StockFetcher Forums · General Discussion · EVALUATING LARRY CONNORS' ETF TRADING STRAEGY<< 1 2 3 4 >>Post Follow-up
Kevin_in_GA
4,538 posts
msg #86481
Ignore Kevin_in_GA
modified
1/18/2010 9:48:13 AM

Just thinking out loud here a bit ...

If you want to follow Larry Connors' trading methods, which several people here already do and others have expressed interest in learning more about, it would be helpful to set up a filter to tell you exactly what to buy, when to buy it, and when to exit. So let's try!

Connors rules for High Probability ETF trading are pretty simple, and should be easy to implement into SF code.

1. Buy only ETFs above their 200 day moving average. This is easy -

Fetcher[
etf
]



This gives 1241 hits (probably includes ETNs as well). A large universe from which to trade. We need to narrow it down a bit.

Fetcher[
etf
close above MA(200)
]



1145 hits. Still too large. We will also want to trade highly liquid ETFs, and have the trading volume large enough so that our purchase does not affect the price.

Fetcher[
etf
close above MA(200)
average volume(50) above 100000
]



Now the trading pool is down to 331. Adding in the volume helps. This is probably narrow enough for now.

2. Buy on a pullback. Connors uses several systems for determining this. The most well-known is, not surprisingly, when the RSI(2) closes below 2.

Fetcher[
etf
close above MA(200)
average volume(50) above 100000
RSI(2) is below 2
]



That reduces the number of hits to 11 - very manageable. At this point, it would be prudent to look at these candidates to see if there was any specific news or issue that affected the stock price that might be hard to recover from in the short term. If there is no dramatic price drop (remember, these are ETFs and therefore should not have violent price swings unless something is amiss), consider these as good candidates.

3. If the price drops after you buy in, add a second unit at the lower price. I would personally rather avoid this - so why not look for ETFs that were candidates 1 day ago that have dropped in price today?

Fetcher[
etf
close above MA(200)
average volume(50) above 100000
close below close 1 day ago
RSI(2) 1 day ago below 2
]



This gives only 1 hit for today. If the basic premise is valid, this variant must also be valid (and potentially more profitable since you have bought in at a lower price). But this may be too restrictive for most traders, and will likely miss some really good trades.

4. Exit when the RSI(2) crosses above 70. I would rather see this as "exit when the RSI(2) crosses back below 70 from above" since a strongly trending stock can stay at high RSI levels for a while - no reason to exit prematurely. No real way to code for both entries and exits in the same filter, but I will add the following lines

Fetcher[
etf
close above MA(200)
average volume(50) above 100000
close below close 1 day ago
RSI(2) 1 day ago below 2

draw RSI(2) line at 70

add column RSI(2) {RSI(2)}
sort on column 5 ascending
]



The "sort on column 5" line insures that the ETFs will be ranked from lowest RSI(2) to highest.

The other approach Connors advocates is to exit once the price crosses the MA(5). Same problem as above for building in both entries and exits in the same filter, but I can at least visualize this for backtesting and profit estimation in the following way:

Fetcher[
etf
close above MA(200)
average volume(50) above 100000
close below close 1 day ago
RSI(2) 1 day ago below 2

draw RSI(2) line at 70
draw MA(5)
draw price line at MA(5)

add column RSI(2) {RSI(2)}
add column MA(5) {MA(5)}
sort on column 5 ascending
]



Pretty simple. Four basic rules, which can to a great extent be programmed into a simple Stockfetcher filter. This results in what is really a four line filter (the first four lines are all that is used to select stocks).

Backtesting the above:

Now let's see how well this works in a quick set of backtests. Since I do not have the higher level subscription here at SF, I can only backtest for four months at a time. You can use the offset function and manually review past trades - it is the most accurate approach, but is time-consuming and boring. For instructional purposes, let's just go with the idea that you buy at the open of the next day, rather than at the close as some folks do.

For these backtests, no stop losses are used (as Connors advocates). Trade size is $2000 per trade (50 open trades max). Exit is when RSI(2) crosses 70 from above.

Period 1: 9/15/09 to 1/15/10

Trade Statistics
There were 57 total stocks entered. Of those, 56 or 98.25% were complete and 1 or 1.75% were open.
Of the 56 completed trades, 45 trades or 80.36% resulted in a net gain.
Your average net change for completed trades was: 2.73%.
The average draw down of your approach was: -2.07%.
The average max profit of your approach was: 3.98%
The Reward/Risk ratio for this approach is: 16.39
Annualized Return on Investment (ROI): 91.84%, the ROI of ^SPX was: 24.10%.

Period 2: 5/15/09 to 9/14/09

Trade Statistics
There were 56 total stocks entered. Of those, 31 or 55.36% were complete and 25 or 44.64% were open.
Of the 31 completed trades, 26 trades or 83.87% resulted in a net gain.
Your average net change for completed trades was: 5.83%.
The average draw down of your approach was: -2.33%.
The average max profit of your approach was: 7.99%
The Reward/Risk ratio for this approach is: 36.32
Annualized Return on Investment (ROI): 193.38%, the ROI of ^SPX was: 52.19%.

Period 3: 1/15/09 to 5/14/09

Trade Statistics
There were 13 total stocks entered. Of those, 13 or 100.00% were complete and or 0.00% were open.
Of the 13 completed trades, 5 trades or 38.46% resulted in a net gain.
Your average net change for completed trades was: -0.38%.
The average draw down of your approach was: -3.53%.
The average max profit of your approach was: 3.82%
The Reward/Risk ratio for this approach is: 0.60
Annualized Return on Investment (ROI): -11.76%, the ROI of ^SPX was: 18.27%.

Period 4: 9/15/08 to 1/15/09

Trade Statistics
There were 3 total stocks entered. Of those, 3 or 100.00% were complete and or 0.00% were open.
Of the 3 completed trades, 2 trades or 66.67% resulted in a net gain.
Your average net change for completed trades was: 13.18%.
The average draw down of your approach was: -1.65%.
The average max profit of your approach was: 16.80%
The Reward/Risk ratio for this approach is: 16.10
Annualized Return on Investment (ROI): 751.45%, the ROI of ^SPX was: -95.98%.

NOTE: The entire gain was from one trade on DTO that netted 40%. The others were at the 2-3% range.

Summary

This system basically performs exactly as advertised - hit rate of about 70-80%, but average gain per trade is around 3%. On a $100,000 trading account (the same size as the backtest) this means you only net about $60-70 on any trade (of $2000 each). After commissions and assuming some slippage, this may yield as little as 1% net.

Thoughts and comment welcome.

Kevin




drew9
171 posts
msg #86483
Ignore drew9
1/18/2010 10:39:28 AM

Nice analysis Kevin. I have been using his strategy for awhile now to quite good success. I have varied it a bit and thought I would pass on my enhancements. I do have the upgraded subscription and did so just to test these out over a longer period. First, in order to enhance the returns, I chose only those leveraged ETF's - 2 and 3 times leveraged. Second, since the trades are not that frequent, I reduced the number of trades to 13. This gives a nice gain yet still allows for decent diversification. Third, I changed the MA to MA(100) as it allows for more trades yet still provides the needed ability to not trade in a downtrend. This backtested better than the MA(200) as well. Fourth, use RSI(2) of 75 for the exit. Does not change dramatically but offers a bit more profitablity. Fifth - do a backtest on each fund. I chose tried to choose a diversified list of different global indices etc. I narrowed it down to 20 for my list. It is important though once you have made these selections to do a backtest on each one. Doing this and keeping only those that perform best really enhances the results. This really opened my eyes as there was a huge variation as some performed quite poor. Sixth, I did manually test adding positions to each hit over the last year where the RSI(2) stayed below the trigger. I use RSI(2) of 15 for my entry. Adding a position each day it stayed below 15 nearly doubled the profitablity. This enhanced the winners and made nearly all losing initial positions profitable or close to it. The problem here is then figuring out the trade size. I had planned on using 13 as mentioned but now this needs to be accounted for. No way to backtest this so it has to be done manually. I am now using 25 knowing I will be adding to positions. If I know you, you will have the backtests for this whipped out in no time! :)

duke56468
683 posts
msg #86484
Ignore duke56468
1/18/2010 10:48:44 AM

Kevin
Thanks for the filter. Are the period 1 and 4 dates the same??

Kevin_in_GA
4,538 posts
msg #86485
Ignore Kevin_in_GA
1/18/2010 10:57:07 AM

Typo on the dates - now corrected.

chetron
2,817 posts
msg #86488
Ignore chetron
1/18/2010 11:45:27 AM

last 2 year....

*******************

Approach Name: kevin on connors
Test started on 02/01/2008 ended on 01/15/2010, covering 493 days
Filter used:
etf
close above MA(200)
average volume(50) above 100000
close below close 1 day ago
RSI(2) 1 day ago below 2

draw RSI(2) line at 70
draw MA(5)
draw price line at MA(5)

add column RSI(2) {RSI(2)}
add column MA(5) {MA(5)}
sort on column 5 ascending



Trade Statistics
There were 113 total stocks entered. Of those, 112 or 99.12% were complete and 1 or 0.88% were open.
Of the 112 completed trades, 90 trades or 80.36%resulted in a net gain.
Your average net change for completed trades was: 1.86%.
The average draw down of your approach was: -2.19%.
The average max profit of your approach was: 2.51%
The Reward/Risk ratio for this approach is: 7.15
Annualized Return on Investment (ROI): 111.89%, the ROI of ^SPX was: -8.82%.

Exit Statistics
Stop Loss was triggered 0 times or 0.00% of the time.
Stop Profit was triggered 0 times or 0.00% of the time.
Trailing Stop Loss was triggered 0 times or 0.00% of the time.
You held for the maximum period of time (25 days) 0 times or 0.00% of the time.
An exit trigger was executed 112 times or 100.00% of the time.

Entry Filter
etf
close above MA(200)
average volume(50) above 100000
close below close 1 day ago
RSI(2) 1 day ago below 2

draw RSI(2) line at 70
draw MA(5)
draw price line at MA(5)

add column RSI(2) {RSI(2)}
add column MA(5) {MA(5)}
sort on column 5 ascending

Basic Setup
Name: kevin on connors
Approach Type: Long
Start Date: 02/01/2008
End Date: 01/15/2010
Benchmark Symbol: ^SPX

Exit Setup
Stop Loss: N/A
Profit Stop: N/A
Trailing Stop Loss: N/A
Minimum Holding Days: N/A
Maximum holding days: 25
Exit Trigger #1: rsi(2) crossed above 70

Extra Indicators
Entry Columns:
Show Performance After: after 2 days
after 5 days
after 10 days
after 25 days
after 40 days


Advanced Options
Selection Method: select by volume descending
Entry Price: open
Conditional Entry: No
Exit Price: open
Maximum Trades Per Day: 5
Maximum Open Positions: 50
Maximum Selected Stocks: All




*******************************
sorry, couldn't resist.......

********************************
Approach Information
Approach Name: kevin on connors
Test started on 10/01/2009 ended on 01/15/2010, covering 73 days
Filter used:
not etf
close above MA(200)
average volume(50) above 100000
close below close 1 day ago
RSI(2) 1 day ago below 2
close between .1 and 1
draw RSI(2) line at 70
draw MA(5)
draw price line at MA(5)

add column RSI(2) {RSI(2)}
add column MA(5) {MA(5)}
sort on column 5 ascending



Trade Statistics
There were 102 total stocks entered. Of those, 97 or 95.10% were complete and 5 or 4.90% were open.
Of the 97 completed trades, 67 trades or 69.07%resulted in a net gain.
Your average net change for completed trades was: 28.99%.
The average draw down of your approach was: -13.75%.
The average max profit of your approach was: 15.12%
The Reward/Risk ratio for this approach is: 13.54
Annualized Return on Investment (ROI): 987.88%, the ROI of ^SPX was: 26.02%.

Exit Statistics
Stop Loss was triggered 0 times or 0.00% of the time.
Stop Profit was triggered 0 times or 0.00% of the time.
Trailing Stop Loss was triggered 0 times or 0.00% of the time.
You held for the maximum period of time (25 days) 2 times or 2.06% of the time.
An exit trigger was executed 95 times or 97.94% of the time.

