| Seeker |
|
Beta Testers, Member, Platinum User, TeleChart
|
| Registered User |
|
|
|
|
| Unsure |
|
| Thursday, October 7, 2004 |
| Saturday, October 18, 2008 9:23:23 AM |
35
[0.01% of all post / 0.00 posts per day] |
|
I have Blocks. I have TeleChart. I'm looking for data.
I'm looking for NYSE stocks only daily market activity data: advancing issues, unchanged issues, declining issues. I assume that advancing + unchanged + declining = total stocks traded, but I'd like to see the total stocks traded (as a cross check).
I'd like to put that data in a table with a daily date, the DJIA closing price, and the NYSE Composite closing price. Then I need to analyze the table.
I don't know enough yet to write code for an indicator. Data analysis comes first.
Can this be done inside Blocks or TeleChart, or can I export that data for analysis in a spreadsheet?
Thanks
Jim
If you have all the answers, you don't understand the questions.
|
T2100 is a very useful indicator, and I refer to it frequently as part of market analysis. I would like to explore studies that consider other momentum and reversal relationships in the underlying data -- the number of stocks traded, and the numbers of advancing, declining and unchanged stocks. I don't see a way to locate that data and put it into a Watchlist tab data element, a PCF, or an EasyScan. Am I missing something in where I should be looking?
If it is not available through TeleChart or Blocks data, can you suggest a data source?
Thanks for your help.
|
|
Thanks! That's a solution that avoids the need to maintain several copies of a PCF! This is one case where less is REALLY more! You've made my day!
|
When I create an EasyScan using a PCF that returns a Boolean value (True or False), the scan editor provides a dropdown selection box to pick the PCF that I would like to use. It then returns the selected PCF to the evolving scan, assuming that the desired condition is True. In some cases, that is appropriate. In other cases, I need to use False, rather than True. For example, another TCNet user and I are sharing PCFs that identify overlapping sets of symbols. We would like to be able to create scans that produce the intersection set (both conditions true) and the outliers (the union set minus the intersection set). This could be accomplished by creating an additional PCF for each case, where the second PCF multiplies the value returned by the first PCF to negate the condition (flip the value from True to False or False to True). That's another item in a long and ever growing list. It would be nice if I could use the scan editor to set the desired value for the Boolean condition to true or false as needed. Is this there, and I missed where to find it, or is this to be addressed in a future upgrade?
Thanks for your help!
|
|
Thanks, Craig! I don't normally think in terms of Ctrl-V as a way to paste. That seems to work when mouse paste and Shift-Insert don't. Thanks for your help in this.
|
I'm running 6.0.13. More often than not, as the new versions appeared, the ability to copy and paste text from Worden Reports -- especially the current note -- has not worked. That function has been reliable in the production version of TCNet. This applies to selecting text and then copying with either Ctrl-C or Ctrl-Insert. Any suggestions on circumventions? I'll be converting my remaining watchlists, watchlist tabs, PCFs, scans, indicators and templates this week-end. I use that function daily.
Thanks for your help.
|
Thanks. I appreciate the input. Sometimes constraints can be worked around or overcome, and sometimes they can't. We'll get on with this the best that we can.
Thanks again.
|
I am preparing to do a longer term study of TRIN and TRINQ against the major indices and some index components. Because of the nature of TRIN and TRINQ, the logarithmic scale is limited to 0 to 10, with the center being 1.00. Is there a way to control the origin as well as the scale differential? That would allow a user to set a template for normal charting (origin = 0, only positive values allowed), and a different template for some kind of special charting (e.g., origin = 1 with positive values allowed, or origin = 0 with negative values allowed).
I don't see a way to access this kind of display control using the Chart Scaling control window. Am I missing something, or is this beyond the limits of TCNet's designed capability for 2000 and 2005?
As an alternative solution, could I use a PCF to generate a data value which could be plotted in the Middle or Bottom chart window?
Thanks.
|
I'm in too, in a tax deferred account for hopefully distant desires or needs. I read Peter's comments to mean "two or more years" -- maybe "three or more years." This isn't a trading position. I'm not sure that there are many good long trading positions in this market.
A couple of suggetions that you might consider before your next foray in this direction. First, don't jump in, taking a full position in one big buy. Volatile stuff like this can move a long way up or down on each swing. Look at the behavior of this particular equity for a good chunk of time -- get a feel for how much and how quickly it moves up, and especially how much and how quickly it moves down. Based upon what you see in its prior behavior, estimate the size and likelihood of further gains from where it is when you're looking at it, and the likelihood, size and speed of a correction. Based on that, draw up a Plan to manage this long term position. What event should cause you to take your first step in? How big should that step be? If the water's cold, what should cause you to get your foot out? If the water's fine, what event should take your next step in? What should cause you to run for the beach? At what point do you reach a full position -- you're swimming? What should put you back on the beach? When you get to the beach, how much of your trading capital do you PLAN to still have? Even if the water is fine, when should you plan to stop swimming? Lunch? Will you go swimming again during or immediately after eating?
