I just checked at stockcharts.com. All BCC there are bullish and will probably signalat gtr1 tonight or tomorrow.http://schrts.co/YJAEgEIr
FWIW, the unusual variant of the "no new high lately / 99 day" signal that I use also reset yesterday, despite its not being a fresh market high.It will now be bullish at the very least until mid March next year.Jim
My timing system said:Potential recession, obey SMA sell signal.Timing: IN stocks (+5.3% vs. SMA) (# 39 consecutive weeks) possibly in recessionThe SMA squeaked close to the 0.0% sell signal not too long ago. But is handily above that now.
My timing system said:Potential recessionI assume this is based on using FRED data for GTT? I noticed the negative Industrial Production Index number, but saw that the Retail Sales number is still positive year-over-year. I'm never sure whether the timing gate should be opened when both numbers are negative, or if just one is negative (as is the case now). Just curious as to what other folks are doing.Thanks for posting this, RayCathy
The 2 FRED indexes are used to DISconfirm a recession. If both are positive YOY, then a recession is disconfirmed. If one or both are negative, then a recession is not disconfirmed.So....if one or both are negative then you should obey any SMA sell signal, should one occur.Note that "not disconfirmed" is not the same as "confirmed". The terminology is tricky, because our minds have a hard time grasping negatives. Which is the reason I spend hours and hours thinking it over, and wrote a computer program (a bash script, actually) to print & email me the outcome. Once you get the program right, it never gets confused or second guesses itself.
Just curious as to what other folks are doing.I follow all three of the FRED indicators (Industrial Production, Retail Sales, and Unemployment), and turn on timing when any of them is negative; which has scarcely happened since Jesse Livermore first published his work in Jan 2016. But it's not something I'd decide by voting. I think you should consider how this form of risk management interacts with other methods that you use. If your normal timing is relatively twitchy, or if your usual holdings are relatively volatile or expensive to trade (e.g. small-cap stocks), you might want to wait until you get two negative indicators. Given that the main goal is to reduce whip-saws, it might make sense to set a higher bar (two signals rather than one) if the underlying portfolio strategy is in fact prone to whip-saws. Or maybe not: if you take on a lot of risk to begin with, maybe you want all the risk management you can get? As so often, a lot of investing comes down to temperament. Baltassar
[...] since Jesse Livermore first published his work in Jan 2016. . . .An aside/question: "Jesse Livermore" is a nom de plume, yes?--Tom
An aside/question: "Jesse Livermore" is a nom de plume, yes?https://jesse-livermore.com/George
An aside/question: "Jesse Livermore" is a nom de plume, yes?https://jesse-livermore.com/If someone is currently using that name, then probably yes.But there was a famous investor named that.http://www.tindakmalaysia.com/showthread.php/6214-Gold-Jim-S...
RayWhen calculating the moving average from the FRED unemployment data, do you include the most recent month in the 12 month average (current month/average of current month plus preceding 11 months) or is the current month compared to the average of the preceding 12 months?Craig
When calculating the moving average from the FRED unemployment data, do you include the most recent month in the 12 month average (current month/average of current month plus preceding 11 months) or is the current month compared to the average of the preceding 12 months?The SMA signal is the moving average of the S&P500. I use the 43 week (~10 month) moving average.The FRED signals are the most recent month compared to the one 12 months ago. YOY = year-over-year.Note that the FRED data gets updated around the 15th of the month, so you are always 2-4 weeks behind. But that's okay, it's Good Enough(tm).A negative SMA of the S&P500 is the signal; the FRED data is the gate which determines if you should ignore or obey the sell signal.You'd have to read the paper: "Growth and Trend: A Simple, Powerful Technique for Timing the Stock Market Posted on January 18, 2016 by email@example.com" to fully understand the logic.Basically, the plain SMA going negative timing signal is too aggressive---it takes you out sooner and much more often than is optimal. In my earlier testing I came up with a couple of ways to tame that, to improve it. One was to wait for the SMA to drop below 4% the current price. The other was to wait for 4 successive weeks of SMA below current price.The FRED gate works better. In my backtests.
When calculating the moving average from the FRED unemployment data...Oh, sorry, didn't see this. I do not use the unemployment data. First off, it is too vulnerable to manipulation (read: lies) for political reasons.Second off, just the IPG and RRSG are good enough. Including employment does not add any value.-----------------From the paper (almost at the end):Real retail sales growth and industrial production growth represent two diverse, independently reliable indicators of the health of the two fundamental segments of the overall economy: consumption and production. The best result comes when they are put together, in combination.......a recessionary indication from either metric turns the strategy’s timing function on. Otherwise, the timing function is off, and the strategy stays longLater on:the individual timing performance of the growth signals is quite weak. ... [This] doesn’t hurt the result for GTT, however, because they are being used as mere overlays for a more price-cognizant trend-followingapproach. GTT respects price and re-enters the market whenever the trend goes positive, no matter what the growth signals happen to be saying.
When calculating the moving average from the FRED unemployment data...Oh, sorry, didn't see this. I do not use the unemployment data. I do track the unemployment data, along with the other two indicators. Here is the original description of the method, which calls for comparing the current unemployment rate to the 12-month SMA:http://www.philosophicaleconomics.com/2016/02/uetrend/When I do the calculation I include the current month as one of the 12 months. I don't think there's anything in the paper that defines the method precisely, nor that there is any logical reason to include or exclude the current month; nor, indeed, that the choice would make a meaningful difference in practice. But I am open to persuasion, obviously.Baltassar
Baltassar and RayThanks for the comments. The reason for the post was that the article did not specify how the SMA was calculated. I know some researchers use current month in their calculations and some don't and I hadn't seen in his previous articles how he does the backtest. But I have been known to miss quite a bit when I read articles.Yes, it shouldn't really make any difference for a signal to use 12 months of data or 13 months.Craig
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |