'BO IDX' stops out...7 losses in a row for this model.

'BO IDX' got stopped out shortly after the open. The low of the day on the tracking vehicle (IDX) so far is just 0.03 below the stop. However, the trading vehcile didn't stop out. This happens sometimes*. Once I noticed this, and that the EMDU09 had rallied back up to near breakeven, I closed the EMDU09 trade. Happy to take a smaller trading loss as compared to the tracking vehicle.

The Modeled position (IDX) posted a 0.80% loss, which is worse than the average losing trade return of -0.56%. The actual Trading position (EMDU09) posted a 0.18% loss (including commissions), 0.62% better than the model.

Perhaps the stop is just too tight or some of the filters need to be adjusted. I may take a stab at rebuilding this model.

* http://stbsmb.blogspot.com/2009/07/out-of-sync-modeled-vehicle-rutx-and.html :

"I wonder if the opposite has ever worked in my favor? Such that the tracking vehicle got stopped out, but the trading vehicle didn't. I do recall that within the past 6 months or so, a trading position get stopped out (within 15 minutes after a particularly volatile FOMC release) at a much better price than the modeled position did, perhaps by as much as 0.50% I have only recently begun to monitor how much of a difference there is between a position's modeled return and its actual trading return, and have discovered that on average its around 0.15% lower than the modeled return."

'BO IDX' generated a Long Entry Signal for Thursday's Close, 7/16/2007

'BO IDX' generated an Entry Signal for Thursday's Close, 7/16/2009. This model is also on the discontinuation watchlist. It's last 6 trades have stopped out. It's Money Managment Stop is small, so the stops are not so painful. However, six stops in a row suggests this model is no longer working, or the recent environment was just not conducive to it.

Here are some of the model's Performance Statistics:

• Model Tracking Vehicle=IDX
• Start Date=7/1/1994
• Number of Trades=80
• Average Trade Return=0.25%
• Average Winning Trade Return=0.68%
• Average Losing Trade Return=-0.56%
• Win Rate=65%
• RAWAL (Ratio of Average Win/Average Loss)=1.21
• Profit Factor=2.25

Disclosure: Long EMDU09.

Disclosure: The performance results shown above are for Model analysis purposes and do not include commission or slippage. The model is built on data from the IDX, but I trade the model with the EMD e-mini. Actual trading results from the EMD usually differ from the model results of the IDX, with the EMD showing somewhat weaker performance data. Nonetheless, I still trade this model with the EMD.

***remember this is an illustration of what i am trading and my thinking...it is not a recommendation for you or anyone else to buy or sell this or any other security***

Extreme Frustration, but keeping my chin up.

For the past few days I have been in the process of culling out models that have not been performing well over the past few years. On the surface this is the right thing to do. However 3 of the 4 recently discontinued/suspended models have performed very well over the past few weeks:

Narrow Bar #1 NDX: generated a trade a few days ago that made about 0.5%. I have suspended the Narrow Bar models due to a flaw in the logic.

NDX OU: This morning, I closed the NDX OU trade pre-market for a 0.22% loss. The model has performed poorly over the past few years and for all intents and purposes is declared dead. The after hours futures have gapped up 25 points, so that model would be sitting on a 2.0% gain due to be closed at 4pm Wednesday.

NDX 14th DOM: I discontinued this model 15 minutes before the close of today's markets. The initial concept was sound, but the logic filters no longer made sense to me and the model has been performing poorly over the past few years, so I discontinued it. This model would have generated an entry signal at 4pm today. Right now, with the gap up in the NQ futures, that model is up about 1.85%.

Oh, the models that I am keeping, and that have generated positions over the past few weeks have either been scratches for small gains, or have gotten stopped out:

BTBT NDX
BTBT RUT
Payrolls NDX
NDX 1st DOM
NDX Summer Rally
NDX 30th DOM

Right now I am very frustrated. My cognitive side still believes in the models that have performed poorly over the past few weeks, and still believes the models that I have discontinued should stay offline. However, my emotional side is very frustrated. This reminds me of the challenging dynamic I experienced from the summer of 2007 to the summer of 2008. Nearly everything I did was wrong. The entry signals I took often turned into losers, and the entry signals I ignored ofter turned into winners. The basis for ignoring those profitable signals was disbelief in robust models. This happened so often, that I became too scared to trade.

This current environment appears to be different. A string of poor timing where good models produce losers and bad/suspect models produce actual/potential winners. I am not going to change how I trade. Except tonight and tomorrow will probably be rooting for the gap up to be sold! Emotionally, I want to be right, cognitively I won't be changing a thing.

Chin up!


p.s. the gap up may not be all that bad...i have been building a longer term investment position in the FAS for my IRA and some custodial accounts...the FAS may be held for another 3-15 months, depending on how it performs...ideally i want it to go down more so i can build up my position...alas i cant control what it does, so if it decides to put in a run, i will have a partial position, which is better than nothing...

Disclosure: Long FAS.

***remember this is an illustration of what i am trading and my thinking...it is not a recommendation for you or anyone else to buy or sell this or any other security***

'NDX OD' also gets the boot: the model is not robust and needs to be rebuilt or discontinued.

'NDX OD' is based on a similar concept as the 'NDX OU', except that OD is a short strat and OU is a long strat. It too was placed on the model discontinuation watchlist back in May. I just performed a more detailed performance analysis of 'NDX OD' and like 'NDX OU' (http://stbsmb.blogspot.com/2009/07/ndx-ou-gets-boot-model-is-not-robust.html), it is also being taken offline. Rebuilding it is a possibility, but my intuition suggests it is dead.

So the PF of 3.07 during the Historical Build Mode was quite acceptable. During the Historical Test Mode, the PF was 1.78. This is definitely below my minimal threshold, but I decided to trade the model anyway thinking that perhaps this lower performance was just a mild setback or a basing period.

During the Real Time Trading Mode, the PF was 0.50. Anything below 1.00 represents negative performance. So the PF has been in a steady decline from the Historical Build Mode, to the Historical Test Mode, to the Real Time Trading Mode. This is also true for all the other stats shown in the table above.

Adios 'NDX OD'! Thankfully I peformed this analysis before getting into the next trade with this model.

'NDX OU' gets the boot: the model is not robust and needs to be discontinued.

At yesterday's close, 'NDX OU' generated an entry signal which I took. In May, I did a cursory review of my models performance over the past few years. This model, due to its poor performance was put on the watch list for possible discontinuation.

This morning, I did a much more detailed performance analysis of this model. I found the results to be so poor, that I closed the trading position pre-market, and have discontinued trading the model.

I find the Profit Factor (PF) statistic to be one of the most telling of a model's performance. During the Historical Build Mode, the PF was 3.68. This means that for every $1.00, the model booked as a trading loss, it booked $3.68 as a trading gain. In other words, for every 1 step back, it made 3.68 steps forward. For my style of trading, I may accept a PF>=2.00, but I really prefer a PF>=3.00.

So the PF of 3.68 during the Historical Build Mode was quite acceptable. During the Historical Test Mode, the PF was 1.64. This is definitely below my minimal threshold, but I decided to trade the model anyway thinking that perhaps this lower performance was just a mild setback or a basing period.

During the Real Time Trading Mode, the PF was 0.74. Anything below 1.00 represents negative performance. So the PF has been in a steady decline from the Historical Build Mode, to the Historical Test Mode, to the Real Time Trading Mode. This is also true for all the other stats shown in the table above.

I should have done this analysis prior to putting the trade on yesterday. Closing out the trade pre-market led to a 0.22% loss. The money management stop for this model is currently at 1.00%. So I sold it at a 0.22% loss to prevent the additional 0.78% loss. I don't know if this particular trade would have gotten stopped out, but I have lost my confidence in the model so I just took this minimal loss and have moved on.

The lesson here for me is to go beyond the cursory performance review of the models that are on the discontinuation watch list, and undertake a more detailed performance review like the one above. Not having done so prior to this morning cost me 0.22%.

Though this model is now discontinued, it can be brought back on line if it can be rebuilt. My intuition strongly suggests that this model is dead.

'Payrolls NDX' is an Express Train with NO STOPS today! The trade is still a loss though and the model remains suspect.

'Payrolls NDX' is an Express Train with NO STOPS today! However the trade is a still a loss and the model remains suspect.

The Modeled position (NDX) posted a -0.36% loss, which is better than the Average Losing Trades return of -0.63%. The actual Trading position (NQU09) posted a -0.53% loss (including commissions), somewhat worse than the model.

The position was close to getting stopped out, down about 1.5% at 1030 am. However it then based into lunch, rallied up to its gap down opening level, hit a new high for the day in the afternoon, and closed just below that high of the day. Today's action saw the NDX carve out a nice bottoming tail on it's daily chart.

However, the trade is still a loss and the model remains suspect. I will have to re-evaluate the robustness of this model and make an adjustment to my trading plan:

• reduce the position size for future entry signals
• rebuild the model
• stop trading the model

'Payrolls NDX' Generated an Entry Signal for Thursday's Close, 7/2/2009; however its robustness in realtime trading is unproven.

'Payrolls NDX' Generated an Entry Signal for Thursday's Close, 7/2/2009. This model is based on the behavior of the NDX on the day the Payroll report is released. It is due to be exited at today's (Monday's) Close unless its Money Management Stop is hit intraday.

This model of market behavior is unproven in real time trading. Prior to Thursday's Entry Signal, the model has produced only 18 trades within a 24 year window. The first 17 were used to build the model and trade #18 was the first one that was executed in real time. It was subsequently stopped out, and that was the only stop out in the 18 trade history. At the time of writing this post, the current trade is down about 1.25%.

This is one of the risks of system trading. Even when building a model with in-sample data, testing it on out-of-sample data, and approving it for trading (because the model appears to be robust), there is no guarantee the model has correctly identified a type of predictive market behavior that will generate profits going forward. If the model gets stopped out today, then we will have a situation where the only two stops in the 19 trade history of the model, are when it began to trade in real time. This does not inspire confidence. In the case of a stop out today, I will have to re-evaluate the robustness of this model and make an adjustment to my trading plan:

• reduce the position size for future entry signals
• rebuild the model
• stop trading the model

Here are some of the model's Performance Statistics:

• Model Tracking Vehicle=NDX
• Start Date=11/1/1985
• Number of Trades=18
• Average Trade Return=0.94%
• Average Winning Trade Return=1.39%
• Average Losing Trade Return=-0.63%
• Win Rate=78%
• Average Win/Average Loss=2.23
• Profit Factor=7.80

Disclosure: Long the NQU09.

Disclosure: The performance results shown above are for Model analysis purposes and do not include commission or slippage. The model is built on data from the NDX, but I trade the model with the NQ e-mini. Actual trading results from the NQ usually differ from the model results of the NDX, with the NQ showing somewhat weaker performance data. Nonetheless, I still trade this model with the NQ and when executed according to the plan, it has generated consistent profitability for me.

***remember this is an illustration of what i am trading and my thinking...it is not a recommendation for you or anyone else to buy or sell this or any other security***

'SHOE SPY' generated a Short Exit Signal

'SHOE SPY' generated a Short Exit Signal at the 4pm Close on Friday 6/19/2009. The tracking vehicle (SPY) took a 0.4% loss, though the trading vehicle (ESU09), took a larger % loss, perhaps due to the ESU09 trading above fair value or something, which obviously did not work in my favor. I imagine that over a large enough period of time, the periodic discrepancy between the SPY and the ES will even itself out.

So this questionable model booked another losing trade. I will have to revisit the following:

• stop trading the model


• rebuild the model


• reduce the position size even further for future entry signals

'SHOE SPY' generates a Short Entry Signal

'SHOE SPY' has generated an Entry Signal on the Short side for the Thursday Open of the SPY. 'SHOE SPY' stands for 'SHort Option Expiration SPY'. The trade is due to be Exited on Friday June 19 at the 4pm Close, unless the Money Management Stop is hit before hand.

This Strategy is based on the behavior of the SPYder's during Option Expiration week, and is built on data that goes back to 1993.

This Short biased strategy performed extremely well during the bear market of 2000-2002 (Trade #32-52 on the Equity Curve Chart above). Though it did hit an new Equity Curve high in January 2008 (Trade 78), it has since booked its largest drawdown, during a period when the S&P 500 dropped by nearly 50%! The model appears to be out of sync with the market behavior during Option Expiration week for the past 18 months. The model may be broken.

I used to trade this model with 100% of capital, but I reduced that to 60% of capital when I put the trade on this morning. As of about right now, the position is about 1% out of the money.

I may need to do one or more of the following:

• stop trading the model


• rebuild the model


• reduce the position size even further for future entry signals

Disclosure: Short the ESU09.

***remember this is an illustration of what i am trading and my thinking...it is not a recommendation for you or anyone else to buy or sell this or any other security***