Hi my friends, this is lonelygrass. Just drop a quick update on my latest study.
In addition to trend following, I have realized that mean reversion trading systems can also be profitable if implemented correctly. This system is based on the idea that price will ultimately revert back to its mean after experiencing extreme price movements.
With this in mind, I have deployed two notable indicators, Bollinger Bands and RSI. When Prices go outside of the Bollinger bands and RSI has arrived oversold/overbought levels, there will be a clear indication on the graph that one should possibly trade against the current trend.
Short term and long term versions are available for use, just choose the one you want. Happy trading and stay hyped!
love this tool. would you be able to share the script with me?
iushakov1
⋅
@lonelygrass, is there a chance you can share the code? I've been trying to play with similar ideas and would like to know the settings you use. Thanks!
mohamadchemist3570
⋅
Hello, thank you for providing it for free, I just had a question, can you send me the script so that I can make a robot that buys whenever the candles fall below the line and sell whenever they rise?
tobyz7
⋅
Hi lg, toby here. Just wondering if you would trigger trades when a stock approaches the upper and lower band limits. I note that you don't indicate Buy signals on the above chart (presumably because price never actually crosses the indicator lower line). But if you bought and sold on close approaches to the line, you'd have a 10% gain between early July and early August and another 10% gain between late September and late October. Another way of looking at this is that possibly your band limits are slightly too wide and a band narrowing could give better results. The traditional parameters are not optimum for every stock (and may drift over time).
That's one problem I have with TA (including my own work): We don't spent enough time optimizing our parameters and, once optimized, we almost never test for drift. For those of us with full daytime jobs, who try to do our analysis at night, this problem is almost intractable.
Further, when you see a Sell signal, as in your chart (I'm not trying to pick on you), and then never have a clear Buy signal, several months down the road you may find that a great stock has gotten away from you. Often, the best analysis is to hold and have DEEP pockets. Otoh, I have noticed stock charts that drop from $60 to $3 in less than a year.
What a quandary! A good approach (for me) is to be really sure I'm buying a quality stock and then use a ZigZag to quantify the percentage dips off peaks and then allow the stock its down runs before the next pivot, After a few months with a decent profit the Sell Stop can be tightened up.
Buy Low/Sell High .... a fantasy!
Toby
lonelygrass
⋅
@tobyz7, Thank you toby. It is so nice of you to offer your comments on this TA. Before publishing this script, I have found that users could rarely get both buy and sell signals with this, especially in a strongly trending market. I wonder this would work best in a ranging market because of its lagging nature and long lookback period.
Your review is such an insight to me! I would post an improved form of this TA, might be a version which follows price more closely and responsively. Alternatives may be donchian / keltner channels. However, there is a long way to go testing and accessing different types of setting. But I can make sure you can see a better version soon.
Besides, I am quite in favor of your ZigZag trading idea. I am also looking for ways to find quality stocks to buy and hold. Would you like to share the system you use to explore good stocks in terms of technical analysis? Any articles or journals ?
Thanks again.
Lonelygrass
tobyz7
⋅
@lonelygrass, Thanks for your comments, lg. Over the years, I’ve used several TA systems for buying stocks. Some work fine for a while and then sour (the ‘drift factor’, I suppose).
Currently I like the standard MACD Crossover, though it lags more than I like so I sometimes try to cheat the system and buy in anticipation of a crossover. Sometimes I get burned. The difficulty is in the confirmation. MACD lines can cross up and down in just a few days, and I don’t like to hold for so short a time. So I look elsewhere. Another system I’ve used is the RSI14 with a buy on crossing up over50 and a sell on the first cross down below 70. Then what do you do with a stock that crosses 50 and then goes flat (maybe oscillating between 60 and 40) and never crosses above 70? Too many false signals.
Then I noticed ZigZag. The rap against ZigZag is that it has no predictive power, and that’s true. But I observe that a good stock has long Zigs up and short Zags down. I like that. The question is the % parameter. Too small and you’re forever trading. Too large and you’re missing valid signals or you hold beyond your comfort zone. Too small is 2%, too large is 50%. I’ve settled empirically on 7%.
The problem I have is programming a scan of the stock universe to find stocks that have just crossed 7% on an up leg. And if I do create a good scan, what auxiliary criteria do I use to narrow down a long list?
This has gotten me somewhat away from pure TA in creating a potential Buy list.
Through TDAmeritrade, I subscribe to ‘Market Edge’, which is a strongly TA-based rating service. I want to consider stocks they raise to a Buy or Strong Buy rating. They also provide a Sell Stop price which I find useful.
Other services I pay attention to are Zacks Ratings (and its subscription advisory letters), IBD scores, ChartMill set-up ratings, Tradingview.com, StockCharts.com, and TC2000.
I took a TC2000 class earlier this year to learn how to better use the software. They presented several scans and other techniques. One is Top-Drill-Down. From the best sectors, find the best industries, and then choose the best stocks. Very plausible, but I set up a hypothetical paper portfolio with 9 stocks (3 sectors, 3 industries each, 1 stock each) and its performance is just so-so.
One notices that, when a stock hits a new high, it often continues to hit new highs. I have set up a TC2000 scan to report stocks with a 500-day new High (within the past 3 days), Price > $3, and Volume > 100000. Also (a class suggestion) I run the same scan for a 500-month High. In class I raised an objection (‘There aren’t many stocks that have been trading for 40 years’) and got the response (‘Well, that’s the point’).
From these multiple sources, I try to keep a Potential Buy list of about 10 stocks. How do I choose which stock to buy? I’m doing it mostly by consensus from the following:
TDA Research Team: Accumulate
Zacks: 1,2 Rating
Market Edge: Strong Buy, Buy
StockCharts.com: SCTR > 70
IBD: Comp Rating: >79
ChartMill.com: Analysis Tech Rating GE 7 and Set-Up Rating GE 7
TradingView.com: Up Trend
Zacks Newsletters: Portfolio Buy
For each condition satisfied, I assign a value of 1 point and choose the stock with the highest value. It is a TA-based system, but it’s not my TA. So far, it seems to work well.
A final comment: Most of the stocks I sell will ultimately come back and rise above my Sell price. This is disheartening and argues for a long-term Buy-and-Hold strategy (hoping never to hit the disastrous big-time loss). My only hope is that the stock I buy will outperform the stock I sold. That motivates one to keep on improving one’s Buy and Sell systems.
Take care, lg. Maybe we can continue this discussion.