Percent above the days low

dhall6938

At first, I thought the first version would keep me just above the day's low (.006) and just above yesterday's low, again (.006) but now I think that I may have had it written incorrectly. [[Close < 1 Day ago High - [0.006*[1 Day ago High - 1 Day ago Low]]] or [Close < 1 Day ago High - [1 Day ago Range *.006]] and [Close < Low - [Low * .006]] or [[Close < 1 Day ago High + [0.006*[1 Day ago High - 1 Day ago Low]]] or [Close < 1 Day ago High + [1 Day ago Range *.006]] and [Close < Low + [Low * .006]]

markd

maybe you could re-phrase what you want in words instead of code. you say % above days low, but then have code for highs. Help! Also, I think it would be clearer if you just use "range" instead of high-low. And definitely, for clarity, don't mix the two when you mean the same thing. And, if you test just one condition at a time until its right, then combine all the tested conditions, it would be much easier to find the errors (if there are any).

dhall6938

Well, for now I am trying to use a percentage above the low of the day so I don't think using range would work in this case. For now, I want the signal only if current price is less than 6/10 of one percent above the low of the day.

dhall6938

I am testing or trying to test each condition separately, one at a time but it's hard to do this one until the markets are open and running. I just started today and I didn't get a chance to check this set of conditions. So, now I have been wondering if what I have written is composed incorrectly. I am sure that I'll figure it out over time and experimentation.

lmkwin

Your 1st sentence talks about Lows or just above the lows. You don't need High in the code.

Low * 1.006 would be what you are saying, I believe. If the Low is $10 you want Close to be less than something just above 10? Like 10 * 1.006? Close is less than Low * 1.006? This would give you results where the close is between between 10.00 and 10.06

If you multiply by 0.006 then you are saying that Close < Low($10) * 0.006 or Close < 6 cents.

dhall6938

Yes, I think your right again. I'll play with that tomorrow if price gives me the chance. As always, Thank you

dhall6938

So far this is working pretty well. What you guys have explained to me, is something that I would have never figured out on my own. However, I would also like to add another limit to my scans. I don't think that range will work for this one. I want to limit my long trades to no more than a certain percent below the day's high and the opposite for short trades. Half of the days range may not be the same as a percentage from the high or the low. I don't know that the percent for longs will be the same for shorts but for the sake of this discussion we can assume that they will be the same. I believe that I don't want to accept any more than 4/10 of one percent below the days high and the opposite for shorts. For shorts, not more than 4/10 of one percent above the day's low. .004 I think would be the decimal number for both of those situations. [Close > High - [0.004*[ High - Low]] for the longs? [Close < Low + [0.004*[ High - Low]] for the short trades?

lmkwin

if you don't want RANGE then forget about High-Low.

4/10 of 1% :

below high is high * 0.996
above low is low *1.004

dhall6938

Thanks. I do want range, but I think that just because price is above or below its mid-point, it may be up or down more than I'd like for this kind of trading. So, I may also want to make sure that it's above or below its mid-point but not less or greater than a certain percentage from its high or its low or even the previous day's high or close or it's range. Not sure yet but I'd like to keep range and add to it as another limiting filter. Thanks again.

dhall6938

Don't know if you need to know or even want to know but I am trading only SPX and occasionally NDX using ETFs. So, a small percentage that may seem ridiculous might be important. Above or below the mid-point but not more than X percent is what I am imagining. Anyway, thanks again

lmkwin

If you are looking for something that is 4/10 percent off yesterdays High then it will be in the upper half of the range. Well, unless the Range (H - L) is less than 8/10 percent.

If StockCharts ever adds the ATRP then it would make it clearer for display on a chart. Until then, I believe, Price Channel 1,-1 is a way to visualize it but it's not going to show the % value of the range. Only the High, Low, and mid point of the range.

You could also play around with the Chandelier Exit parameter 1,0.5. The CE looks at ATR for the specified period and allows you to select the number of ATR for the offset of the line. So 0.5 is half of the ATR for the period. CE defaults to working off the high. To work off the low, 1,0.5,short would be the 1 period parameter and 0.5 ATR's You could put 0.25 for a quarter ATR.

dhall6938

For now it seems to work well but so has many other ideas that I have had over the years. I'm trying to trade a 10 minute chart using daily scans, lol. I can hardly wait for SC to offer intra day lower time frame scans. I'm experimenting ( takes a long time because I can't test any lower time frames after hours ) with long trades above the mid point with filters and short trades below the mid point with filters.

dhall6938

Maybe it's easier to understand what I am imagining and how to control it a little better if I try to describe it like thie. On a day when the range is huge, I may want to insist on something more stringent than price being just above or below the previous day's mid-point while not interrupting a small range day. In my mind, adding a maximum percentage might help.

markd

With that last explanation, I think it would probably work out better if you used range exclusively. The results will adjust "dynamically" with range expansion and contraction, and will work the same way across symbols (whereas per cents of prices may or may not be appropriate across symbols).

So below the mid point but above say the bottom 10 % of the range, or, above the midpoint but below the top 10% of the range would look like this:

// below mid point

and [close < low + [range * .5]]
and [close > low + [range * .1]]

// above mid point

and [close > low + [range * .5]]
and [close < low + [range * .9]]

I'm not clear on how you are using values from yesterday and today, so I left out the time modifiers. You can use the time modifiers - even with range, I think - like 1 day ago range instead of high minus low - and it should still work (provided you keep them all straight, of course).

dhall6938

Thank you again. It's so interesting to me. It's hard on my brain but I enjoy it quite a lot. I'm going to play or experiment with your range idea above. Not to split hairs but range will not contract, it will only expand. Also, I am not certain about using percentages either, but I am trading only SPX and NDX....no stocks at all. I use leveraged ETFs but I take my signals from the two indexes. I had four winners and one loser on Friday last week. That's pretty good for me. I am out on the close of every day so it's day trading and I have never been able to be successful day trading over about a weeks time even though I have tried a thousand times. I'm optimistic again but I have been hopeful before. You guys are great

lmkwin

Range is a function of High for the period - Low for the period. It can contract, expand, or remain the same.

Range > 1 days ago Range
Range < 1 days ago Range

Weekly Range > 1 weeks ago weekly Range
Weekly Range < 1 weeks ago weekly Range

Monthly Range > 1 months ago monthly Range
Monthly Range < 1 months ago monthly Range

Each pair will produce different results.

Be careful on the Monthly in particular. Running it today will compare April vs March. April has 1 days worth of Monthly range at the moment. So perhaps 1 month ago vs 2 months ago would be more "useful".

dhall6938

SC must be down today, my scans won't generate anything today. Range for the same day can only expand. Once the high and the low are established they only change when the high moves up or the low moves down. I don't mean to be argumentative, it doesn't really matter for this situation except for comparing to the previous day.

lmkwin

You are correct on the daily range @dhall6938 . I wasn't considering daytrading.

markd

When I said range will contract and expand, I was thinking of the effects of using per cent of price vs per cent of range to establish the boundaries you seem to be looking for.

The number you get from per cent of price (e.g., the low) will land in a different part of the current day's range every day.

So lets say some per cent of the low today is 20. It might vary over the day or from one day to the next between say 18 and 22, but for simplicity, lets say 20. If the range for the day so far is 30 pts, the low plus 20 lands in the top third of the range - way off the low. But if the range so far is 80 points, it lands you in the bottom quarter which might be better but still might not be acceptable.

But if you use low plus 10 per cent of range so far, you will always get a value that is 10 per cent above the low no matter what happens to the size of the range within the same day or from day to day.

dhall6938

I understand your thoughts and I think your right about that it's more what I think I want to try.

dhall6938

I have tried day trading a gazillion times over the years just to give up again and again. So far, this effort seems to be better than all of my other efforts in the past, as best I can recall. Can't start counting my chickens just yet but this one makes my heart go pitter patter right now. Thank you guys one more time. As it turns out, the daily criteria helps a great deal to filter out a whole lot of losing trades. If this works over time, I'll come back to thank you again. David