Second, do some data sheet work. This may end up happening before you answer many of the questions above, because you should feel the need for insight before you can make thoughtful choices. I've done 16 of these on low beta stocks like GDT and high beta stuff like SIRI and ARBA. You'll learn a lot just by going through this process for anything that you think you might take a position in. Here's my drill:
1. Download daily Open High Low Close data for your potential date. [Remember, you're dating, not getting married. You're not signing up for the "or worse" part unless you're planning on going down with the next Entron, WorldCom or Tyco.] You want enough time that you see a meaningful range of ups and downs in the context of market experience. I went back to September, 2002 as my starting point -- this catches the broad recovery from the bear market bottom. Sort these ascending by date.
2. Add two columns -- Hi (aka cumulative high) and % Change from prior day. You need a formula to do the high: If today's close > yesterday's cumulative high, use today's close as the new high. Otherwise, copy yesterday's cumulative high.
3. Add another column -- Below Hi %. This is Today's Close / Today's Cumulative Hi - 1, displayed as a percent. I format these showing x.x% with negative in red.
4. Propogate those down through all of the dates.
5. Now look at what you have. Looking down the % Change column, what is a big move up or down for this stock? 4%? 17%? Highlight notable moves up and down. Then go down the Below Hi % column. This emphasizes the magnitude of pullbacks. If you don't see a lot of double digit red numbers, you're doing something wrong or you have found the stock of a lifetime. Those numbers will probably convince you that you need some kind of loss cut strategy for this stock.
6. Add pairs of columns for each loss cut level that you would like to evaluate. I started out with 8% (IBD's absolute worst case limit), 10%, 11%, 12% and 13%. These are really easy. If the loss cut % is in the column's bottom heading line (say, row 3), then the loss cut for this cell is Today's Cumulative Hi * (1 - % in row 3). The right hand of the column pair is an IF that tests Today's Close < the Stop just calculated in the cell to the left. If less is true, display "Sell", otherwise display a blank. Propogate that for all of your history. Now look at that. Do you see any patterns that would shake you out of the position? How loose does the loss cut need to be to avoid most of these whipsaws? Remember that this means that you are giving up that much of your ultimate trading capital as the cost of staying in the trade longer.
7. Pick a loss cut percentage that you will tentatively live with for this position. For the rest of this iteration of the exercise, you can'[t change that number.
8. Make a copy of the worksheet so far. Start at the top (e.g., 3Sep2002). What is the price that is your loss cut percentage ABOVE the closing price on 3Sep? Note that number. Add another column for the Reset Buy Point. Put that number in the cell on the 3Sep row.
9. Go to the next day. If the price closed lower, calculate a new lower buy point based on the lower closing price. Repeat this until the closing price is ABOVE your tentative Buy Point. Chances are, this is a new Hi. I marked my Closing Prices for buys in bold green, and my sells in bold red, so that I could easily find them again during review. Use this process to trade all the way through the exercise period. For this simple exercise, assume that all of your trading capital goes in a single buy, and comes out in a single sell for each trade. After each sell, you must reset a Buy Point using the same strategy you used to set the first one. Set a Buy Point. Buy. Hold. Get stopped out. Reset the Buy Point. Keep doing this until you run out of market history. Another column right of the Reset Buy Point gave me a place to record the percentage gain or loss on the individual trade. It's easy to add another column to the right to turn that into a cumulative update of sample trading capital.
10. Chances are, if you always bought after the first close above your Buy Point, results didn't turn out well. What happens if you buy at the close of the NEXT trading day (row)? If you run through the market data, faithfully executing this buy and sell discipline, what do you see as a pattern of trading results for this strategy on this stock with this market history?
11. Is buying the first day after a buy signal too quick? Try the example again looking for two consecutive closes above your buy point. Is the loss cut too quick? Try two consecutive closes below your sell point. If that looks too loose, try three closes above. Does an asymmetrical approach work better -- more days to confirm before getting in than to get out? Does more than the next day on a Loss Cut target FOR THIS STOCK REALLY add to returns? Or does it just add complexity and anxiety?
Note that we have focused on price here -- price alone. That is the most important indicator. In this exercise, avoiding the use of other indicators (MS, TSV, BOP, MACD, OBV, et. al.) clarifies what you are trying to learn about the character of a stock at the onset. After you have a feel for the way that price acts over time, try other indicators to see which ones are relevant to predicting or confirming price behavior FOR THIS PARTICULAR STOCK and which are not. If an indicator doesn't work for a particular stock, ignore it. But consider that some stocks change character and behavior as their price (and market cap) move up or down into different ranges. You may need to change indicators as a stock moves from range to range.
