StockFetcher Forums · Backtesting Support · ROI-positive, chart-negative<< >>Post Follow-up
juha
52 posts
msg #47785
Ignore juha
11/4/2006 3:27:38 AM

Hi!
I'm unable to understanw Why I have positive ROI=531.69% at some filter, but performance chart for same filter is negative???? May be I don't understand the meaning of chart? Please, see screenshot at http://www.dx200.biz/sf/
filter itself is
Fetcher[set{ClOp,close - open}
set{x1,clop / open}
set{pct, x1 * 100}
set{y,count(pct is above 3, 40)}
set{pr1,price one day ago}
set{pr2,price two days ago}
set{pr3,price three days ago}
set{ss,ind(SPY,rsi(2))}
price more than 2% below pr1
pr1 more than 2% below pr2
pr2 more than 2% below pr3
y > 2
ss above 50
and price is above 20
and volume is above 50000
]



thanks


traderblues
195 posts
msg #47798
Ignore traderblues
11/4/2006 9:39:37 AM

The chart you are looking at only keeps track of positive vs negaative trades. If you look at your stats, only 31% of your trades are positive. So as you move along the chart, there are more negatives than positives, so the chart keeps going farther into the negative.

The bottom chart is the one that tracks the total equity. If you look at your chart, the equity ends up positive, but not by much. I would have expected to see a larger ending balance. A couple of days ago someone pointed out a problem with the chart. Supposedly it got fixed, so you might try re-running the backtest and see if the results are better in the bottom chart.


juha
52 posts
msg #47805
Ignore juha
11/4/2006 6:13:51 PM

traderblues, thanks for answer. I've re-run my filter and several other filters...same result, while I have good positive ROI the charts looks ugly, for example bottom 2 charst show NO gain at all, while ROI=531.69% :(
what is uncorrect, ROI or charts?
thanks


rharmelink
81 posts
msg #47813
Ignore rharmelink
11/5/2006 4:19:22 AM

More inconsistencies?

-- Given that the backtest covers a 2-year period and ends with $100,775.43, shouldn't the annualized ROI be 0.39%?

-- The text says the annualized ROI is 500%+ for the backtest and 10.74% for the benchmark. However, the performance chart indicates 0.76 and 9.88...


stockfetcher
975 posts
msg #47839
11/6/2006 8:36:52 AM

Hi,

Thanks for the feedback.

The Annualized ROI is a simple calculation of:

247 * (AVERAGE_GAIN_LOSS ) / ( AVERAGE_DAYS_HELD)

* Note: 247 is an estimate on the number of trading days per year.

In the above backtest you have the following stats (or at least the one in your account with the same name):

AVERAGE_GAIN_LOSS: 0.748 %
AVERAGE_DAYS_HELD: 0.3478

This gives:

247 * 0.748 / 0.3478 == 531.69


So assuming you were able to make around 3 trades every day over the entire year, with an average of .748 % gain, your annualized (non compounded) return would be 531%.

Obviously this is a simple measure but one which is typically included in backtesting.

The chart is indicating a number of aspects which help to shed additional light on your backtesting run. First, you'll want to reduce the maximum number of open positions to get a better feeling for what you might be able to do with your money.

Next, your approach has few trades over the entire period (in comparison with the above). The key one is that you're not actually able to make 3 trades every day. You've actually made 345 in the 504 days of your backtest - or - .68 trades every day.

Using a variation on the above Annualized ROI, you might more closely determine the percent gain by:

247 * 0.748 / (504/345) == 126.47%

or 252% for two years.


The bottom plot is indicating an overall compounded return on 100,000 (i.e. re-investing gains). Its also showing what you could do with 100,000 assuming you're allocating up to 250 open "slots" for trading. Given this run allocates 250 slots, the transaction size is $400. Since you have very few open concurrent positions, in actuality, of the 100,000 allocated, you're really working with $400/$1200 at any given point (i.e. you're really only looking to see where ~$400 ends up after the two years).

On the final day of this run, the total running/gain loss(compounded) is $1038.38 (or around 260% gain on 400 dollars).

So in the end, assuming you use the variation of the Annualized ROI on investing $400 across 2 years, you have:

$400 * 2.52 = $1008

Comparing with the compounded of $1038.38.



Thanks,
Steve
StockFetcher.com Support















glider
59 posts
msg #47850
Ignore glider
11/6/2006 6:02:17 PM

In my opinion, very little attention should be given to ROI. I'm not sure what useful purpose it serves. The most important numbers you should be looking at are as follows:
Average Net Change: Personally, this number needs to be at least 2% taking into account trading costs and slippage.
Percent Win: At least over 50%.
I find if these two are good, the other numbers tend to fall into place.

The bottom graph, Estimated Equity works fine for me. It assumes $100,000 investment and divides it by the maximum open positions. The way to use this is to determine your average investment and divide 100,000 by it. You then enter this number in the maximum open positions field.
For example: You buy $2,000 of each stock. 100,000/2,000 = 50.
50 is your maximum open positions.
To figure out how much you would make on this filer, calculate:
Per Trade Amount X Average Net Change X Number of Stocks
Example: 2,000 X 2.75% X 160 = $8,800.
The ending amount listed on the graph should be a little higher beacause of compounding plus of course the initial $100,000.
Hope this helps to clarify things a bit.

glider




juha
52 posts
msg #47857
Ignore juha
11/7/2006 12:51:18 AM

Steve, glider, thanks !
Now it's clear for me.


juha
52 posts
msg #47858
Ignore juha
11/7/2006 1:21:01 AM

Steve, are you going to implement chart for ROI? It would be nice...
or chart for "Using a variation on the above Annualized ROI, you might more closely determine the percent gain by:
247 * 0.748 / (504/345) == 126.47% " ?



StockFetcher Forums · Backtesting Support · ROI-positive, chart-negative<< >>Post Follow-up

*** Disclaimer *** StockFetcher.com does not endorse or suggest any of the securities which are returned in any of the searches or filters. They are provided purely for informational and research purposes. StockFetcher.com does not recommend particular securities. StockFetcher.com, Vestyl Software, L.L.C. and involved content providers shall not be liable for any errors or delays in the content, or for any actions taken based on the content.


Copyright 2018 - Vestyl Software L.L.C.Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus