Alerts

dhall6938

I am curious to know what purpose the hourly alerts, or any other alert, have in lieu of the continuous alerting. Do you know if the hourly alert sends out a notice at the end of each hour if the alert criteria has been met any time during the last hour even if the criteria has unfolded before the hour is up? I have learned over the last 6 months or so that my scans / alerts trigger during the day often, but completely unfold by the close of the day with no valid signal. That happens a lot.

dhall6938

I will add that I have considered waiting until after the close each day to run my scans or check my alerts, but the overnight gaps can be awful and very detrimental to my trade plan especially for stop placement. It gets kind of hectic trying to place orders right at the market close and too many times the order isn't filled until later at a worse price.

markd

Continuous alerting would require a lot more server and communications capacity, which costs money.

Here is the documentation explanation:

"Since alerts are based on daily data, users will only be notified once per day that a particular alert has triggered. For example, if you have an alert that looks for Intel to cross above $22 and the stock crosses above and below $22 several times during the day, you will only be contacted the first time the price crosses above $22. "

https://support.stockcharts.com/doku.php?id=alerts

dhall6938

I read all of that this morning, but I still don't understand why they offer a variety of alerts. Is it for their benefit to save them money?

dhall6938

I have about 40 scans set up to run continuous, but I have learned that I can't act on them until the close without some serious risk of being stopped out the same day. If it helps them save some money, I could change all of them to an hourly notification?

dhall6938

I am curious about something else that maybe you'd know. Not sure that I can word my question very well but here goes nuttin. If using the hourly alerts, and the alert triggers early into the hour but has gone away, come undone, by the end of the hour, will the alert still trigger or does the signal have to be valid at the time of the alert?

markd

As I read it, if an alert runs continuously, it will trigger the first time the condition is true, and not again.
If it runs hourly, it triggers the first time the condition is true at the time of the hourly scan, and not again. If it was true before the hour, and/or after the hour, but not at the hour, you won't get an alert.

I would guess that the hourly alerts tend to be more "stable". That is, if the condition occurred during the previous hour and persisted until the hour, maybe it's more likely to not come back. But it doesn't necessarily have to be that way.

That's how I read it. I don't use them, so I don't actually have experience with them.

dhall6938

Thanks. What you describe about continuous alerts is true, I have experience with them. I get too many false signals that way though, so I am going to change them to the hourly version and watch. If the hourly works as you describe, I think it would help to eliminate some of my false signals. I'm gonna guess that if the conditions are met, the very next hourly alert will trigger even if the signal has completely unfolded. I'll watch and see how it goes. Thanks

dhall6938

I have come to enjoy the alerts because I don't have to run my scans all day long every day. They aren't perfect tho, especially if you have some weekly criteria in the scan. It can be late so I have tried to figure a daily equivalent to the weekly or an approximation as close as I can get. In any case, they make things easier

lmkwin

I use Alerts, both Price Alerts and Advanced Alerts. I have never had one sent twice in a day, including my ones set to Hourly. But mine are set to Email, if that matters. I don't have any set to SMS or Workbench notification.

I have this one set to tell me when the DistFromSMA crosses 0 or a PctRelative threshold is crossed, for a group of ChartLists.



Most of my Advanced Alerts are set to run After Close though. And I don't have any Advanced Alerts that use price fields Open, Close, High, Low, so I don't know how that would play into it. The only Price field you can use on Price Alerts apparently is Close. As we know, intraday, the Close is a moving target, so the Low will also be the "Close" and the High will also be the "Close" at some point during the day.

markd

@dhall6938

I wonder if you are familiar with Van Tharp's "R" concept of risk.

I think his book "Trade your way to Financial Freedom" is still in print. But if you don't want to read the whole book (or spend the money), here's a pretty good intro.

This is a teaser for a practice trading game available on his website (he's gone, but his company lives on). I have no opinion on the game because I haven't played it. But the explanation of R further down is very good. It might have something you can use for setting your stops, since that seems to be an issue in your posts.

https://www.vantharp.com/trading/wp-content/uploads/2018/06/A_Short_Lesson_on_R_and_R-multiple.pdf

dhall6938

I tried to set my alerts to email but there is some kind of glitch within SC that won't allow it. I'm sure that it's something that I have done wrong. Trying to figure out what it is that's wrong when SC won't communicate with me has made it impossible to figure out how to correct it. I have been using advanced alerts sent to my phone running continuous all day. I no longer think I need that kind of alert, so I am going to change most of mine to an hourly alert instead. If the hourly alert does not trigger and send me a notice unless the criteria are met at the time of the alert, it may help save me some false signals. And it should help their systems run more efficiently and it sounds like it may save them some money as well. A win win win

Thanks very much and I have not heard of Van Tharp. I stopped reading traders books a long time ago after reading Bill O'Neil's book and the books that he recommended. I tried for years to master what he describes, and I could never do it. Probably my personality is the reason, and I didn't know for two decades that he used weekly charts mostly for the pattern and the daily for entries and exits. At some point, I decided that if I could not develop my very own unique thing, I would just never be a good trader. It's frustrating, very challenging, overwhelming and just plain hard to master. After decades of trying as hard as I could, I have finally developed a trade plan that works very well trading indexes, not stocks, using leveraged ETF's. Using EOD data is all I have at SC for back testing my plans. At first glance it looks, and it seem easy but it's not easy to implement. I don't know why it took me so long, but I have finally realized that way too many times I get an intraday signal that completely unfolds before the close and I lose money, usually the next day on a gap in the wrong direction. Trying to run my scans that may not trigger until after the close, all right at the close and enter trades all within a minute or so of the close is nerve wracking as hell. It's hard to do. I have not really done a serious back test to see how well I'd do if I just waited until the next day opening to enter orders giving the scan machine all night to do its thing and confirm the previous day's signal, but I have done enough to see that overnight gaps hurt often. I usually get a much worse entry price and if I use the same stop, which is a few pennies below the entry day, I lose a lot more when I am stopped out. Having said all of that, my plan almost always doubles my whole account in a years' time using EOD pricing that sometimes doesn't occur until after the close. My little plan, in theory, works so well that I am probably able to just wait until the next day to enter orders and that is my next chore to accomplish and see how I would perform. I also have a gazillion other ideas that I want to try that have nothing to do with my primary plan.

dhall6938

The last time I ran a whole year of back testing was several years ago but the last couple of times I did it, I had about a dozen entries in a years' time, both long and short. My average loss was less than 3% and much closer to 2%. My worst loss was about 8%. I had almost a 7-to-3-win loss ratio and profits averaged well over 15%. If I can find a way to execute this plan, it's gonna be wonderful. In the meantime, while I am trying to master the execution of that plan, I am trying several of my many other ideas.

markd

If you are making money in real time, you must be doing something right.

The main point Tharp makes is, as long as your profits are consistently several times your risk, you can be wrong more often than you might expect and still make money. He defines risk as the difference between your entry and your stop. So if your stop is a $1 ("R", for risk), you want a profit target of say 3R to 5R - $3 to $5. If you don't see 3-5 dollars available in the trade when the stop has to be a 1 dollar, you shouldn't take the trade.

Of course, you have to have the temperament to take 3 or 4 one dollar losses while waiting for the 5 dollar payoff to come through. Not everybody can do that.

Also, there is something to be said for wider stops. There is a certain amount "noise" in prices - random fluctuations that don't disturb the trend. Provided the necessary profit is there in the long run, a stop of a few pennies probably isn't enough to let the stock get back on track. Have you noticed when you get stopped out whether trade would have been profitable after the drawdown?

But there is more to be said for better entries. Many books recommend, or at least imply, that you should enter on the signal. I think that is just asking for trouble. The signal of course is in the direction of the trend (for instance crossing an MA, or MACD Line crossing Signal, etc.). Everybody sees the signals and pros know how much counter trend price movement other traders can take, so they push prices against the trend until other traders cave or get their stops hit. THAT is the entry moment - the REACTION to the signal, not the signal itself. And then place the stop below the reaction low. It's less likely to get hit, because a second reaction less likely, and if it does get hit, it probably means the trade really was wrong from the start. Of course sometimes there is no reaction and you miss the trade. But so what, you lost an opportunity, not money in hand, and there is always another trade waiting.

dhall6938

Yes sir. Understood. If I can figure out a way to enter and exit a trade only on the day's that I end up with a valid signal, I'll make a ton of money. Since you've been so willing to talk about it, I'll add that I looked over some charts yesterday to again see what might have happened if I just waited for the next day open to enter or exit. Yes, bad entries could be avoided completely not losing anything but a trade, and I realized again, that taking profit is just as hard as entering a trade. Too many times the dam gaps take way too much profit even turning a decent gain into a loss. It may still be a profitable system but not nearly as good. I have had all kinds of ideas on how to handle it. I have tried using RSI(14) as a filter and that one does help eliminate a lot of faulty short entries. I Have tried watching an hourly chart and setting an intra day stop at the low of the signal candle and that helps sometimes but there is still a loss and that does not help with exiting a profitable trade. The gaps are just an awful thing to have to deal with.

dhall6938

BTW, the vast majority of my trading and what I am referring to here is trading indexes, not individual stocks. The indexes are so much easier to predict than stocks. I know that you have no idea what it is that I am trying to do but I think that I have spent so much time trying to pin down precise entries and exits that I may have overlooked what might be the best entries and not having intraday information available to the scans is a huge limiting factor. Anyway, here is my latest filter for both longs and shorts. These don't always give me the very best entry but, without trying them in real time, they appear to help eliminate at least some of my false signals. My favorites list 23 are indexes. INDU, SPX, IWM etc. I trade 7 of them. I like to use a crossing criterion whenever I can rather than just greater than or less than because it helps to eliminate a ton of signals that I don't want to see.

[Favorites List is 23] and [[[Close X 1 Day ago High] and [Close > 1 Day ago High]

and [[EMA(20,RSI(14)) > SMA(20,RSI(14))] or [RSI(14) > EMA(20,RSI(14))] or [RSI(14) > SMA(20,RSI(14))] or [RSI(14) > 55.0] or [[MACD Hist(48,103,36) > 1 Day ago MACD Hist(48,103,36)] and [MACD Line(48,103,36) > 1 Day ago MACD Line(48,103,36)]]]]

or [[1 Day ago Low X Today's Close] and [Close < 1 Day ago Low]

and [[EMA(20,RSI(14)) < SMA(20,RSI(14))] or [RSI(14) < EMA(20,RSI(14))] or [RSI(14) < SMA(20,RSI(14))] or [RSI(14) < 45.0] or [[MACD Hist(48,103,36) < 1 Day ago MACD Hist(48,103,36)] and [MACD Line(48,103,36) < 1 Day ago MACD Line(48,103,36)]]]] ]

dhall6938

There is one more thing that I have learned to do the hard way. If I wait for at least the first hour and a half or longer, the morning price action that triggered a signal often evaporates.

markd

If you are trading very short term, and losing a lot to gaps after racking up a profit, maybe you could try setting your stops at or just below or just above a target price ABOVE the current price (on a long), instead of using a trailing stop (or use a one cancels other order). You might be getting gaps because price reaches recent resistance (like a recent high from a prior leg peak, or higher high in the last leg down) and short term traders are dumping there because it was resistance before so might be again.

The usual advice is to "let your profits run", but that's more for swing trading where you expect to to have some short term draw downs ( a couple or three days) in the longer term trend of several weeks. Maybe "defined profit" trading, where you get out on getting to your expected price would work better - or at least cause less heartache.

dhall6938

Possibly true. I have spent my whole time trying not to anticipate a top and I have a fairly complex system of that I'd call swing trading. I have developed a plan to hold on to longs as long as possible but depending on where the markets are within my analysis, I may only be able to hold for a few days or a couple of weeks. A 2 to 4 month run up will produce big gains for me but if it doesn't last at least a month, a gap while trying to take profits can really kill my overall performance. When I'm short, I do my best to get out on the way down and not wait for a reversal like I do with longs. One bounce on a short trade can eat up every penny of short profits and they happen fairly often.

lmkwin

Excellent comments @dhall6938 and @markd .

I like to say that I'm always learning and I've been "investing" for over 40 years. In 2009 I started learning about PnF charts. It took me about 2 years to "figure out" Point and Figure chart uses and then another 2 years to unlearn the process that I had learned from the books and websites. Most teach, using cherry picked examples, some utopia that doesn't exist. I was surprised to see Dorsey Wright was preaching in their books and website, about using their methodology and then seeing that they didn't actually use their methodology in their practice. They just TALKED about it.

Dorsey Wright is one of the pioneers in the ETF world and has created a plethora of 'indexes' based on their "momentum" methodology. The indexes are are used by ETF sponsors for their ETF offerings. DWA used to be based on their "relative strength" methodology, but they found that using relative strength has very poor results on portfolios that get recreated/ rebalanced quarterly. So they changed to "momentum", which still suffers the same pitfalls, but allows more gray area to enter their selection processes. Once they got bought out by Nasdaq, their selection process changed again.

So I unlearned their system as my "backtesting" showed that an alternate method that I had worked on with other PnF chartists, actually worked better and was something that could actually be applied. Kicker was, it had great results. It wasn't until I added a 'final' part to my system that all of it fit better with my personality and intentions to cut losses and let winner run. That last, and probably most vital, part of my "system" was adding something called "patience".

dhall6938

If you are interested and if you can see this chart, the blue lines are valid long entries and the red lines are valid short signals. There are just as many intra day false signals as there are valid signals.

https://stockcharts.com/h-sc/ui?s=$SPX&p=D&yr=0&mn=6&dy=0&id=p27255540789&a=1483748839&listNum=10

dhall6938

You guys are very nice. Thanks

dhall6938

I don't imagine that you're interested, but I think that you could set up an alert to trigger three times a day. I haven't done it, but I may try a version of it in the future if I don't figure out a fairly reliable way to eliminate some of the false signals. Set up scan XYZ continuous, then set up the same scan XYZ hourly and then gain XYZ close.

markd

Actually, if you are taking daily time frame signals intraday, before the daily bar is completed, it's not so surprising that many of them don't work out, since the signal is not REALLY a daily time frame signal. The daily bar hasn't completed yet, so any signal that does occur is based on incomplete data. In practice, a daily signal at 10:30 am COULD turn out to be the same as the final signal, if the value at 4 pm is the same (or further into signal territory) as it was at 10:30.

It's possible that you could get a better price by getting the jump on the 4 pm signal, but I think your experience is that that potential advantage more often turns into a loss.

As I suggested above, being the first one into a trade may not always be the best thing. In a runaway market it could be, but most of the time the market retraces the same ground several times AFTER a signal, giving you multiple opportunities to get in if you DON'T get in on the signal itself (provided you still believe the signal after price moves against it). If you DO get in right at the signal, then you have to ride out those retracements hoping your stop doesn't get hit, so they look like threats rather than opportunities.

dhall6938

I understand that a dally signal is not valid until after the close of that day. As I have mentioned, it's very difficult for me to get into or out of a trade after the close unless I wait until the next day. As I have also mentioned opening gaps can be awful so I could just use limit orders to protect against big gaps on buy orders but that may not work very well with a sell order for a trade that I am already in and need to get out of. The gaps in the chart that I posted in theory worked in my favor fairly often but leveraged ETF's don't always follow their underlying exactly especially on opening gaps. An index might open up while it's ETF might open down moving in the opposite direction. Anyway, I'm still trying to filter out or otherwise mange false signals so that I can get into trades on the signal day to avoid the overnight gaps. If I get into one early enough, I do get a better price that might help to absorb some losses that I'll incur on the bad signals, but it has not worked out very well so far. Not taking any action until the next day would be super easy and no stress at all so that is the next study that I think I need to perform. It takes forever to do a back test manually like I have to do here on SC but I am beginning to think that it's a necessity. Thanks

dhall6938

Today's opening gap up is a perfect example. I scrambled to unload all of my shorts at the close on Friday and I didn't get it all done until about 10 minutes after the close which made me nervous because I didn't think that I would get them all sold. If I had not been able to unload everything Friday, I would have lost a huge percentage of my gains on today's gap up.

dhall6938

To your point that pull backs may be something to consider. I have looked at pull backs after my entry and there are a number of them that do pull back up above my entry, but the best trades just blast out and keep going to I would lose all of them and that doesn't work well trying to get out of an existing position, lol. Anyway, if there is a way to manage false signals it's a very good system.

dhall6938

FYI, I got a bearish signal for symbol SLV at the open this morning. Actually, two minutes after the open. I ran my scan manually to confirm the signal and SLV was selected. So, I changed SLV from a continuous a all-day signal to an hourly signal and it has not triggered. So, at this point I am going to assume that once a signal, no matter how you have it set up, has triggered, it will not trigger again even if you change the alert settings. At this poing it appears to me that the hourly will trigger an alert even if the signal has unfolded by the end of the hour.

dhall6938

Sorry to continue to bombard you but one thought that I have had and can't check without intraday information is that I may just run my scan's during the last half hour of the market day and place MOC orders for everything that is selected. and when I check the scans after the close just place a market sell order for the next day open for those that did not meet the criteria at the close. It sounds crazy but if the plan can double my account in a years' time, it might actually be a valid strategy. I wished they would offer intraday scanning because the only way to verify a crazy idea like this one is by trial and error.

markd

I don't understand how, using daily data, you get a bearish signal at the open. The bar is only a few minutes old. But, if it's working for you, that's great.

I'm thinking it might help to validate your daily signals on an hourly chart, if you're not already doing that. In other words, get the signal from the daily scans, but check the patterns on the hourly.

For instance, you said, you got a bearish signal on SLV this morning (8/21), but the hourly chart shows it's been basing and broke above its MAs this morning (on my hourly chart)



You can see that it stopped making new hourly lows last week and moved sideways into its MAs (green ma is ma of highs, red is lows), and broke above them today. That's not bearish, although it might have looked like it on an incomplete daily chart/indicator.

lmkwin

Although you mention only looking at specific ETF's and such, here's two issues you might encounter with your MOC plan/scans. I could be wrong but I don't believe that the After Hours and Pre-market trading are captured in the scan engine on StockCharts. Although the impact of those two lightly traded markets can have a significant impact on the Open price, the scan engine isn't going to alert you to anything until 20 minutes after the open the following day.

If you are using TD Ameritrade and their ThinkorSwim platform, you can set up all sorts of custom scanning and alerting and displaying, including intraday, down to the smallest time frames, including the extended hours market activity. It's a pretty slick platform if you are into setting up or modifying indicators of importance to your system. I didn't start truly modifying stuff until last year. As I may have mentioned prior, I always try to be learning.



Just an FYI, they DO NOT offer Point and Figure Charts though.