In a broadly uptrending market, did an evenly balanced strategy of delay after a price-based buy or sell trigger increase or decrease your trade capital for this stock? Should you adjust the trade management strategy to improve the likelihood of better results, or should you look for a different trade target? You can easily test a number of approaches to see how sensitive the results could have been to the choices that you might have made. You can also consider placing a trading floor under some stocks (or all stocks) because the pattern of price movement is too unpredictable and rapid to for you to manage in your trading circumstances.
Now look at the part of the exercise trade history for 2004. When would you have gotten in? Would you have been shaken out? Would you have gotten back in? Would you still have been in at year end? What does 2005 look like?
You need patience and discipline to identify and wait for places to get in, where the risks to your trading capital are reduced. For longer term positions, you need to manage the building of a position, not just of a single trade. To survive, you have to learn that anything can go against you. I'll bet that any of the choice sets that you picked for this exercise went against you most of the time in the last 28 months.
If you think that it's just CMGI, try SIRI with a buy in August, 2004. That's another low price, very volatile issue that could have made you and lost you a lot of trading capital in just the last six months.
When you look at low priced stocks like CMGI and SIRI, consider the change of character that occurs as they move under $2, then under $1, then back above $1, then back above $2, then above $3. How does the change in price affect the proportion of institutional investors to individual investors? Did something attract the attention of hedge funds -- either coming in long or going short? You may want to run some scans to see how under $5 stocks last September have done in January as compared to under $10 stocks, under $20 stocks, $20-50 stocks, and over $50 stocks. Apple is a good example of an over $20 stock that more than tripled in 2004.
Finally, consider beta and leverage. I think of beta as a kind of leverage on the rate of returns. Margin is another kind of leverage. Leverage can work for you or against you. Five months of gains from the lows of 2004 were wiped out in three weeks in January, 2005. If you ever thought about trading high beta stocks on margin to leverage your returns, the exercise that I described here should be a wake up call. You can easily see the impact by adding a few more columns to your exercise spreadsheet. Here's how to do that:
Put an "account cash balance" in the heading of a new column -- like you did for the loss cut columns. Put a number in that cell to represent the trading capital that you are tentatively willing to risk on this position. Next to that, put the "margin equity rate" that your broker requires for the stock that you're considering. [Yes, these are different for different stocks, and yes, brokers can and do change the rates without telling you, and yes, brokers can and do change the rates both down and up. Of course, rates tend to go up when prices are going down.....] On the rows where you identified a buy event, reference the buy price in a formula. To keep it simple for this example, ignore margin interest as an expense, but this is real and it's not trivial. Let's also assume a Margin Call Loss Cut Level of 8%. Your loss cut price is 8% below your proposed purchase price. [Actual results may vary with market conditions and trading habits.] Your buying power at the loss cut level = [Account Cash Balance / Margin Equity Rate * ( 1 - Margin Call Loss Cut Level %)]. Examples:
$20,000 Account Cash Balance, 8% Margin Call Loss Cut Level
100% Margin Equity Rate for a $5.00 price = 3680 shares for $18,400 of buying power
60% Margin Equity Rate for a 10.00 price = 3066 shares for $30,667 of buying power
50% Margin Equity Rate for a 15.00 price = 1333 shares for $36,800 of buying power
40% Margin Equity Rate for a 20.00 price = 2300 shares for $46,000 of buying power
30% margin Equity Rate for a 40.00 price = 500 shares for $61,333 of buying power
As the price falls, the higher margin rates that you could have applied in 2004 really bite now. Compare CMGI at 100% (no leverage) to ARBA at 60%, to ASPT at 50%, to A (Agilent) at 40%, to CREE at 30%.
My suggestion: until you have a track record of money making trades, and have developed a process to respond to events within one market day, minimize your use of low priced stocks to leverage returns and completely avoid margin. Patience is a hard thing to learn, but it is essential. You must be patient while you learn, and apply lower risk approaches to the limited trading which you allow yourself. It's slower and harder to replace trading capital lost than it is to lose it. Five months of market-wide gains were wiped out in just three weeks, with the worst damage in the first and third weeks of January. If it takes you as long to get out as it took you to get in, then you are programmed to lose trading capital. You must be able to get out ten times faster than you got in. How much time does it take you from the time that an event occurs until you can appropriately act upon it? If it's more than one market day, you should strongly consider automating your sell decisions with Stop Loss orders. If you're making money with an upward moving price, great -- move the stop loss price up to track your gains. If the market takes you out, so be it -- you planned where your trading capital floor would be no matter what. In a declining market, cash isn't a bad place to be, however you get there.
|
|