| |
| Simulator Updates / Progress Notes |
« View previous topic :: View next topic » |
| Author |
Message
|
| nelaina |
 Posted: Thu Apr 13, 2006 4:07 pm Post subject: |
 |
|
CFO
|
|
| Back to top |
|
 |
| Dave Rathbun |
| Posted: Thu Apr 13, 2006 4:22 pm Post subject: |
|
|
CFO
|
| nelaina wrote: | | what is "churn rate"? |
That's a very good question. 
It's supposed to be a measure of how often a portfolio manager turns over or "churns" their assets. For example, if a mutual fund has assets of a million dollars, and they make a million dollars worth of trades during a time period, then they have a 100% churn rate. However, it's not that simple. 
The basic formula that I found doesn't really apply, I don't think, to the game here. But I'm trying it out anyway. The bottom line is that a higher number means you've moved your money into and out of the market more times than someone with a lower number. Someone with zero trades has a churn rate of zero. Interestingly enough, someone with only buys (or only sells) also has a churn rate of zero. It's only when you buy then sell then buy again that your churn rate goes up.
So if you start with 10K and buy two positions of 5K each, your churn rate is zero. But if you start with 10K and buy 10K of stock, sell it as 12K, buy something at 12K, sell it at 15K, well, your churn rate is higher because you keep turning over (churning) a sale into another buy, and into another sale, and so on.
That's the way it's supposed to work, I'll post the actual formula that I'm using later, as I said earlier, I'm supposed to be offline tonight.  |
|
| Back to top |
|
|
| InvestingMac |
| Posted: Thu Apr 13, 2006 4:48 pm Post subject: |
|
|
Investing Manager
|
Looks great!
I particularly like the "trades this month". Should prompt a few more peaks between portfolios.
Interesting that you decided to implement it in one of my most active trading months. Previously I have been sitting on slow longer term growth. I decided to start switching things up to test how the new system deals with buying and selling positions. So far so good. |
|
| Back to top |
|
|
| nelaina |
| Posted: Thu Apr 13, 2006 4:53 pm Post subject: |
|
|
CFO
|
| ok, makes sense...grim's churnrate is pretty high! |
|
| Back to top |
|
|
| vetelmo |
| Posted: Thu Apr 13, 2006 5:14 pm Post subject: |
|
|
Moderator
|
churn it up BABY!
I like the trades per month stat for sure! 
Thanks for all the work you've been doing Dave!! |
|
| Back to top |
|
|
| Dave Rathbun |
| Posted: Thu Apr 13, 2006 5:39 pm Post subject: |
|
|
CFO
|
I'll admit the reason I wanted to put in the monthly trades and the churn rate was to see if my assertion that active traders will now have an advantage in the game turns out to be true. I've added churn and trades to the rankings page (the member list) and observed an interesting situation.
Top member as I write this is Grim, with 20 trades and an astounding 500%+ churn rate. This means that he's turned over his entire portfolio five times already this month.
But number two on the list is nickspcs, with only one trade.
Number three is jdunn72, with 55 trades  but only a 27% churn rate. As I posted earlier, I'm not sure the formula for churn is working the way I think it should, but this implies that jdunn72 has turned over 27% of his portfolio this month. That would mean that if he started with 100K he has bought, sold, and bought again using about 27K of his funds. Or sold - bought - sold - bought - sold ... since he has so many trades (55) it could be that he has 25 buys and sells, each for about 2K or so. Anyway, the numbers certainly do raise some questions, and might make the "sneak peek" feature more attractive.
If you want to know what nickspcs single trade was to get that 9% return, maybe it's worth a few e$.
BTW, the list is also sortable by monthly trades and churn percent. jdunn72 is leading the active traders with 55, PDJ Enterprises is second with 38 (but a 65% churn) and vetelmo is third on the active trader list with 39 trades. Fourth on the list is shaks who has a 36% churn rate and a -9% return rate.
crazycow is a hair behind Grim in the total return race, and has a 100% churn on only 4 trades.
Interesting stuff. |
|
| Back to top |
|
|
| Im Not Warren Buffett |
| Posted: Fri Apr 14, 2006 7:01 am Post subject: |
|
|
CFO
|
| While we are adding things to the rankings list, why not net worth? We have portfolio value and e$, so would net worth be possible? |
|
| Back to top |
|
|
| Dave Rathbun |
| Posted: Fri Apr 14, 2006 7:08 am Post subject: |
|
|
CFO
|
The last time I tried the screen got ugly as there's not enough space. Plus net worth has to be caluclated each time I need to display it, unlike the cash and portfolio value which are stored. (Portfolio value is updated every 15 minutes now, and cash is updated after each transaction.) Net worth = cash + portfolio + shop items... and the way shop items are stored is causing the issue.
So for space and performance reasons I've left net worth off of the rankings (memberlist) page for now.
