i like mean reversion short-term strategies. once i learned about the VTO report it made sense to me & i find that it works! one thing i like about the VTO system is that it uses QQQQ so you're not taking on single stock risk. what i don't like about it is; very few trades per year.

so i was thinking....

what about a sector/ETF strategy, something like this:

for any given sector ETF - if today's price is above its 50 DMA &
if today's price is at or below its 10, 1.5 bottom bollinger band &
if the 5 day RSI is below 50...

BUY!

how does that sound? it seems to me that the behavior of a group is easier to forecast than the behavior of an individual. trading sectors would minimize the "bad news" effect that can plague a single stock for a long time. it also (theoretically) eliminates the risk of bankruptcy. of course the bounces will be smaller but... you can't have it all.

any tho'ts?

-ja

Joe,

I think anytime the price is below the lower BB, the RSI will be below 50. I don’t know that for sure, but I would imagine that would be the case.

I tested ETF’s using VTO rules once and I wasn’t too thrilled with the results. But if you use the rules that we’ve been using lately -- sell at soon as it closes profitably, it may work a lot better.

Also, you will really limit your opportunities if you require it to be an ETF, below the lower BB and above the 50 DMA. That’s not going to happen every day. And since we’re just looking for mean reversion, I’m not sure if that requirement improves your odds or not.

The current individual stock BB system will probably have to be messaged somewhat. There are a few stocks right now that are going to have to make pretty good moves in the next five to ten days if they’re going to be profitable.

Larry

Larry,
I'm going to toss out an idea. In the current form the individual bollinger band system produces a lot of small wins but a couple of big losses can erase a big part of that. I'm finding that most of my really big winners GPS, EYET, SHRP, PLL and PDCO (which I happened to catch in the middle of the day) come from finding stocks that are severly depressed below the lower band Intra-day (most occurred at the open) but snap back by the close. Maybe this is a good way to get some large winners to offset some of the big losers. What do you think?

Jim

P.S. I'm also hoping the momentum strategy will also produce some big winners.

Jim,

I know that intra-day moves bought substantially below the lower BB and sold when they bounce is a very good way to do it. My problem is that as far as posting them on the web site is concerned, I have to assume that most people can’t trade them intra-day. So the practicality of order placement is one of the factors I have to consider.

One way to do it may be to find stocks of good companies that had bad news reported the previous day, get an idea from the pre-open where they may open, put a limit order in that you know would be a certain percentage below the lower BB, and if you get filled put an order in to sell on the close or in the after-market.

Larry

Using my idea here are the sector ETF’s that can be bot on Monday’s open:

IBB, IGV, XLP, & XLV

Etf’s getting close:

XLI, XLY, & IYT

These are sector etf’s only. You could add broad based & country etf’s to get more signals if you wanted.

I agree that the RSI measurement is probably redundant – just use the BB or the RSI – not both.

-ja

Larry,
Almost all of the Intra-day moves have occured on the open. I'll be happy to post them immediately if anyone would like to trade them. The Nasdaq stocks are pretty easy to spot by checking the premarket ticker on CNBC.
For those who have a stockcharts subscription just create a favorites list of the S&P 500 and Nasdaq 100. You can quickly scan your favorites (in summary mode), Intraday, to find stocks heavily in the red. I like to watch the stock till I see the the trend reversing before I buy.

Jim

Since we're exploring ways to improve our odds with the mean reversion system has anyone noticed that almost all trades that close within one day will also close higher again within 2-3 days (about 80-85% of the time on the second day).

Gary

It just occured to me that by waiting till a one day trade closes up and buying right at the close you could greatly improve your chances of winning, avoid all large losses and keep every trade to 2 days or less.

Gary

gary,

are you saying that - once you get a signal, wait a day, then put the trade on & hold for 2 days? is that it?

-ja

Joe,
Sort of. Once you get a signal, Ex. the trade closes profitable, you buy at the close that night. About 90-95% of the time the trade will close higher again on the next day or the day after. I just looked at all of the trades we have placed plus the extra ones that I placed and most of the Nasdaq 100 for the last 6 months and found that the strategy works whether the signal occurs on the first day or any other day. Once the trade becomes profitable as per the original rules you almost always get a second up day. The big plus is no trade lasts more than 2 days so the odds of a big loss are very small. A combination reversal and momentum system

Gary

Gary,
Would it work if you used a stop or are those second day moves not straight up?
h

Howard,
I'm not sure what you mean by the "second day moves straight up". I don't think you would need to use a stop since no trade lasts longer than 2 days and momentum is in your favor. I wasn't able to locate any trade that took a large loss. I'll look at the S&P for the last 6 months tonight and see if I can find any trades that would have taken a large loss.