Entry Filter
not etf
close above MA(200)
average volume(50) above 100000
close below close 1 day ago
RSI(2) 1 day ago below 2
close between .1 and 1
draw RSI(2) line at 70
draw MA(5)
draw price line at MA(5)

add column RSI(2) {RSI(2)}
add column MA(5) {MA(5)}
sort on column 5 ascending

Basic Setup
Name: kevin on connors
Approach Type: Long
Start Date: 10/01/2009
End Date: 01/15/2010
Benchmark Symbol: ^SPX

Exit Setup
Stop Loss: N/A
Profit Stop: N/A
Trailing Stop Loss: N/A
Minimum Holding Days: N/A
Maximum holding days: 25
Exit Trigger #1: rsi(2) crossed above 75

Extra Indicators
Entry Columns:
Show Performance After: after 2 days
after 5 days
after 10 days
after 25 days
after 40 days


Advanced Options
Selection Method: select by volume descending
Entry Price: open
Conditional Entry: No
Exit Price: open
Maximum Trades Per Day: 5
Maximum Open Positions: 50
Maximum Selected Stocks: All


BarTune1
403 posts
msg #86490
Ignore BarTune1
1/18/2010 12:03:50 PM

It is what it is .... and I've had luck following the rules .....

And Drew, thanks for noting your observations .....

Not sure about the "universe of ETFs" selected ..... but personally I avoid commodity based ETFs like the plague because of their ability to run and because they are least effective at "reverting to the mean" .....

Country funds are susposed to be the best candidates (for reversion to the mean) .... and actually, I'm at work now, but I might be able to dig up a symbol list of country ETFs that might bolster performance (or help in risk management) .....

straken
469 posts
msg #86497
Ignore straken
1/18/2010 1:18:01 PM

Kudos Kevin,
I liked the way you built your filter as you detailed your synopsis for this strategy. You would apparently make a good instructor for a wide body of students with varying degrees of academic abilities.
Mary (Babelfish) 4 money, should take notice before attempting to explain the reductionist method of chaotic math.



drew9
171 posts
msg #86498
Ignore drew9
1/18/2010 1:24:02 PM

Sites for leveraged funds.

Not entirely complete but good places to start.

http://etf.stock-encyclopedia.com/category/leveraged-etfs.html

http://etfdb.com/type/equity/leveraged/

Some are thinly traded so beware as the spreads can be wide. Go for the highest volume in each category you select for your list.

fortyfour
185 posts
msg #86500
Ignore fortyfour
1/18/2010 1:39:09 PM

www.etftrends.com .... is good for general info

www.tickerspy.com ........ not an etf site but very good at trackin sub - industries

www.finviz.com ....

chose GROUPS....INDUSTRIES...CHARTS.... for the best presentation
of industries I have seen. You get the chart of the industry and with a
click.... the companies that make it up. IMHO much better than IBD industry tables.
Maybe slightly off topic...but...it may help to support an etf trading strategy if you
dont follow a large universe of etfs..


Havent read Connors book.......Does he have a lot to say about money management and how, or ,if
he sees a difference in money management vs. stocks and etfs?

Thanks..... Great thread....


drew9
171 posts
msg #86504
Ignore drew9
1/18/2010 2:10:30 PM

It seems as though Connors likes to show strategies and backtest them over a period of many years. He has about 15 different strategies in his books but this thread highlights what I think is the best one available - at least for my trading personality. He seems to want to hold back a bit for future books or to keep to himself the money management piece. He does talk about a scaling-in approach and how well it has worked for him and his peers but only enough to wet your appetite. Using this strategy he makes it clear though that stops hurt the performance. I also agree after trying various ones with my filter and it did not help the results. I think that was the hardest thing to overcome as well as buying even more on the dips with the scaling-in approach. Kind of goes against conventional wisdom but it works for these ETF's. I think the great thing about this strategy as Kevin mentioned is that it is a nice fit for the capabilities in SF!

StockFetcher Forums · General Discussion · EVALUATING LARRY CONNORS' ETF TRADING STRAEGY<< 1 2 3 4 >>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 2016 - 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