However, I already started a "two line" display with the trade count and the churn rate, along with the post count and posts this month. So I could try squeezing net worth in underneath something else, perhaps. I just can't put a sort on it. If the code for the shop had used a proper data model  then I could do something with it, but your items are stored as a delimited text string, and that doesn't lend itself to doing efficient calculations on the fly.
Or we could remove something, and replace it with net worth instead of trying to squeeze in a second line under something else. |
|
| Back to top |
|
|
| Dave Rathbun |
| Posted: Sat Apr 15, 2006 2:37 pm Post subject: |
|
|
CFO
|
| Dave Rathbun wrote: | I have made a change to the portfolio update process.
So the process now does two things. First, it counts the number of valid quotes returned by Yahoo. If that number is zero, then the process is abandoned. It will try again in 15 minutes. |
I looked over the logs for this week, and there were 4 instances where quite a few people saw -100% (or lower) returns. As designed, when this happened, the background process dumped out the quote values returned by Yahoo for debugging purposes.
It seems that what is happening is that Yahoo is returning partial results for the quotes. For example, in one case it should have returned quotes for 867 different stocks, and I got quote data for only 520 of them. So it's not a complete failure, but a partial failure. I can check for this in the code.
As said earlier in this topic, if you wait 15 minutes odds are good that the next run will be fine. All four errors occurred on April 13 (was that Friday? ) and interestingly two of them were at 10 minutes before the hour, and two were at 5 minutes after the hour. I don't think that was significant. |
|
| Back to top |
|
|
| Dave Rathbun |
| Posted: Sat Apr 15, 2006 2:51 pm Post subject: |
|
|
CFO
|
I have made a change to the quote engine. This is the same set of code that is called from nearly everywhere on this site, not just the 15-minute portfolio update process. This means that it could impact trading, your "my return" page, the portfolio page, essentially everything.
The good news is that I have tested it today, and it seems to be working okay. But the code change is related to getting errors from Yahoo, and I cannot simulate those. The changes are designed to check each packet of quote data returned, and verify that I got as many quote rows back as I sent symbols out. If the numbers do not match, then the process sleeps for a few seconds (to let whatever network issue(s) expire) and then tries again. It will retry each bundle 3 times before giving up and moving on to the next group.
So while it is still possible that we may see the -100% returns, I am hoping that this code change will make the system more robust. I will continue to monitor the log files and see how this works out. |
|
| Back to top |
|
|
|
|
 |
| jdunn72 |
| Posted: Mon Apr 17, 2006 7:21 am Post subject: |
|
|
Member of The Month
March
Joined: 26 Jan 2006
Posts: 1407
This Month: 0
11661.05 e$
Net worth: 11,661.05
Portfolio Value: 0.00
Monthly Return: 0.00%
Trades this month: 0
Churn Rate: 0.00%Items
|
| Dave Rathbun wrote: | | But number two on the list is nickspcs, with only one trade. | He's not there no more....
| Dave Rathbun wrote: | Number three is jdunn72, with 55 trades but only a 27% churn rate. As I posted earlier, I'm not sure the formula for churn is working the way I think it should, but this implies that jdunn72 has turned over 27% of his portfolio this month. That would mean that if he started with 100K he has bought, sold, and bought again using about 27K of his funds. Or sold - bought - sold - bought - sold ... since he has so many trades (55) it could be that he has 25 buys and sells, each for about 2K or so. Anyway, the numbers certainly do raise some questions, and might make the "sneak peek" feature more attractive .................................. BTW, the list is also sortable by monthly trades and churn percent. jdunn72 is leading the active traders with 55, PDJ Enterprises is second with 38 (but a 65% churn) and vetelmo is third on the active trader list with 39 trades. Fourth on the list is shaks who has a 36% churn rate and a -9% return rate.
crazycow is a hair behind Grim in the total return race, and has a 100% churn on only 4 trades.
