

    As to your economic contributions at SA, you are maintaining focus on important issues while using data. This is consistent with your website manifesto :) http://bit.ly/yPp5Ez

    Jeff
    Jan 14 01:13 PM | Likes Like |Link to Comment
  • Evaluating Recession Forecasts: What Every Investor Must Know [View article]
    Thanks to everyone for some great comments -- stimulating more discussion.

    The next installment will provide even more such opportunity, I promise!

    Jeff
    Jan 13 09:58 PM | 2 Likes Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    Conventional -- There are several researchers working on the ECRI dates. One problem with the trading angle is that that stocks do not always move with the economy, but I obviously I agree that the forecasts add value.

    I'll have more on the ECRI method, maybe next week.

    Good idea -- and not too hard to try.

    Jeff
    Jan 13 08:36 PM | 1 Like Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    bixbubba -- Your determinism idea is not practical, nor is it a good idea. It is typical of how non-economists approach problems -- something that I have written about as "light switch thinking."

    Relationships are never 100% or 0% regardless of past experience. While I can state that we have never seen a recession with readings at the current level, it is certainly possible that there could be a systemic shock great enough to cause one. My personal preference for using this model is to assign a probability to various scores. This is actually preferred by most economists.

    And by the way, if that is your approach, you must really hate the ECRI where there is interpretation from factors they don't even report!

    I know that you would like to have a 100% certainty, but we don't get that in social science.

    As to the double dip in 81, you are free to misinterpret the model if you insist. It calls cycle peaks and troughs. It cannot forecast a new peak unless there is a trough. It is that simple.

    I am glad that this topic has stirred your interest, and I hope you continue to join in the discussion on future installments.

    Jeff
    Jan 13 08:32 PM | Likes Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    Conventional --

    First -- and I want to say this nicely, but still make the point -- quantitative research has been my life for decades. I have developed hundreds of models and tested thousands. I taught these classes at a top graduate school. I have worked on this project for several months. Do you really think that I have not already considered all of the things mentioned in today's comments? You guys do not even know the basis for the model. You will really have fun next week!

    Second, on Hussman, I think my conclusions were very gentle, but I'm not going to repeat what I wrote in the comments there.

    Third, on data revisions -- Notice that the Dieli method was accurate on the 2007 weakness, even early as you noted above. More importantly, you are describing revisions in the GDP, which are common and large. The real question is the chance for revision in the four series followed by the NBER.

    Finally, and most importantly, you are wrong about bias in the process. I stated the criteria and followed them -- all good principles for evaluating research. I guess you might be surprised to learn that nearly all of the best entries show low recession odds. You will hear more about them in a future installment.

    So I hope you keep reading and commenting, but I have a little challenge for you. Why don't you try to do the same type of analysis you did here on the ECRI. Or Hussman. Or Shilling.

    Just write a little piece explaining that their methods are not as noisy as Bob's and/or showing how they did better in real time. I rarely allow guest authors on my blog, but if you can do it, I'll run it.

    More next week:)

    Jeff
    Jan 13 04:51 PM | 2 Likes Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    Screamin187 -- Bob and I have discussed the "misses" in the series. There are two readings which, in a very strict sense, can be interpreted as misses. One was in the high turn of 1960 in which the indicator did not get to the 200 range during the NBER interval. The second was in 1966 when it went negative but there was no NBER recession interval. In the latter case, the economy did slow, but not the degree that the NBER considers necessary to call a recession.

    The late 70's - 80's double dip was not a miss, as I explained in another reply. The indicator never showed an end to the "first" recession. The NBER chose to treat it as two events. Since it was in recession territory all the time, it is incorrect to call this an error.

    I hope this clarifies.

    Jeff
    Jan 13 04:20 PM | 2 Likes Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    Conventional --- As to the double dip in the 80's, the indicator was accurate. If you notice, the line never crosses 200 to the upside, so it did not indicate a cycle trough. The NBER chose to divide this period into two different recessions, while the indicator never left recession territory.

    In the 2005-08 period, there is some noise. The recession probability gets very high at 200 and even higher at 100. In 2005 there was some noise from Katrina sending the indicator briefly below 200. Bob started warning clients in June of 2006 of a 2007 event, and yes, this was a little early.

    One of the things an experienced researcher treats as a warning are results that are "too good." It is easy to slap on a few more variables and exclude the cases that do not seem to work. You are just kidding yourself when you do so.

    Jeff
    Jan 13 04:14 PM | 1 Like Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    screamin187 -- The chart shows an indicator of impending cycle events, not economic strength. As I have tried to emphasize, cycle turns come at economic peaks. The question is whether we are close to an economic peak.

    As to changing the model -- most of the other recession forecasters change and backfit their models. That is why the predictive power is overstated. A strength of the Dieli approach is that he has NOT tinkered with it.

    Jeff
    Jan 13 03:47 PM | 2 Likes Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    Conventional -- I covered the Hussman blindside hypothesis yesterday: http://bit.ly/yU7AYP

    Everyone really needs to read each article in the series, since one builds on another. If you think of the process as forecasting business cycle peaks and troughs, it will be easier to understand.

    Jeff
    Jan 13 01:55 PM | 1 Like Like |Link to Comment
  • Best Recession Forecaster: Robert F. Dieli [View article]
    Conventional -- You can get an idea of the rate of change in the indicators just from the history. When we do reach 200, it typically is a nine-month warning.

    The main point of my first article in the series is that forecasting recessions is a matter of finding peaks in the business cycle. The indicator says that we are not close to a peak. It is contrary to most people's psychology that if things are bad, this must be a recession.

    More next week, and thanks for the good question.

    Jeff
    Jan 13 01:50 PM | 2 Likes Like |Link to Comment
  • Evaluating Recession Forecasts: What Every Investor Must Know [View article]
    Jeremy - You are obviously not a regular reader of my stuff or you would know that my investment posture changes regularly and is reported on a weekly basis.

    You then take this mistaken impression and infer that I would distort research results to support a particular viewpoint. This is completely wrong. I have many different investment programs and I beat the market by being right most of the time. I want to discover what is going on and adjust positions accordingly. I have frequently written about the mistakes that happen when you start your analysis with a conclusion.

    I have spent many months on this project. If you keep reading with an open mind, I hope you will find it useful. You will also see that the methods I have discovered began with a hypothesis, tested it, and then proved it in real time.

    I strongly suggest that you go back and read my explanation about why every recession begins with strong data. The Hussman statement is factual but creates an incorrect impression that we might be blindsided. It is a factually correct statement that seems relevant, but is no more meaningful than saying the sun will rise in the East.

    Thanks for giving me a chance to elaborate, and I hope you continue reading this series.

    Jeff
    Jan 12 04:51 PM | 4 Likes Like |Link to Comment
COMMENTS STATS
1,504 Comments
2,418 Likes