Contents of TR-INI55.ZIP

This file includes:

weightmix2_jim_mom2.ini (fixed) - Dave Howarter , 03/05/2008, This is a modified version of the combination of 5 buy/hold funds chosen because they hold different asset classes so they give a smooth ride that Dave posted last December. He replaced hsgfx by prpfx because this combo gives better annual return with only slightly higher Mdd. For the last year it gave Ann = 19.7%, Mdd = 3.68%.

rebwtmix.ini, yearly_signals.ini (corrected) - Bruce Robinson , 03/06/2008, Slight correction to the yearly signals since it needs to alternate buys and sells. Bruce has maintained for a while that a daily rebalance such as WeightMix does is a good approximation of a periodic rebalance such as monthly or yearly. But, a periodic rebalance is straightforward in Amibroker, so he's been doing these models there. Some diverging long fund / bear fund models show a slight, but noticeable difference when periodically rebalanced at practical intervals such as quarterly or yearly. So, he thought for a while that it wasn't possible to model the general case with weighted mixes in Trade, but in doing the AB version, Bruce thought of a way to do it with a some unusual math manipulation in Trade. So, here it is. If interested, make sure that you read the comments at the beginning for instructions about how it is run - Also, here's a program to generate an end of year rebalance signal for RebWtMix.ini -

cashplus.ini - Terry Steffen , 03/15/2008, Here's nice way to earn 39% on your cash. This uses 3 treasury ETF's, BWX, SHY & SHV.

jnbndmm.ini - Terry Steffen , 03/16/2008, This is Don Bell's fine modification. No ST bond involved. This is strictly Bond to Mmkt.

mowinhol.ini - Terry Steffen , 03/17/2008, MoWinHol went long on Friday for its first trade of 2008. Back to Mmkt at Tuesday close.

jncashbnd.ini (updated) - Terry Steffen , 03/17/2008, Here's a version of the above for JeffNat. And another go at JNCash.

gqsig.ini (revised) - Terry Steffen , 03/23/2008, Here's a program to iterate for the next GQsig signal. The way Terry plans to use this program is to enter the values for each. Then manipulate one of the values,like gld, to see what gld will have to do to trigger a signal for any given ndx value. By playing with this, he can quickly determine if a signal is likely or not.

tltt.ini (revised) - Terry Steffen , 03/29/2008, A modest improvement. Here's one for trading gov't bonds. It works with TLT and GVPIX. The code comes from Don's ProJnk which started with Bill Barnard. This is a 1 day delay signal even though it says zero delay in the SPT. That is because you can improve the result if you are lucky enough to fire off a zero delay trade. There is code to separately change the buy and sell delays.

tlttr.ini, delta.ini (fixed) - Bill Barnard , 03/30/2008, Bill found that he had introduced two errors into it when he revised it for Don last year to incorporate separate Buy and Sell thresholds. Unfortunately, in the present case, the results from the corrected code are not as good as before, but perhaps with some changes in parameter values the good results could be regained.

ileev.ini - Rich Schiebel , 04/21/2008, Here's a buy-and hold consisting of 1 part EEV and 2 parts ILF: EEV has only been around since 11/1/07 - before that, Rich replaced it with a -2 warp of EEM. Here's the numbers: last 4 years: AnnRet 17.2%, Mdd 5.1% on 12/28/05, last year: AnnRet 21.3%, Mdd 4.0% on 08/16/07, since EEV inception: AnnRet 29.6%, Mdd 3.7% on 11/27/07.

tlttipbonds.ini - Mark Keitel , 04/25/2008, Mark has just added the bonds and the ranking part to it.

x4a.ini - Dave Howarter , 05/01/2008, After tonight's update this run bombed in the middle with no error message, just a bleep sound from DOS. Dave had this problem one previous time and recognized it was a conflict between file names in the last INI file processed, X4A.ini and some other file that had created a file with a name like X2. The problem is the other file had created an X2.fnu, as I recall, and this caused Trade to bomb with no error message. Dave had already changed the creation of x1.sig to x1joe.sig which solved the first such problem. So today Dave changed the x2.sig to x2joe.sig and x3.sig to x3joe.sig. Since this did not change the intent or the results Dave did not change the file name. Dave hopes that Joe Porinchak will forgive him for modifying his creation.

spdrs.ini - Terry Steffen , 05/01/2008, Here's another shot a sector rotation.

digdugb.ini - Terry Steffen , 05/03/2008, Here is a long short OIL funds signal. It has a 4 day delay on Buys and a 2 day delay on Sells.

sms.ini - Terry Steffen , 05/07/2008, This is Terrys version of Bill Kellett's strategy. Compare to Don Bell's.

jeffnat.ini (update), jn7.ini (updated) - Bob Stevens , 05/13/2008, This creates the updated famalies of JeffNat clones for FT gathered by Dale Dann.

ncasel.ini - Lynn Grubb , 05/23/2008, ranks a family you specify by a modified NCALPHA. It also maintains a portfolio of the 4 top ranked funds in the family. The command line is: TradeFR NcaSel.ini family index signal.

idxpc1.ini idxpc.ini - Terry Steffen , 06/01/2008, Here's the first crack at a signal using the Index PC, ixdpc.ini. This is a zero delay signal. This actually might be possible to trade because CBOE updates the ratio hourly, however ixdpc1.ini is a 1 day delay version.

idxsee.ini - Terry Steffen , 06/01/2008, Here is the updater for the Index Put Call ratio. This is combo signal using both the Idxpc1.ini and ISEE.ini. To trade it, you would use ISEE as the master signal and when it is on a sell, trade Idxpc for something to do. Idxpc is in the market less than 10% of the time.

isee_drh.ini - Dave Howarter , 06/05/2008, Dave has made a few modifications to Terry's ISEE.INI. First, he had a problem with the fact that it used HoldMinDays = 20 to achieve its good results. Dave did some tests and saw that it did not do nearly as well with shorter HoldMinDays. However, 20 trading days is almost a month and during the time while he was waiting for the results to materialize he would never know if the signal had started to fail or if it was just in one of its dormant periods. Therefore, Dave prefers to have exit methods that are based on the strength of the fund or some index, rather than just on time itself. So he tried some options and found that he could get as good or better results than the original if he used HoldMinDays = 1 and triggered the exit by a moving average crossover. Deve found that the best moving average crossover was the price of the traded fund crossing below its DEMA. A DEMA is a type of moving average that can be coded in Trade. Next Dave changed the long fund to ADRE which gave him better results than UMPIX. Then he noted that the long trading system was in cash quite a bit of the time when ADRE was trending upwards, so he added a buy condition that is based on an SMA crossover of the traded long fund, which gave a moderate improvement in the ANN and Mdd. With those changes Dave was able to get a little more than twice the Ann of the original ISEE.INI with only a slight increase in the Mdd. This gave him an Ann = 40.8, Mdd = -7.1 for the long trade and Ann = 62.0, Mdd = -7.1 for the long/short trade. Dave didn't change Terry's approach for the short trade.

bbmc_iwm_qqqq.ini - Dave Howarter , 06/09/2008, This this is similar and not quite as good as the bbmc3.ini Don Bell posted but without a JeffNat account so you can't trade PCGRX that you used in bbmc3.ini. Dave added AccuPairTrade code to switch between IWM and QQQQ, depending on which one was generally stronger and then let the BBMC signal trade with either IWM or QQQQ depending on which the AccuPairTrade chose as strongest. This produced better results in 2007 - 2008 and gave very consistently good results over the last 5 years. That combination gave Ann = 29.2, Mdd = 7.6 from 3/12/03 to 6/6/08. This was with delay = 1 trading so since all decision are "next day", it is easy to understand what to do.

tlttr.ini (revised) - Terry Steffen , 06/15/2008, TLTR.ini was discovered to include Holy Grail, which means you should not be using it. Here's a new version which does not include any ZigZag or H.G. Thanks to Don for making it more productive.

sigperf3a.ini, sigperf3b.ini - Bruce Robinson , 06/19/2008, These 2 scripts are by Bruce from 2002. I find that they were buried in a published .zip file and am pleased that they now see the light of day.

adhdg1.ini, adhdg21.ini, adhdg3.ini, adhdg.bat - Daniel Thurman , 06/20/2008, In Feb, Terry Steffan posted a trade file to hedge ADINX. Daniel has taken his work and tried to jazz it up. He is using 2 fnu files, hdg2b and hdg3s, b for buy (long) and s for sell (short). Hdg2B = 50% ADINX, 25% SLPIX and 25% UOPIX. Hdg3S = 50% ADINX and 50% UCPIX. The first file is titled adhdg1 and uses the ratio of ADINX/RUT-I. The second file is adhdg2 and uses the ratio of ADINX/Sp-CP. The third is adhdg3 and uses the ratio of ADINX/OTC-C.

complite8.ini dat8.bat - Daniel Thurman , 06/20/2008, One step further with the adhdg files. Next Daniel took 2 signals, adsig2, adsig3 and added bblong to the mix. He then played around with Bruce Robinson's complite.ini and weighted the sigs as follows: 0 = fn20s 3 = fn20l 1 = MM 2 = fn20l 2 = fn20l 1 = fn20l 3 = fn20l 0 = fn20s Fn20l = 50% ADINX, 25% UUPIX and 25% UOPIX MM = 50% ADINX and 50% MM (You never sell your ADINX) Fn20s = 50% ADINX and 50% UCPIX The file is complite8.ini and can be run with dat8.bat.

complite8c.ini dat8c.bat - Daniel Thurman , 06/22/2008, Here is a file that uses a pretty balanced attack. It is a Bruce R. complite file called complite8c.ini and you can run it using dat8c.bat. It uses 3 sgnals: BBLong - uses SPX HiLoVlc3 - uses NDX and VLE-I Adsig3 - uses ADINX and OTC-C The signals are weighted equally and are used as follows: 0 = FN20S l = MM 2 = FN20L 3 = FN20L 3 = FN20L 2 = FN20L 1 = MM 0 = FN20S FN20L = 25% each of MDPIX,OTPIX,SLPIX,UUPIX FN20S = 100% SHPIX. ProFunds does not have a 1x Emerge fund so UUPIX is used. You have to include Emerging Markets to take advantage of the developing world.

gdx5.ini - Terry Steffen , 06/28/2008, Here's a next day gold fund signal that has worked since 1/1/06. Prior to then, it didn't work. It requires the gold miners fnu file gdm-x which comes from AMEX.com You don't have to update it because the ini uses GDX after it was born.

roy-ash.ini - Terry Steffen , 07/04/2008, Here's an oversold signal that rarely ever Buys but it could happen Monday. It is triggered by the ratio of new highs to new lows using a parameter that Roy Ashworth uses. It's a next day signal. Its 8 for 8 since 1999 and produces a 12% return with only 5% market exposure. Of course the problem is it has only been tested 8 times. Warning: There is no stop loss built in so if RYMDX drops by more than 5.8% from the buy date...Sell.

roy-ash_dh.ini - Dave Howarter , 07/04/2008, In Dave's words: 1. The problem in 1994 and 1998 was that it went just under the threshold, then a little above it to create a buy, then down again, which meant you really didn't want to do those buys. I eliminated them by requiring the 0.131 threshold, but that could still be subject to other whipsaws in the future. So to try to minimize the whipsaws I added a condition that the SMA must reach a low level before the buy took place and then buy when a second higher level was reached. This might eliminate some future whipsaws. In testing trading Rut over the entire database it eliminated one unproductive buy so 16 of 17 buys were winners, although the Ann remained the same. This did not affect the RYMDX trading you were testing. I set the low level requirement at 0.128 because that was the lowest that did not lower the Ann with RUT trading. That is not much of a difference from 0.131 so it might be better to set it a little lower, to try to eliminate more whipsaws, at the expense of removing some good trades. Anything down to 0.122 eliminated two trades and reduced the Ann trading Rut from 8.0 to 7.3 with the same Mdd. This also shows that 0.128 is a "magic number", so I am not very convinced that small changes in levels are not just curve fitting. 2. I also noticed you used the RSI5 based on MDY for the sell condition when you were buying RYMDX. I tried changing the sell RSI5 to be based on RYMDX and it increased the Ann from 12.5 to 13.3.

isee.ini (latest version) - Terry Steffen , 05/25/2008, Here's a "new & improved" version of ISEE without the 20 day hold on either side of the signal. With a stop-loss fix, this one works all the way back the beginning of ISEE (4/1/02).