Also going long QCOR and ACT today, as they've triggered buy points for me.

Wouldn't Tuesday's gain be considered a 3rd day follow through? Index was down on 3/27, so day 1 would be 3/28. I think that's why IBD didn't call the FTD.

Day 1 starts on the recent low for me which was 3/27.

PS, I realize this is one day less than IBD or what Scott explains in his blog, but it works for me! As long as you backtest and validate the 'tuning' of your timing model, you're good to go. Just don't over-optimize!

Some rally! With the NASDAQ down 2.25% at the time of this writing, it feels like a false rally to me. I know that IBD called the uptrend in tact with the S&P hitting record highs...but the NASDAQ has been the leading index and it looks quite unhealthy with loads of distribution days compounded with spending 9 out of the last 11 days under the 21-day EMA. Today it breached the 50-day in surging volume. No, I have no confidence in this current 'uptrend'.

Yup! Doesn't make the strategy any worse.  This is just another data point which happens to be on the losing side.

If you noticed yesterday undercut the resumed uptrend day, i considered that the killer to this rally attempt.

That's true too. My model will definitely signal a correction. The 50day line is an automatic sell signal for me so I will liquidate all my TQQQ Monday at the open. I looked more into what counts as "day 1" of the rally attempt and there was barely any difference in results when backtesting the IBD rules or mine.

I will see if giving a sell signal when the uptrend day is undercut makes a big difference. That is usually a death knell, so I should investigate it as a way to further reduce losses.

thanks!

Great timing. Following your advice, I put all my money in the market on April 3rd. Excellent call

Maybe you need to find a strategy you're comfortable with if you can't deal with am occasional shakeout.

dtempest, this is a forum to express our thinking.  some good thoughts, some not-so good thoughts.  I don't know ma.mustillo.  but i do know he has put himself out there.  markets turn.  its part of life and trading.  no one plunges, so your comments are an attempt to put him down are not one of support.  find out a way to make his system better.  maybe come up with something he should consider.  be positive rather than condescending.  i think you'll get more out of this forum that way and maybe support from another reader....perhaps me....when a trade goes against you.

My comments are what are called (in some circles)... sarcasm. If you can't take a little sarcasm, then maybe you shouldn't post. Successful people are open to criticism, and a critique of their methodology, even if delivered in the form of sarcasm. I could have been much more rude.

And let's be honest - it's very funny when someone says "good news, my model has triggered a nascent rally." and then a couple of days later, the market goes into a swan dive. It's like when playing golf with your buddies, and one of them says "watch this awesome drive" and then hooks the ball out of bounds.

I do admire ma.mustillo for putting himself out there, and just like in golf - he must feel very humbled right now.

I have been at this for almost 20 years, 15 of those using CAN SLIM, or a variation thereof, and I have found that nothing consistently works well. What works for me, if I feel that the market is oversold and should go higher, is to buy one stock that is moving higher in a base, or breaking out of a base and see how that stock works out. If it makes some progress and gives me a cushion so that I could get out with a small loss if it should turn against me, then I look for another stock to buy. And (this is sacrilege) for CAN SLIM ers, the best stocks often don't have good ratings, at least in terms of their EPS and SMR ratings, although they usually have good RS numbers and good group strength. What you want to look for are stocks breaking out of long bases, where the base has shaken out many people along the way, by dropping below the 50, 100 and sometimes 200 days moving averages, on strong volume. I look for stocks where the earnings next year are expected to be double or more  what last years reported earnings were. I also look for stocks where they climb quickly within the base, and are able to hold their gains. That means on a weekly chart, the right side of the base is almost all blue. If there is some pink, the stock must have closed down just slightly. All of this takes a lot of work. It takes patience, and looking very closely at the charts, and screening through a lot of stocks. You can't simply screen for good CAN SLIM ratings - that's too easy, and if it's easy, it's probably going to disappoint (paraphrasing WON).

I'm open to any critique of my methodology, and I welcome sarcasm:)

All I do is try and find an edge, backtest it, and make sure to cut my losses short. Nothing I've done has been easy and it took a lot of research and development. This loss doesn't phase me one bit, because it's the sum of returns that matters and not any individual event. The biggest mistake is breaking your rules.

Great discussion. We all share basic principles. We only place a trade where we perceive an edge and we cut our losses if the trade goes against us.  We fully accept the risk of each trade and do not slam a system because one signal did not work.  But we do re-evaluate and learn.  We follow our rules but we also tweak our rules.  There is no perfect system and no system works all the time.  But many systems are good enough to be profitable in the right environment.  The great challenge for all of us is to sense when the time is right for our system and to manage risk so we live to invest another day.

Just been for a cycle to clear the mind.  Here are my thoughts about market timing methods.  I think the outstanding feature of current conditions is the Fed keeps stepping in when a correction starts.  So my thinking is we need rules to allow us to get back in quicker.  We need rules to distinguish between a minor correction and a market that is breaking down.  I came up with my own rule that unless the market corrects more than 8% I will start getting back in when the market rallies without waiting for a follow through day.  This is based just on eye balling the Nasdaq and has not been back tested.  But it makes inherent sense.