Gary

When I checked 30 of the S&P stocks for the last 6 months the results weren't as good. About 67% profitable. Losses were kept to a minumum though. I don't believe any loss was more than 5%-6%. Not sure if its worth it to take more losses in order to have smaller losses and only have capital tied up for 1-2 days. My inclination is to trade by the normal rules but let the trade run for one or two days after it's profitable looking for the second up day.

Gary

After going over all of the Nasdaq 100 and quite a few of the S&P I determined that 90% of the time you will benefit by holding for the extra 1-2 days.... but only in May when the market was rallying strongly. During the first 4 months of the year it was a toss up. Since its hard to determine in advance when exactly the market is going to rally I'd say the best bet is to just use the rules as Larry has them posted

Gary

larry,

when scanning stks do you look at the relative WIDTH of the bolliner bands? seems that the wider the band the bigger the bounce.

so if i get a two signals, one with a narrow channel, one with wide, i would opt for the wider one.

is that logical?

thanks.

-ja

Joe,

No, I haven't really looked at it yet, but I agree -- it is logical.

Larry

Larry and Joe,
Wouldn't that be about the same as taking the greater ATR? If we used a ratio of ATR to price we'd get the same metric as the bollinger band width if all charts were to the same scale. Since they aren't I wonder if the ratio would be a better, if more difficult to use, measure.
h

does anyone know how to re-set the "lower bollinger band" default on stockcharts.com's basic stock scan? i want to scan for stocks using the 10,1.9 criteria rather than the 20,2.0 default setting.

thanks.

-ja

Joe,
Here are the settings that I use:

Global Filters
Average Vol > 100,000
Average Price > $10.00

Additional Technical Expressions
first line:
Indicator: Close....1 day ago... >=... Lower Bollinger Band...10,1.9...1 day ago

Next line:
Indicator: Close...0 days ago...<...Lower Bollinger Band...10,1.9...0 days ago

However it doesn't catch every signal. I also have a favorites list of the S&P and Nasdaq 100 that I scan through (10 per page). You can also scan through in summary to quickly spot stocks that are down by a large percentage Intra-Day or last close.

Jim

thanks... that's better than what i have now.

-ja

I just discovered that under Gobal Filters: group option

You can specify S&P 500 or Nasdaq 100.

Maybe by using that option you could catch all the signals.

Jim

jim (or anyone)

i'm looking to do a stockcharts.com scan within a set group i've created. i know how to use the global filters to scan within my chosen group... but i want the scan to pull up ONLY NEW bullish SAR dots. (not ALL the bullish SAR charts).

any ideas?

thanks.

-ja

Joe,
I suspect you'll need to set the "days ago" indicator to "0".

Jim

Larry,
I have another shorting idea that maybe you wouldn't mind taking a look at.

Short at the end of the day anytime a stock gaps up but shows a filled black candle stick. Sell anytime the trade closes profitable. Hold a max of 15 days.

It doesn't appear to be a trade that occurs a lot but it appears to have a very high success ratio and a modest number of the trades produce very large gains. I only looked at the Nasdaq 100 and a few of the S&P but I'm guessing you could apply the rules to any stock.

Jim

Larry,
I've been testing the S&P for the candlestick method this morning and there seems to be a very definite advantage to using the Nasdaq 100 as opposed to the S&P. About 94% profitable for the Nasdaq, 85% for the S&P. Would there be any reason why the Nasdaq would perform so much better or do you think I'm just seeing an anomally that would correct over time?

Jim

regarding the scan... er... i don't get it.

i want to scan for NEW bullish weekly SAR dots in one of my folders that i've created.

can anyone sorta spell it out for a "software-challenged" guy?

thanks.

-ja

Joe,

I think this will give you what your looking for.

Global filters
Group: what ever you want to scan
Predefined: Parabolic SAR Buy

Additional Technical Expressions
First line:
Parabolic SAR... .02,0.2 .... 1 weeks ago > ...Close... 1weeks ago

Next line:
Parabolic SAR ... .02,0.2 .... 0 weeks ago < Close...0 weeks ago

Looks like LIZ qualifies in the S&P today.

Jim

For anyone who's interested (and might like to check my results) I'm getting somewhere around 80-83% profitable trades for the B/B system testing back to Nov. of last year. However if you wait until 2 more down days follow after the bollinger band signals a buy and then buy on the open of the fourth day I'm getting about 89-92% profitable trades. (Must have 3 down days in a row before taking the signal).

Jim