Interesting stuff. | It is very interesting stuff. I like this a lot. As far as the math goes I can tell you that, I try to keep about 25-30% of my worth available for taking secondary positions to support my primary positions and I buy and sell these for additional profits and or repositioning strategies so, your numbers are jiving for me. Occasioanlly I will buy new positions and sell out of current holdings as soon as they no longer meet my criteria (which has nothing to do with gains or losses but, is modelled after my initial Experimental Portfolio thread that I started here). Thanks to the vision of einvesting and your coding I have finally developed a trading strategy that may well prove successful, for me. Again thanks and keep up the good work of course you've earned a well deserved break and everything looks great. Congratulations on a job well done. |
|
| Back to top |
|
|
| jdunn72 |
| Posted: Mon Apr 17, 2006 7:29 am Post subject: |
|
|
Member of The Month
March
Joined: 26 Jan 2006
Posts: 1407
This Month: 0
11661.05 e$
Net worth: 11,661.05
Portfolio Value: 0.00
Monthly Return: 0.00%
Trades this month: 0
Churn Rate: 0.00%Items
|
| Dave Rathbun wrote: | | I'll admit the reason I wanted to put in the monthly trades and the churn rate was to see if my assertion that active traders will now have an advantage in the game turns out to be true. |
Well, it may turn out to be that the best active traders will do better and that will show applied knowledge, hard work and diligence (which almost always takes the lead in any endeavor) pays off. Also, it would be equally interesting to see a list of the top twenty most active traders and the top twenty least active traders. That way you can have a real perspective and take an average of the percentage gained from each list, then you will have a more accurate representation to the truth in this matter, personally I think it will be about equal but my curiosity is piqued. |
|
| Back to top |
|
|
| Dave Rathbun |
| Posted: Mon Apr 17, 2006 7:53 am Post subject: |
|
|
CFO
|
| jdunn72 wrote: | | Also, it would be equally interesting to see a list of the top twenty most active traders ... |
You can do that now, sort of. If you click the Rankings page, and set your sort to Descending order by Monthly Trades you can get the stats for the 20 most active traders. As I write this, 14 of the 20 most active traders are negative. Of the 6 positive, only 3 have a return of 4% or higher for the month.
I can't do the same for the "least" active traders as there is a billion-way tie for first place (I'm exaggerating a little) with all of the folks with zero trades. I think what you probably were looking for was a listing of folks with at least one trade, and I can't generate those numbers using the screens that are available, I would have to write that query on the database directly. |
|
| Back to top |
|
|
| jdunn72 |
| Posted: Mon Apr 17, 2006 8:02 am Post subject: |
|
|
Member of The Month
March
Joined: 26 Jan 2006
Posts: 1407
This Month: 0
11661.05 e$
Net worth: 11,661.05
Portfolio Value: 0.00
Monthly Return: 0.00%
Trades this month: 0
Churn Rate: 0.00%Items
|
| Dave Rathbun wrote: | | jdunn72 wrote: | | Also, it would be equally interesting to see a list of the top twenty most active traders ... |
You can do that now, sort of. If you click the Rankings page, and set your sort to Descending order by Monthly Trades you can get the stats for the 20 most active traders. As I write this, 14 of the 20 most active traders are negative. Of the 6 positive, only 3 have a return of 4% or higher for the month.
I can't do the same for the "least" active traders as there is a billion-way tie for first place (I'm exaggerating a little) with all of the folks with zero trades. I think what you probably were looking for was a listing of folks with at least one trade, and I can't generate those numbers using the screens that are available, I would have to write that query on the database directly. |
Not, exactly; I should have been more clear..... based on portfolio returns the top twenty most actives(with the highest returns) and top twenty least actives(with the highest returns). You see what I mean now? |
|
| Back to top |
|
|
| Dave Rathbun |
| Posted: Mon Apr 17, 2006 9:15 am Post subject: |
|
|
CFO
|
| jdunn72 wrote: | | based on portfolio returns the top twenty most actives(with the highest returns) and top twenty least actives(with the highest returns). You see what I mean now? |
I feel like an idiot, but no, I don't see what you mean.
You want me to sort the list by total return, and pick out the 20 most active and the 20 least active? And compare their returns? The 20 most active is easy, or at least I thought it was. I thought I was answering that before. You'll have to quantify "least active" better. Do you mean one or more trades, or is zero trades enough to qualify? If that's the case, then I can give you the twenty highest returns for those folks with zero trades, but I would have to write that query directly on the database.
For example, here are the top twenty return percentages for folks with zero trades:
54.61 38.67 21.35 20.25 20.08 18.68 17.24 15.15 12.46 12.33 11.79 11.03 10.88 10.71 10.43 10.09 9.05 8.93 8.65 8.37
The average of these 20 is 16.5375%.
Here are the returns for the top twenty active traders, sorted in order by return (but queried by order of trading activity)
19.66 8.52 5.88 3.45 0.76 0.52 -0.5 -0.67 -0.68 -1.85 -3.43 -3.63 -4.61 -5.29 -5.65 -5.69 -5.77 -7.83 -8.78 -10.32
The average of these is -1.2955%
So sorting the returns by trading activity shows a definite difference.
Now if you're suggesting picking the top forty returns, and sorting those by most / least activity... well, there are only 10 active traders of any kind (one or more trades) on the top 40 returns.
Maybe you can give me a more specific example of what you're looking for, because I suspect the fact that I am closer to the data is blinding me to what you're trying to accomplish. |
|
| Back to top |
|
|
|
|
|
| |
stock market game on the web. At eInvesting you compete for monthly cash prizes while you interact, make virtual dollars, purchase items, and trade in the realistic stock market simulator. We make investing FUN!