I carefully follow MarketSchool in Leaderboard and was intrigued with recent rule changes.  Essentially MarketSchool has adjusted the rules to allow for the same thing, getting back in without waiting for a FTD.  This system is much more systematic and I like the new rules.  Basically the buy switch can stay on even after a correction starts. This is called a "powertrend under pressure". I am not yet sure about the precise rules that will turn

the buy switch off.  And I am not sure if the buy switch would stay on in the situation where a powertrend never did start.

A final comment - I appreciate the observations of dtempest about CANSLIM.  These are the kind of posts that make the blogs valuable.

Absolutely, dtempest is right, growth stocks can become out of favour. Check out Minervini's book. Earnings and sales are important, but first he checks if the stock is trending up, then he looks at upwards revisions in earnings estimates and if the stock continually surprises estimates. Then you rationalize this with it's actual fundamentals to confirm if you should trade it and pronounced acceleration seems more important than magnitude. I like this approach because it worries mostly about who is trending and avoids capitulating stocks.

As for market timing, I've been finding my model gets me out of the market much sooner than IBD's calls, and gets me back in a bit sooner or at the same time as IBD. Thus I tend to avoid the churn while IBD is countint 6+ DDs. As I've said above, the risk to getting in sooner was negligible.

If you're curious here is the trailing performance of the strategy:

Actual end date is: 04-Apr-2014

CompReturn: 15.0849%

Total number of signal changes over 278 days: 25

CAGR: 13.4678 %

Max drawdown: 4.2513 %

End to end return: 30.6165 %

These numbers are based on the NASDAQ composite itself. Trading the TQQQ will not net you triple the gains but usually between double to triple. The drawdown usually increases by about the same multiple as well.

Thanks.  I will check out Minervini's book.

Getting all mopey about this signal not working this weekend, and then QCOR brightens up my Monday!

Does your model still signal a nascent rally? You mentioned getting mopey about your signals not working...just curious what your model puts the market action at?

Oh it's been a correction. What I meant with being mopey is that I deal with my innate irrationality by reviewing my models and revalidating them to make sure they correctly represent real-life trading conditions, tweak its fidelity, etc. It's a good way not to get discouraged and understand that it is just another data point on the roads to riches. What helped definitely was seeing that I should be expecting a max loss per trade of 13% when this loss was only 10%. So this kind of loss is expected and hasn't invalidated my model.

The buyout in QCOR netted me some relief at least... sold on that news.

ma.mustillo... when you see the 6 or 7 or 8 days of distribution, do you sell everything, or sell portions of your portfolio? In my own experience, it's often a mistake to go all to cash. In fact, I'm trying to formulate a way of staying in stocks that don't fall in proportion to the market. I use that formula for buying stocks, and it works extremely well, and am trying to find a way to apply it to selling.

I only sell my index ETF when it reaches the 6 DD limit I set (used to be 10 but I find it too risky and not worth the exposure). For stocks, though, I don't automatically sell, nor do I have rules to sell into strength. I simply sell when the stop is hit. The stop is a 3xATR stop for now. I don't like selling on discretion at all.

In terms of selling stocks, I have found that if a stock closes the week (on a weekly chart) greater than 4% below the 10 week moving average, it's usually a good sell signal. Many, many stocks will close the week slightly below it, but not usually more than 3 to 4% below, and then go on to recover. The low of that week when it closes just below it (10 week ma) can also be set as a stop. It's a sell rule that often keeps me in the big gainers for as long as possible.

I've definitely been guilty of setting stops too tight, probably my biggest hindrance to success.

Good news, my model has triggered a nascent rally. Is this the case again?  (just pulling your leg :))

I have found that rather than waiting for the follow through day that we all wait for, it's more effective to buy a stock that we've been eyeing, when the market moves up from a very over sold condition. You want to look for a stock that has been bucking the trend over the past few weeks and is close to making a new high. If your timing is good, a move up from an over sold condition will give you a cushion in the stock if the rally should falter. If the rally continues and we do get the follow through day - you are already in a stock and it should be making good progress. You can then look for other stocks to buy.

What do you consider oversold? Are you looking at RSI or bollinger bands of the market? I try to keep to a systems trading approach so I'm not too good with eyeing things.

I'm glad my crap rally call caused so much discussion! :)

I just click on the stochastics button for the Nas or S&P 500 and you'll see when the market is oversold, but wait until it begins heading higher again, from below 20. It's not foolproof and the market can quickly break down again, but if your stock choice was good, you will usually have a decent gain and can sell without a loss, or with a very small loss if the market turns tail.

What is your current read on the QQQ?

>> What do you consider oversold? <<

I find Chaiken Money Flow indicator to be a better short term overbought and oversold signal generator. Tends to give more accurate short term (2-10 day trade) play support.

Still not back long on this strategy.

Been spending this profitless time designing backtesting software that includes delisted stocks, portfolio risk management, margin, and hopefully shorting soon. This is why corrections can be good, too!

still not a rally for my model. My threshold is 1.4%, 1.2% is not good for long term CAGR.

NOW it`s a rally. :) Too lazy to blog about it.

ma.mustillo, thanks for the update and all the prior information regrading your system.  

Well this time sure is making up for the terrible last signal! Another summer bull in the works.