iBC Machine Daily Update- 5/15/08by Danny on May 15th, 2008 at 10:50:40 pm |
This is the inaugural iBC Machine Daily Update.
The updates won’t ever be this long, I just wanted to take some time to explain The Machine and how it works to aid in your usage of it.
A little bit about The iBC Machine:
The Machine is a count-based indicator. That means it counts the number of stocks meeting certain conditions, over certain common time frames. The point of counting the number of stocks meeting certain conditions is that is paints an internal view of the market that you miss if you simply watch the price.
The way the machine is generally interpreted is at extremes, or when some measures cross over.
The iBC Machine was modeled off other oscillators, and most specifically, the Worden T2 indicators.
Here is an example of the T2108, which counts the % of all stocks trading above the 40 day MA. I plotted it against the S&P500, and added the upper and lower boundaries, which I just eye-balled. The way a lot of these work is mean-reversion. When you hit the extreme, you bet the other way. This is the core of Woodshedder’s RSI(2) strategy, which I very much enjoy.
This particular indicator is a better “bottom caller” than “top ticker,” but, I can tell you, I’ve fine-combed each top in the oscillator, and it did perfectly coincide with a market pullback. Some were half a percent, others in the double digits. The reason I put this up is for an illustrative example of the type of reasoning that inspired The iBC Machine. This oscillator also illustrates why bottoms are easier to call than tops.
Moving Averages are not concrete enough for me however. The iBC Machine measures price action that is both normal and anomalous.
Each day, people are buying or selling a market home to over ten thousand tradable securities. Each day there are a thousand opportunities and ten times as many interpretations thereof. When keeping track of all the upgrades/downgrades, Uncle Ben’s, and Gasparino sit downs, it is all too easy to have information overload.
By focusing on only the number of stocks meeting extraordinary conditions as defined by price movement, volume, or rate of change, you can quickly paint the enthusiasm and mood of the market in a unique way that gives you a tradable edge. You tune out the mundane, and focus on the core.
The exact parameters are secret, but it measures conditions being met that day, over the course of a month and over one quarter. It measures other conditions on a monthly and quarterly time frame. All of the conditions overlap, so visually, when they all say the same thing, that is the most opportune time to buy [sell].
This is, again, purposefully similar to other commonly used indicators. Ever wonder “Whats with using the 50 and 200 day averages all the time?” These are close to the measurements of the number of trading days in a quarter (65) and year (260). So measuring relevant and anomalous conditions of stocks in the market over these time frames is par for the course.
Then, I compared the spreads of these indicators. I use a regression to the mean assumption, which I proved exists in this indicator, to determine again the proper timing for buys and/or sells.
Then you can compare the ratio of spreads to have an even more acute view of the indices, again over numerous time frames. When the numerous time frames agree is when you have the highest success rate and the biggest directional edge.
Now that you have some background on the machine, let me go over tonight.
Here is a picture plotting the REAL buy and sell strength of each day, along with the ratio between the two.I made the ratio visible, as it is designed to show how much stronger today’s buy interest is over yesterday. This reveals the imbalance between buying and selling, which is what we care about more than how much the S&P moved that day. Highlighted and annotated I’ve shown how there was clearly more buy interest at the March 18th lows, and that the buy interest was extremely high.
For the last few days, buy interest has been tepid. The May 12th rally had more buy interest than today’s 94 point rally did. It’s a little hard to tell, but the line above May 12th is high than today’s line.
Moving onto the monthly time frame (not shown in chart), which is what I use to assess overbought or over sold conditions, it is registering as slightly overbought, so until that condition wears off, I am hesitant to put on new trades.
The longer term measure, the quarter indicator (not shown in chart) shows that stocks are still positive, and that this mini bull rally is still alive, even if a bit extended. The way to use the quarterly information is under the assumption that if a trend in the marketplace has existed for more than one quarter, it is real, not specious, and deserves your attention.
How do you assemble the pieces to come up with some analysis that gives me an edge? Here’s the picture I see:
We are a smidgen above the 200-day moving average, which was serving as resistance. I need three days of close above the 200-day and a successful test for me to consider it as support.
In other words, our current location is something to be wary of. You know that buy interest has been weak lately, that is has gotten weaker as prices have gotten higher, and that the monthly indicator is reading mildly overbought. The bottoms in Jan and March were met with strong buy readings. If this market is going to break above 1420 / 13,000, meaningfully, I’d expect to see a lot of participation, not such a narrow rally as evidenced by the lack of daily buy strength. See how it’s coming all together?
The iBC Machine is showing relatively weak buy interest up here, and though we remain in an intermediate term (quarterly) uptrend, the more opportune time to add to positions or engage in momentum trading is when the Monthly oscillators become less overbought, or we have a high conviction buy day. While is is not uncommon for overbought conditions to persist for days following the first reading, they always resolve in one of two ways–The indices trade lower, or the indices flatline while overbought conditions are worked off.
All these factors mean that the edge is to NOT PUT NEW TRADES on, and be quick to take profits on the recent profitable trades you may have made. Conversely, you could cut the risk in half on trades, or, at the absolute minimum, prepare your stomach and mind for your positions to trade off over the next week.
The beauty of this is that it brings us to the “show me stage” of the rally. If we go up more, then The Machine will be too overbought, which increases the likelihood of some selling. Best case scenario is that we dick around here and test the 200-day. Then we have another huge buy interest day, reflected as a spike on the chart. This should alleviate the overbought conditions, all while showing real buy interest, giving the continuance of the rally more credence.
Remember, the key to the spread chart shown above is understanding the strength of the move, not playing a reversion to the mean strategy. That is how you read the other charts, which I will post in other updates.
Well, there it was! First iBC Machine Report!
Damn, quite a doozy. If you have any questions, recommendations, comments, additional clarifications, whatever, be sure to go fuck yourself. No, c’mon, I’m just kidding, of course I would love to hear ‘em.
Good stuff Boone. I’ll look forward to reading this every evening. As I take a gander at the machine tonight, it seems that things are really really extended. Today looks like the market is the most extended, in terms of the spreads and ratios, since October. In fact, you’ll notice the one spread last peaked near this level on Oct. 31, 2007, which if memory serves, was the top.
May 15th, 2008 at 11:13:16 pmyeah that’s the other concern. Since that is visually easy to understand, and you have access to the machine, you know. But yeah, that main qtr spread is at the upper end of its range since the dawn of the rally, and coupled with the uninterested buyers as of late, color me unimpressed. But still rich of course.
May 15th, 2008 at 11:19:07 pmWow that was quite a report.
Thank you for sharing and I look forward to more of your reports using your secret formula. Sounds like a good way to cut through all the noise.
Though you may want to double check the 50 days in a qtr and 200 in a yr. that’s a little short.
Thanks for all the work Danny!
May 15th, 2008 at 11:21:42 pmI understand that an overbought market can stay overbought for some time, yet the oscillators can un-wind (wood calls them extended, I believe) by price just trading sideways or up & down in a tight range, then proceed to bust up and out. Wood also recalls peak level on Oct. 31 suggesting we are at the top now and are imminently going to head down, bigtime (no doubt these are satanic influences of the Fly). Anyways, Danny, can you give us an opinion of where we stand, I realize we are right at the 200 day EMA and it could go either way, but since the trend has been higher highs and lower lows, I believe we will continue an uptrend, instead of crashing and burning, get the US, Europe and Far East into a recession, unemployment lines, etc. This market does not want to go down and I think if Oil prices drop bit by bit, we will see a nice steady move to the upside for the stock market. Anyhow, what do you think is going on?
May 15th, 2008 at 11:29:00 pmI like it, great report!
Here’s to many more reports.
May 15th, 2008 at 11:31:05 pmLTM- thanks for reading it. I look forward to creating the reports and sharing the machine with those interested. I really believe people will find a lot of utility out of it, whether they trade short or longterm time frames.
Please give me feedback along the way.
One trading program I generated solely from the quarterly signals had the following results, vs being long the DOW the whole year:
% of trading days invested:
74 %
Volatility
10% (versus 14% for the dow)
Return for the Year
10% (versus 6% for the dow)
So you got 400 bp higher return, with less vol. , if you had invested the money in a Money market fund the 26% of trading days you weren’t long, you may have near doubled the average return.
This is with zero leverage, and six total commission generating trades (3 buys, 3 sells)
So, I have my mom trading the slow signal, which as I stated only changed three times in 2007 (a change = a buy or a sell), while I focus on using leverage with the fast signal, which gave me ~7 % return in a month.
Re the days, [edited] you’re right.
Boca - Here, here!
May 15th, 2008 at 11:34:50 pmbuylo, you may not be serious, but just in case you are, I’m not predicting the markets to “head down, bigtime.”
I do think a 5% pullback is in order.
You are correct that overbought conditions and oversold conditions (extended) can be worked off by time and a tight trading range.
I too wonder if the market will rally more if oil drops.
May 15th, 2008 at 11:38:30 pmDanny, seems like 65 is closer to the number of days in a quarter.
May 15th, 2008 at 11:40:45 pmBuylo - I think this regarding the current state of the market.
” The iBC Machine is showing relatively weak buy interest up here, and though we remain in an intermediate term (quarterly) uptrend, the more opportune time to add to positions or engage in momentum trading is when the Monthly oscillators become less overbought, or we have a high conviction buy day. While is is not uncommon for overbought conditions to persist for days following the first reading, they always resolve in one of two ways–The indices trade lower, or the indices flatline while overbought conditions are worked off.
All these factors mean that the edge is to NOT PUT NEW TRADES on, and be quick to take profits on the recent profitable trades you may have made. Conversely, you could cut the risk in half on trades, or, at the absolute minimum, prepare your stomach and mind for your positions to trade off over the next week. ”
The market is still bullish, as evidenced by the longer trend indicator (qtrly) being positive. I think the market is precarious here. But, that doesnt mean I think crash and burn, or that the party is over. I will buy if conditions look good.
Conditions looking good mean not overbought, increased buy interest, and a tightening of the LT spread, which is also extended. I expect it to take a few weeks to a month for all of these conditions to reset in such a way that it looks attractive for me to take on a bunch of risk.
However, you can never tell the future. All I can tell, for certain, is that buying is weak and that we are a touch extended.
Massive buy interest will answer my question, or else I expect a sell-off. Right now I would guess mild, but the point is that you are not buying new stocks on the cusp of what will most likely be some market weakness.
May 15th, 2008 at 11:41:05 pmshed - yeah you are right. being as that it is built into the machine, I should know. I’m tired. D’oh! corrected.
May 15th, 2008 at 11:44:32 pmI will give you feedback and hope that it leads to some good timed decisions! Your going to have to give me your fast signal warnings! Please keep explaining your thought process as you post these so we can continue to understand and interpret them correctly.
As far as trading days
January: 21 days
February: 19 Days
March: 19 Days
April: 23 Days
May: 21 Days
June: 20 Days
July: 22.5 Days (The 0.5 is because July 3rd ,2008 is a half day for the market)
August: 21 Days
September: 21 Days
October: 23 Days
November: 18.5 Days (The 0.5 is because Nov. 28,2008 is a half day for the market)
December: 21 Days (I counted the .5 day on Dec. 24, 2008 and the .5 day on Dec 26, 2008 as 1 day)
This brings the total number of trading days in 2008 to: 250 trading days in 2008.
Not trying to be a pain in the ass, just don’t want any calculations to be fucked up!
May 15th, 2008 at 11:47:04 pmit’s built upon 1 for each day (duh), 20/ month, 65/ qtr, no year. Pretty much, I just put the wrong thing in by mistake. Like, I had other numbers in my head, throwing me off. Why? not sure. I was kind of typing that whole thing and analyzing data for 5 hours straight, so my logic lapsed, and I tried to make an incorrect point.
The point is meaningless to the analysis, btw. But yeah, I’m glad I can count on my readers to catch those things. Ur definitely not a pain in the ass. Obviously, and pf course I will make pains to ensure basic mistakes like that occur as infrequently as possible.
May 15th, 2008 at 11:57:58 pmhttp://www.worden.com/TeleChartHelp/Indicators/Worden_s_Market_Indicators_T2s.htm
May 16th, 2008 at 1:20:16 amhttp://www.worden.com/
May 16th, 2008 at 1:21:12 amThanks for posting that link.
It was actually after I read very document that the development of the iBC began in earnest.
I recommend reading that for a more thorough understanding of count-based indicators.
May 16th, 2008 at 1:34:10 amD — going to Vegas this weekend, you should swing out.
May 16th, 2008 at 1:41:34 amCue Mr. Burns: eeexcellent…
May 16th, 2008 at 2:22:04 amword dpeez
jake - whats the mission of the trip? maybe
May 16th, 2008 at 3:01:10 amDanny,
i use them all the time. they are one important piece of the puzzle.
May 16th, 2008 at 8:46:20 amyeah I use them too. They are quite helpful. I prefer my machine though of course
May 16th, 2008 at 2:02:13 pm